News and Agenda Archive

News

Best Student Presentation Award for Jamal Amini

At the 2017 Symposium on Information Theory and Signal Processing (Delft, 11-12 May), organized by the IEEE Benelux Chapter, Jamal Amini received a best student presentation award. Congratulations!

More ...


Dielectric Shimming

Introductory article in ETV "Maxwell"

More ...


Rob Remis appointed Associate Professor

Per 1 January, Rob Remis has been promoted by the Dean to the rank of UHD (Associate Professor). Congratulations!

More ...


Two IEEE SPS Best Paper Awards

Selected for the 2016 IEEE Signal Processing Society Best Paper Award:
Cees H. Taal, Richard C. Hendriks, Richard Heusdens, and Jesper Jensen
“An Algorithm for Intelligibility Prediction of Time–Frequency Weighted Noisy Speech”
IEEE Transactions on Audio, Speech, and Language Processing, Volume 19, No. 7, September 2011

Selected for the 2016 IEEE Signal Processing Society Young Author Best Paper Award:
Ahmed Alkhateeb, Omar El Ayach, Geert Leus and Robert W. Heath, Jr.
“Channel Estimation and Hybrid Precoding for Millimeter Wave Cellular Systems”
IEEE Journal of Selected Topics in Signal Processing, Volume 8, No. 5, October 2014.

The awards will be presented at the Awards Ceremony at ICASSP 2017 in New Orleans, LA


Rob Remis wins STW Open Mind 2016 award

At their annual congres, STW awarded 5 grants (each 50 kE) to research teams to enable them to explore 'risky research' ideas. Martin van Gijzen, Andrew Webb and Rob Remis presented one of the winning proposals: an affordable MRI instrument based on permanent magnets (as opposed to superconducting magnets) for detecting hydrocephalus.

Short movie presenting the idea.

More ...


Vacancy: Team manager for Electrical Engineering Education (EEE)

The Faculty of EEMCS is creating a special team to fully focus on teaching using our unique and innovative ‘Delft method’. This method integrates practical and theoretical electrical engineering education and trains students to be hands-on, theoretically versed electrical engineers ready for a future career in science or industry.

We are looking for a team manager specialising in Electrical Engineering Education (EEE) who will be both a group leader and a teacher in his/her capacity as the role model of EE Education.

More information


Geert Leus named 2016 EURASIP Fellow

To recognize outstanding achievements in the broad field of Signal Processing, each year the European Association for Signal Processing (EURASIP) elevates a select group of up to maximum four signal processing researchers to "EURASIP Fellow", the Association's most prestigious honor.

The EURASIP Board of Directors (BoD) has awarded prof. Geert Leus as one of the 2016 Fellows, "for contributions to signal processing for communications".

The award consists of a certificate presented during the Opening and Awards Ceremony at EUSIPCO 2016, held in Budapest (Hungary) on August 30, 2016.

More ...


Best paper award for Geert Leus

IEEE Sensor Array and Multichannel Signal Processing Workshop, Brazil ("Stationary Graph Processes: Nonparametric Spectral Estimation" with Antonio Marques en Alejandro Ribeiro)


7 July 2016: Opening of CryoLab for Extremely Sensitive Electronic Measurements

The CryoLab of TU Delft's Faculty of EEMCS has been opened on Thursday 7 July by the dean Rob Fastenau. TU Delft scientists from the Tera-Hertz Sensing Group, Jochem Baselmans and Akira Endo, will be leading a team of young scientists and engineers working in the lab on astronomical instrumentation. The first instrument, DESHIMA (Delft SRON High-redshift Mapper), is being developed to be operated on the ASTE telescope in the Atacama Desert in Chile. The goal of the research is to create 3D charts of so-called submillimetre galaxies that, in contrast to 2D charts, also show distance and time.

The large number of superconducting detectors, and the advanced electronics developed at SRON, allows DESHIMA to map a very large volume of space at once. While Endo leads the development of DESHIMA, Baselmans will soon install the next cryostat for testing novel THz array antennas, that will enable his upcoming instrument MOSAIC to target multiple galaxies at once. In the future, the CryoLab is envisioned to also host new coolers from QuTech. Superconducting electronics used for astronomical instrumentation and quantum electronics have much in common, because they both push the limits of what can be observed.


Rob Remis elected best teacher at EWI

By student election (1700 votes), Rob Remis was elected as best teacher for Fac. EWI in 2016. A decade ago, Rob won already once the title 'Best teacher in EE'. This has now been extended to comprise the full faculty (EE, Mathematics, Computer Science). Later this year, Rob will compete for the title of 'Best teacher of TU Delft'.

The annual election is organised by the student associations of the Faculty (ETV, Christiaan Huygens), based on voting and written motivations.


New project "tASk-cognizant sParse sensing for InfeREnce" approved

STW project by Geert Leus

More ...


New project "Earlier recognition of cardiovascular diseases" approved

Atrial fibrillation (AF) is a progressive disease and associated with severe complications such as stroke. Early treatment of AF is of paramount importance as it inhibits disease progression from the treatable (recurrent intermittent) to the untreatable (permanent) stage of AF. However, early treatment is seriously hampered by lack of accurate diagnostic instruments to recognize patients who will develop new onset AF or progress to a severer form of the disease.

The goal of this project is to develop age and gender based, bio-electrical diagnostic tests, the invasive and non-invasive AF Fingerprint, which consists of electrical atrial signal profiles and levels of atrial specific tissue/blood biomarkers. In daily clinical practice, this novel diagnostic instrument can be used for early recognition or progression of AF by determination of stage of the electropathology. As such, AF Fingerprinting enables optimal AF treatment, thereby improving patients outcome.

The project is a collaboration between Erasmus University (Dept. Cardiology), VU Medical Center (Dept. Physiology), and TU Delft (Sections CAS and Bioelectronics), and will fund 4 PhD students.


New book by Amir Zjajo: Brain-Machine Interface

low-power analog front-end circuits for brain signal conditioning and quantization and digital back-end circuits for signal detection

More ...


Michel Antolovic granted PicoQuant Young Investor Award

On February 14, 2016, Michel Antolovic was granted the prestigious PicoQuant Young Investigator Award at Photonics West in San Francisco for his paper titled 'Analyzing blinking effects in super resolution localization microscopy with single-photon SPAD imagers. The paper shows the first localization super resolution images obtained with a SPAD camera. The analysis includes specific timing properties of fluorescing molecules in vitro with unprecedented accuracy thanks to one of the worlds single-photon fastest cameras that was created in the AQUA laboratory. The timing properties are aimed to be used for optimizing fluorophore blinking or separation of fluorophores, enabling multichannel super resolved imaging.


Sundeep's PhD Defense

On 25 January, Sundeep Chepuri defended his PhD thesis on "sparse sensing". Here are some pictures

More ...


Happy 2016!

Here are some pictures of the New Year Reception of the Microelectronics department

More ...


TU Delft Female Fellowship Tenure Track Openings

Academic openings at all professor levels

More ...


QuTech enters in collaboration with Intel

Intel and QuTech, the quantum institute of TU Delft and TNO, have finalised plans for a ten-year intensive collaboration, along with financial support for QuTech totalling approximately $50 million.

More ...


New STW project: "Good vibrations"

Today STW announced that Rob's proposal "Good Vibrations" in the Open Technology Program will receive funding. The project will utilize the power of so-called Krylov subspace reduction techniques and develop solution methodologies for wave field problems in complex media.

The project will fund 1 PhD student: Jorn Zimmerling.

More ...


CAS Outing a big success.

The CAS outing this year focused on socializing, team building and having a great time and it was a big success. The group went to Outdoor Valley in Bergschenhoek for a bit of serious fun. Thanks all for participating and a big hug for the organizers.

More ...


Sundeep Chepuri wins ICASSP Best Student Paper Award

The ICASSP paper "SPARSE SENSING FOR DISTRIBUTED GAUSSIAN DETECTION" by Sundeep Chepuri and Geert Leus won the best student paper award. This is quite a prestegious achievement. Congratulations!

More ...


Else Kooi Award ceremony at ICT Open

Professor Charbon, dr Daniele Raiteri and professor Nauta

The 2015 Else Kooi Award has been granted to Dr Daniele Raiteri for his scientific research on Technology-Aware Circuit Design for Smart Sensors on Plastic Foils. The Else Kooi Award is an annual award for young researchers in the field of applied semiconductor research conducted in the Netherlands. The award comes with a prize of 5,000 euros.

Raiteri has received the award during a special ceremony at the ICT.OPEN symposium on March 25th. The award was presented by the board of the Else Kooi Award foundation professor Nauta, chair of the foundation (TU Twente) and professor Edoardo Charbon. Edoardo Charbon from the microelectronics department of the EEMCS faculty holds the position of secretary of the Else Kooi Award Foundation.

Dr Raiteris research is focused on organic semiconductors. This emerging technology has specific features which severely complicate the design of circuits and systems, such as low transconductance, gain and speed, as well as high component variability. Dr Raiteri has devised several new solutions that have shown to be extremely robust to variability, achieving significantly better gain-bandwidth products in amplifiers and exceptional signal-to-noise ratios in voltage-controlled oscillators.

Photos by: Thijs ter Hart

More ...


Dony's PhD defense

Some pictures appear here!

More ...


New STW project: "SuperGPS"

Gerard Janssen acquired, with his colleagues Jeroen Koelemeij (VU Amsterdam, PI) and Christian Tiberius (CiTG), a new STW project called SuperGPS.

The project addresses the problem that currently, GPS is not sufficiently accurate and reliable to enable autonomous driving. The central question is: "How do we realize highly accurate and reliable positioning using extremely accurate time-frequency reference signals, distributed through hybrid optical-wireless networks?.

The project aims at a hybrid optical-wireless system for accurate positioning, navigation, and network synchronization, to complement or even replace satellite navigation technology. This system is accomplished through a terrestrial grid of radio antenna pseudolites, synchronized with extreme accuracy through the fiber-optic telecommunications network. The key deliverable of the project is a pilot demonstration of SuperGPS technology under real-life circumstances.

The technology will be developed with support and feedback from potential users in telecommunications (Royal KPN N.V.), mobility (TNO and Volvo), and Dutch high-tech manufacturers, as well as stakeholders from the scientific and R&D community, including the Dutch metrology institute VSL, the Dutch keepers of atomic time UTC, and physicists and astronomers in need of better time and frequency signals.

More ...


Best paper award for Amir, Carlo and Rene

Amir Zjajo, Carlo Galuzzi and Rene van Leuken won the Best Paper Award for the paper "Noise Analysis of Programmable Gain Analog to Digital Converter for Integrated Neural Implant Front End" at the International Conference on Biomedical Electronics and Devices (Biodevices 2015; Rome, Italy).

More ...


New STW project "From coil to antenna"

The STW project "From coil to antenna: development of innovative transmit array elements for MRI of the body at 7 Tesla" has been granted funding. Rob Remis is one of the applicants. The PI is Prof.dr. A.G. Webb (Leids Universitair Medisch Centrum), also Univ. Utrecht is involved.

More ...


L3SPAD honored

The STW HTSM "L3SPAD: A Single-Photon, Time-Resolved Image Sensor for Low-Light-Level Vision" program has received funding. The program is led by Edoardo Charbon.

Description

Low-light-level (LLL) image sensors have been receiving great attention because they have various applications ranging from fluorescence microscopy to automotive sensing, from safety monitoring to 3D vision for robots. Traditionally, however, LLL image sensors have been used for military purposes because of their prohibitive costs. The appearance of monolithic solid-state complementary metal-oxide-semiconductor (CMOS) processes for the design and fabrication of photon counting image sensors has paved the way to enable low-cost and high-performance LLL image sensors. In this project, we will realize a gated 1.3Mpixel photon-counting image sensor in a standard CMOS process. The target sensor, with high timing resolution, low noise, and high photon detection efficiency, is the perfect candidate to meet all these technical and cost specifications.

More ...


Best MSc student of TU Delft

This afternoon, Jorn Zimmerling won the competition for best MSc student of TU Delft of this year. Jorn was an MSc student of Rob Remis and Paul Urbach, and is now a PhD student with Rob at CAS.

TU Delft news article (in dutch)

More ...


QuTech appointed as 'national icon'

The Ministery of Economic Affairs has named 4 innovations as 'national icon'; QuTech is one of them. "National icons are innovations which generate future welfare and help to solve mondial problems." The icons will receive a national support podium, including a minister or secretary of state as ambassador.

In the Department of Microelectronics, prof. Edoardo Charbon and dr. Ryochi Isihara are 2 of the 5 EWI faculty members directly involved in QuTech.

More ...


MP Jan Vos visits PARSAX

On Friday 7 November, Jan Vos, MP for the PvdA, visited the TU Delft Climate Institute. The theme of the visit was climate change, TU Delft's research and the usefulness of and need for climate monitoring. The programme included a demonstration of cloud simulations in the Virtual Lab and a visit to the PARSAX radar. Thanks to the rain, it was possible to obtain good live measurements.


Board of Directors of EURASIP

On Sep 1, Alle-Jan was elected as incoming member of the Board of Directors of EURASIP, with a 4-year term starting 1 January 2015. EURASIP is the European Signal Processing association.

More ...


Stefan Wijnholds finalist for Christiaan Huygensprijs 2014

Yesterday, 25 June 2014, Stefan Wijnholds received an "honorable mention" as finalist for the Christiaan Huygensprijs 2014, rewarding the best PhD thesis work in ICT over the past 4 years. The awards were handed by the Minister of Education (dr. Jet Bussemaker).

After a tough preselection by each university, a list of 32 candidates at a national level were judged by the jury. Out of these, 4 finalists were nominated who received a certificate in a ceremony in Voorburg.

Stefan received the honor for his PhD thesis on calibration and imaging for the LOFAR radio telescope. He was a PhD student with Alle-Jan van der Veen in 2006-2010, while being employed by ASTRON in Dwingeloo.

More ...


New STW project for Rob Remis

Rob Remis was granted an STW project "Dielectric enhanced MRI". The main applicant of this project is Andrew Webb (Leiden Univ.), coapplicants are Rob Remis (CAS) and Bert-Jan Kooij (MS3). This will fund 2 PhD students in Delft.

The project aims to improve MRI imaging by inserting "bags" with dielectric materials between the magnets and the body. This should provide better illumination, in particular when using high-tesla fields. This has already been applied in practice but the effect is theoretically poorly understood. The project should provide the EM theory related to this case.

More ...


Millad's SAM'14 paper in final selection for award

Millad's paper "Application of Krylov based methods in Calibration for Radio Astronomy" has made it to the final round of the IEEE Sensors and Multichannel (SAM) 2014 student paper competition. This selection has been made based on ranking by the TPC members. The final poster competition is Sunday, June 22, 2014 in A Corua.

More ...


New project: DRIFT

The future SKA radio telescope will produce large amounts of correlation data that cannot be stored and needs to be processed quasi real-time. Image formation is the main bottleneck and requires order 350 peta-flops using current algorithms. Another bottleneck is the transportation of station data (samples) to the central location where they are correlated.

The project aims to reduce the transportation bottleneck by time-domain compressive sampling techniques, allowing the recovery of full correlation data from significantly subsampled antenna signals, and to introduce advanced algebraic techniques to speed up the image formation. Ideally, we would even skip the intermediate covariance reconstruction.

The project is funded by NWO in the "Big Bang, Big Data" program and is carried out in context of the ASTRON-IBM DOME project.

More ...


Prof. Bastiaan Kleijn in Delft for 2 months

Professor Bastiaan Kleijn is a part-time professor in the CAS group. He will be physically present in the period 1 May-1 July 2014.

His expertise is speech and audio signal processing. He will be collaborating with Richard Heusdens and Richard Hendriks.

More ...


New project on improved hearing aids

Richard Hendriks has acquired a new STW project aimed at improving the intelligibility of speech for users of hearing aids.

With a prevalence of about 11 %, severe hearing loss has become a serious problem in our society. While the current generation of hearing aids can be of a great help in certain situations, they generally are not able to provide the hearing-aid user a natural impression of the acoustical scene. An often-reported problem for hearing impaired people is the inability to understand speech in complex acoustical environments as well as the inability to localize sound.

Due to the development of wireless technology, it is possible to equip hearing aids with more powerful noise reduction algorithms to further increase the intelligibility. However, these more powerful multichannel noise reduction algorithms sacrifice naturalness of the sound environment, also when state-of-the-art binaural noise reduction algorithms are used.

This project aims at developing signal processing algorithms to help hearing aid users in these situations, by providing them a natural impression of the acoustical scene.

More ...


European Conf Antennas Propagation

TU Delft is platinum sponser and exhibitor at the EuCAP 2014 - The 8th European Conference on Antennas and Propagation, to be held at the World Forum in The Hague, The Netherlands, on 7 to 11 April 2014.

The Microelectronics (ME) department from the faculty of Electrical Engineering, Mathematics and Computer Science, includes research groups actively engaged on teaching and research in the field of antennas and propagation.

Located within the microelectronics department, the mission of the THz Sensing Group is to introduce breakthrough antenna technology that will revolutionize THz Sensing for Space based and Earth based applications. In the long term the research will enable multi Tera-bit wireless communications.


Alle-Jan van der Veen appointed EURASIP Fellow

The award is for contributions to array signal processing applied to communications and radio astronomy. In 2014, four researchers have been recognized as Fellow.

More ...


New CAS website

Today, CAS switched to a new website. Please enjoy, and let us know if something is not working right (contact: Alle-Jan van der Veen). CAS members can apply for a user account to maintain their own bio-page.

More ...


Georg Kail new postdoc at CAS

Georg Kail is a new postdoc at CAS, working with Geert Leus on distributed localization

More ...


Przemek Pawelczak new Assistant Professor

In July 2012, Przemek Pawelczak was awarded a VENI research grant from NWO. This grant (EUR 250k) allows the researcher to fund his own research for up to 3 years. The topic of the research is "Intelligent spectrum use in emergency networks", and it will explore statistical methods to guarantee quality of communication in Cognitive Radio Emergency Networks.

Following this, Przemek was appointed as Assistant Professor in the Embedded Software group and started in January 2013.

More ...


A new professor

On 26 Sept 2012, the Board of Directors of TU Delft has decided to appoint Geert Leus as Antoni van Leeuwenhoek Professor in the CAS group. This is a `personal' full professorship aimed to promote young, excellent academics to Professor at an early age so that they can develop their academic careers to the fullest possible extent

Agenda

Quantized Distributed Optimization Schemes; a monotone operator approach

Jake Jonkman

Recently, the effects of quantization on the Primal-Dual Method of Multipliers were studied. In this thesis, we have used this method as an example to further investigate the effects of quantization on distributed optimization schemes in a much broader sense. Using monotone operator theory, the effect of quantization on all distributed optimization algorithms that can be cast as a monotone operator was researched for two different problem subclasses. The averaging problem was used as an example of a quadratic problem, while the Gaussian channel capacity problem was an example of the non-linear problem subclass. A fixed bit rate quantizer was used in combination with a dynamic cell width, to analyse the robustness of distributed optimization schemes against quantization effects. In particular, we have shown that for practical implementations it is possible to incorporate fixed bit rate quantization with dynamic cell width in a distributed optimization algorithm without loss of performance for both problem classes.

Additional information ...


Msc Thesis Presentation

Multi-FPGA Interconnection Simulation

He Zhang

The scalable simulation of neuron communication needs a large amount of computing resources. The high throughput of data cause the high requirement of interconnection network. This thesis is aimed at the finding the proper multi-FPGA connection for the neuron network. First describe the characteristics of the network in terms of the topology, routing and flow control. To find the proper connection, analysis of the throughput for the different network with different traffic pattern by considering the hopcount and bandwidth are made. It shows that the multicast is a good solution. Based on the interconnection router architecture, a simulator is built to make a cycle accurate simulation in systemC and test different traffic pattern by unicast and multicast routing. To break the limitation of FPGA ports, the source synchronous serdes connection is built by using the primitive in the Xilinx FPGA. With the requirement of bandwidth, the possible solution of number of channels and the overhead are anaylsed.


Msc Thesis Presentation

Full-Custom Multi-Compartment Synaptic Circuits in Neuromorphic Structures

Xuefei You

Neuromorphic engineering, aiming at emulating neuro-biological architectures in efficient ways, has been widely studied both on component and VLSI system level. The design space of neuromorphic neuron, the basic unit to conduct signal processing and transmission in nervous system, has been widely explored while that of synapse, the specialized functional unit connecting neurons, is less investigated. In this thesis, a current-based phenomenological synapse model with power-efficient structures, consisting of efficient synaptic learning algor ithms and multi-compartment synapses, has been proposed. A vertical insight is given into the design space of spike-based learning rules in regards to design complexity and biological fidelity. Due to various biological conducting mechanisms, the receptors, namely AMPA, NMDA and GABAa, demonstrate different kinetics in response to stimulus. The designed circuit offers distinctive features of receptors as well as the joint synaptic function. A better computation ability is demonstrated through a cross-correlation detection experiment with a recurrent network of synapse clusters. The analog multi-compartment synapse structure is able to detect and amplify the temporal synchrony embedded in the synaptic noise. The maximum amplification level is 2 times larger than that of single-receptor configurations The final design implemented in UMC65nm technology consumes 1.92, 3.36, 1.11 and 35.22pJ per spike event of energy for AMPA, NMDA, GABAa receptors and the advanced learning circuit, respectively.

Additional information ...


Msc Thesis Presentation

A 32 x 32 Spiking Neural Network System On Chip

Ester Stienstra

In this thesis a prototyping system on chip of a 32 x 32 spiking neural network is presented. This network has been designed in UMC 65. In order to determine which neuron model to use three different analog CMOS neuron models are studied. One of these models is used in the network. The network consists of arrays of synapses and neurons, 32 synapses for each neuron. In order to be able to control all the synaptic inputs and read all the neural outputs, logic is presented that minimizes the number of pads needed, while maintaining controllability and keeping all the important information in the neural signal. Simulations are performed to determine the influence of several behaviors of the neuron and the synapse on the output of the network. Also a floorplan and place and route design for the chip are presented.


Msc Thesis Presentation

Source-Synchronous Interface with All-Digital Data Recovery

Shizhao Zhang

This thesis proposes a low-cost high-efficiency source-synchronous interface for high-speed inter-chip communication. The interface is composed of LVDS transceivers as external I/O buffers, and an alldigital data recovery, which can calibrate the received data phase to be aligned to the 90◦ phase of the received half-rate reference clock, for error free data sampling. The proposed data recovery adopts a fulldigital scheme, which uses time-to-digital converters (TDC) as phase acquisition, a digitally-controlled delay line (DCDL) to calibrate the phase, and a finite-state machine (FSM) as the control unit. Reference clock generated from phase-locked loops (PLL) or delay-locked loops (DLL) is not needed for the proposed data recovery. The interface is implemented in UMC 65 nm Low-leakage technology, with circuits designed at both transistor-level and RTL-level. The postlayout simulation shows the proposed interface works properly with data rates from 412.4Mbps to 1.25Gbps in all process corners. The total layout area is 688 ñm × 87 ñm, and the total power consumption is 16.74 mW.


CSI-EPT: Towards Practical Implementation

Jiying Dai

The information of electrical properties of biological tissues within human body can be useful information for diagnosis, tissue characterization, hyperthermia treatment planning and MR safety control. Contrast Source Inversion Electrical Properties Tomography (CSI-EPT) is an MR base imaging modality, which reconstructs the electrical properties of the object. However, this method has not been implemented in practice due to the following issues: 1) The exact transmit phase, which is a necessary input of regular CSI-EPT algorithm, is not available from MR acquisitions. 2) The incident field, another input of CSI-EPT reconstruction, cannot be measured directly. 3) The reconstruction loses sensitivity at those regions with low electrical fields. 4) The sufficiency of the two-dimensional configuration, which has been used in most published results as an approximation


Signal Processing Seminar

Thomas Sherson

Thomas is going to present his recent research.

Additional information ...


Network topology inference from graph stationary signals

Network topology inference from graph stationary signals

Antonio Garcia Marques (King Juan Carlos University)

We address the problem of identifying a graph structure from the observation of signals defined on its nodes. Fundamentally, the unknown graph encodes direct relationships between signal elements, which we aim to recover from observable indirect relationships generated by a diffusion process on the graph. The fresh look advocated here permeates benefits from convex optimization and stationarity of graph signals, in order to identify the graph shift operator (a matrix representation of the graph) given only its eigenvectors. These spectral templates can be obtained, e.g., from the sample covariance of independent graph signals diffused on the sought network. The novel idea is to find a graph shift that, while being consistent with the provided spectral information, endows the network with certain desired properties such as sparsity. To that end we develop efficient inference algorithms stemming from provably-tight convex relaxations of natural nonconvex criteria, particularizing the results for two shifts: the adjacency matrix and the normalized Laplacian. Algorithms and theoretical recovery conditions are developed not only when the templates are perfectly known, but also when the eigenvectors are noisy or when only a subset of them are given. Numerical tests showcase the effectiveness of the proposed algorithms in recovering social, brain, and amino-acid networks.

Additional information ...


MSc Defense of Boliang Xu

Packet loss concealment for speech transmissions in real-time wireless applications

Boliang Xu

Packet communication applications cannot guarantee correct delivery of every packet. Congestions and interferences in the network lead to lost packets. However, real-time applications require timely delivery of data or information and always tolerate packet loss to achieve this aim. When some speech packets are lost, packet loss concealment (PLC) is used to replace the missing speech.

In this thesis, after investigating packet loss characteristics in realistic wireless networks and problems in existing PLC algorithms, we propose a new PLC scheme named adaptive PLC, which is composed of three algorithms: odd-even interpolation, waveform similarity matching and silence substitution. Adaptive PLC adjusts the algorithm to use depending on loss situations. Odd-even interpolation recovers the loss by interpolating odd or even samples in a packet. Waveform similarity matching estimates waveform segments from correctly received or already recovered packets. Silence substitution just fills in the missing part by zeros.

The adaptive PLC achieves improvements in speech quality relative to each single PLC algorithm and other existing PLC algorithms.


Signal Processing Seminar

Jie Zhang

Jie is going to present his recent research.

Additional information ...


Signal Processing Seminar

Bahareh Abdikivanani

Bahareh is going to present her recent research.

Additional information ...


MSc CE Thesis Presentation

An Accurate System-Level Device Aging Assessment and Mitigation Simulation Framework

Evelyn Rashmi Jeyachandra

As technology scaling enters the nanometer regime, device aging effects cause quality and reliability issues in CMOS Integrated Circuits (ICs), which in turn shorten its lifetime. Evaluating system aging through circuit simulations is very complex and time consuming. In this thesis, a framework is proposed, which allows for the evaluation of long-term aging effects of ICs and the corresponding measures to counteract premature failure. The focus of this work lies in the abstraction of low-level aging models to system-level models, in order to facilitate swift high-level simulation, without any knowledge of underlying circuit dynamics.

Two major aging mechanisms, namely Negative Bias Temperature Instability (NBTI) and Channel Hot Carrier (CHC) degradation are considered for analysis. System-level aging management is performed with the prototype of a System-on-Chip (SoC) including a Management Unit (MU), which counteracts aging by employing Dynamic Voltage Scaling (DVS), Dynamic Frequency Scaling (DFS), and Adaptive Body Biasing (ABB). The simulation platform prototype is based on System-C and a 65-nm technology library. This SoC simulation computes path delay using characterized models, which represent the aged behaviour of individual circuit elements. Results show that the obtained values are within 2% of circuitlevel simulation values.

Furthermore, the System-C implementation has a shorter execution time with an approximate speedup of 15 times over conventional circuit simulators (e.g. Cadence NCSim).

Additional information ...


Microelectronics IoT Pitch

and Summer Drink

All ME-MSc’s and ME-employees are cordially invited to make a 2-minute pitch for an interesting and unexpected IoT application.

The format is free like the level of seriousness and feasibility are but there is a meaningful purpose as the event is meant to inspire the definition of technology integrating projects in the field of IoT. The pitches will be ‘graded’ by measuring the intensity of the applause. The pitch wil be followed by the yearly Summer Drink of the Microelectronics Department.


Signal Processing Seminar

Image domain gridding for radio astronomy

Bas van der Tol
ASTRON

Fast implementations of a "non-uniform FFT" operation

Additional information ...


Signal Processing Seminar

Pim van der Meulen

Pim is going to present his recent research.

Additional information ...


Signal Processing Seminar

Shahrzad Naghibzadeh

Shahrzad is going to present her recent research.

Additional information ...


Signal Processing Seminar

3D direction of arrival estimation of multiple audio sources with spherical microphone arrays

Despoina Pavlidi
University of Crete, Department of Computer Science, Heraklion, Crete, Greece

Abstract: Direction of arrival estimation plays a central role in numerous signal-processing applications, such as smart home automation, surveillance systems, etc. Until recently the research community was mainly interested in single-dimensional direction of arrival (DOA) estimation by deploying linear or planar microphone arrays. Nowadays the focus has turned also towards spherical microphone arrays, which enable the more accurate capturing of the acoustic wavefield, hence enabling two-dimensional DOA estimation, i.e., the azimuth and elevation of an active audio source. In this talk we will present our proposed methodologies for DOA estimation in the 3D space. Our first proposed method relies on energetic analysis. We estimate the sound intensity vector on selected time-frequency elements of the spectrum and post-process the estimates utilizing 2D histogram representations. We enhance our approach by applying beamforming around local intensity vector directions. We call our hybrid approach spatially constrained beamforming (SCB). Our second proposed method improves the performance of two grid-based approaches, namely the steered response power (SRP) and the multiple signal classification (MUSIC) algorithm, both formulated in the spherical harmonic domain. We propose to derive local DOA estimates from the power map for SRP and the pseudospectrum for MUSIC. From these local DOA estimates we form a 2D histogram that we process to derive the final multiple sources directions.


Signal Processing Seminar

Millad Sardarabadi

Millad is going to present his recent research.

Additional information ...


Signal Processing Seminar

Jamal Amini

Jamal is going to present his recent research.

Additional information ...


Signal Processing Seminar

Rob Remis

Additional information ...


DCSE/SIAM Seminar

Untangling nonlinearity in  inverse scattering with data-driven reduced order models

Vladimir Druskin
Schlumberger-Doll, Boston

We consider an inverse problem for the acoustic wave equation, where an array of sensors probes an unknown medium with pulses and measures the scattered waves. The goal is to determine from these measurements the structure of the scattering medium, modeled by a spatially varying acoustic impedance function.

Many inversion algorithms assume that the mapping from the unknown  impedance to the scattered waves is approximately linear. The linearization, known as the Born approximation, is not accurate in strongly scattering media, where the waves undergo multiple reflections before they reach the sensors in the array. Thus, the reconstructions of the impedance have numerous artifacts.  In this talk we show that it is possible to remove the multiple scattering effects from the data registered at the array, using a reduced order model (ROM). The ROM is defined by an orthogonal projection of the wave propagator operator on the subspace spanned by the time snapshots of the solution of the wave equation. The snapshots are known only at the sensor locations, which is enough information to construct the ROM. The main result of the paper is a novel  algorithm that uses the ROM to map the data to its Born approximation.  We develop the algorithm from first principles and demonstrate its accuracy with numerical simulations.


MSc thesis presentation

A Low-Complexity CMOS Receiver for UWB siqnals

Ernesto Huaman

Ernesto's MSc thesis presentation on localization using UWB and its implementation in CMOS


MSc CE Thesis Presentation

Digital Neuron Cells for Highly Parallel Cognitive Systems

Haipeng Lin

The biophysically-meaningful neuron models can be used to simulate human brain behavior. The understanding of neuron behaviours is expected to have prominent role in the fields such as artificial intelligence, treatments of damaged brain, etc. Several neuron models exist, which vary in a level of accuracy complexity, speed, etc. In this thesis, a general simulator is presented, which can implement Hodgkin-Huxley(HH) model, Integrate and Fire model and Izhikevich model in the same architecture in a real-time. The different neuron models can be selected in the simulator to evaluate various network configurations, such as the amount of the neuron cells in the network, properties of the neuron models and so on. The simulator communication cost grows approximately linearly with the number of the neuron cells. Similarly, implementation over multiple Field Programmable Gate Array(FPGA) devices is possible. At last, this simulator is synthesised and validated on a FPGA device. The pipeline is added to reduce resource cost and latency.


Signal Processing Seminar

Matthew Morency

Matthew will present his recent research findings.

Additional information ...


Conferences

2017 Symposium on Information Theory and Signal Processing in the Benelux

Additional information ...


Signal Processing Seminar

Bastiaan Kleijn

Bastiaan is going to present his recent research.

Additional information ...


Signal Processing Seminar

Geert Leus

Geert is going to present his recent research.

Additional information ...


Conferences

Eurosensors dead-line


Signal Processing Seminar

Jörn Zimmerling

Jorn is going to present his recent research.

Additional information ...


Signal Processing Seminar

Aydin Rajabzadeh

Aydin is going to present his recent research.

Additional information ...


Signal Processing Seminar

Andreas Koutrouvelis

Andreas is going to present his recent research.

Additional information ...


Professoren in de Arena

Professoren in de Arena: De bionische mens, van protheses naar upgrades

Wouter Serdijn, Just Herder, Harrie Weinans, Project March

Op 28 maart gaan drie hoogleraren, waaronder Wouter Serdijn, met elkaar in debat over 'de bionische mens'. Wat is er mogelijk en hoe ver kun, wil en mag je gaan? In drie korte minicolleges praten de heren u bij en worden ze vervolgens stevig aan de tand gevoeld door cabaretier, columnist en TU-docent Jasper van Kuijk. In de discussie die daarop volgt, wordt het publiek van harte uitgenodigd mee te doen.

De sprekers van deze avond zijn:

Just Herder - Professor of Interactive Mechanisms and Mechatronics

Harrie Weinans - Professor of Tissue Biomechanics and Implants

Wouter Serdijn - Professor in Bio-Electronics

Project March

Deze editie van ‘Professoren in de Theaterarena’ wordt georganiseerd i.s.m. het ‘Explore your Brain’ evenement van de TU Delft Library in het kader van het 175 jarig bestaan van de TU Delft.

Over Professoren in de Arena

In nauwe samenwerking met de TU Delft en de universiteiten van Leiden en Rotterdam zetten wij in een theatrale setting steeds drie spraakmakende hoogleraren op het podium rondom een actueel thema. Deze onderwerpen worden van verschillende kanten belicht, vanuit de harde wetenschap en/of maatschappelijke en ethische hoek. In een magazine-achtig format met korte colleges, stand-up colums wordt u bijgepraat en doet u mee in de discussie.

Locatie: Theatercafé, Theater de Veste

Additional information ...


Signal Processing Seminar

Complex factor analysis and extensions

Alle-Jan van der Veen

Additional information ...


Signal Processing Seminar

Reconstruction of a Novel Wheel-Rail Contact Signal for the Wheel Defect Detection

Alireza Alemi
3ME

Wheel Impact Load Detectors measure the rail response to monitor the railway wheel condition. Due to the wheel dimension, the mounted sensor on the rail makes a partial observation of the wheel. Therefore, using multiple sensors is common to cover the wheel circumference. To combine the partial information collected by multiple sensors, a fusion method is required. Normally, the features of the measured data are extracted and then the feature-level fusion is applied. This research proposes a fusion method at the data-level to reconstruct a new wheel-rail contact signal. This method transfers the data from the time domain to the space domain using the space relation of the sensor intervals, and the wheel circumference. The signal reconstructed is used for the wheel defect detection.


Microelectronics Colloquium

Microelectronics Department Colloquium

Daniele Cavallo, Vasiliki Giagka, Fabio Sebastiano, Rob Remis

On Wednesday March 15 the next Microelectronics colloquium wil take place, including four lectures by staff members.

Please register online by completing the form.

  • Vasso Giagka
    Flexible bioelectronic medicines

    Abstract: Bioelectronic medicines are the next generation of neuromodulation devices: small active three-dimensional neural interfaces able to modulate nerve activity by targeting a specific neural region. They aim to treat a number of conditions, such as diabetes and asthma, in a tailored (per individual) and reversible fashion, avoiding the side effects of conventional drug-based interventions (pharmaceuticals). They achieve so by recording signals from the respective nerves, extracting information and using it as feedback to electrically stimulate the neural region in a closed-loop manner.

    Current technologies for active implants have not yet managed to achieve the miniaturisation and integration levels required for the development of bioelectronic medicines. For such breakthrough devices, novel concepts need to be explored, developed, and tested.

    In this talk I will present my current activities as well as my vision on realizing the first flexible three-dimensional graphene active implant, for safe chronic neural stimulation and recording from the peripheral nerves.

  • Fabio Sebastiano
    Cryo-CMOS for Quantum Computing: does it work?

    Quantum computing holds the promise to change our lives by efficiently solving computing problems that are intractable today, such as simulation of quantum systems for synthesis of materials and drugs. A quantum computer comprises both a quantum processor and a classical electronic controller to operate and read out the quantum devices. The quantum processor must be cooled at cryogenic temperature in order to show quantum behavior, thus making it unfeasible to wire thousands of signals from the cryogenic quantum devices to a room-temperature controller.

    While this issue can be solved by placing also the electronic controller at cryogen¬ic temperature, which electronic technology is the best choice for its implementation? This talk will address the challenges of building such electronic controller, and answer whether a standard CMOS technology can be employed for the required analog and digital circuits operating at 4 K and below.

  • Daniele Cavallo
    Advanced Antenna Arrays for Modern Radar and Communication Systems

    Abstract: Several of today’s radar and wireless communication applications are shifting their operation to higher frequency to fulfil more demanding requirements on resolution, compactness and data rates. For this reason, there is a growing need to develop low-cost integrated circuit transceivers working at millimeter and sub-millimeter waves.

    However, on-chip antennas are currently characterized by very poor radiation efficiency and extremely narrow bandwidth. My approach of combining the concepts of connected arrays with artificial dielectrics will solve the inefficiency problem and enable high-efficiency on-chip antenna designs.

    Similar concepts can be also realized at microwave frequencies in printed circuit board, allowing for low-cost phased array antennas with state-of-the-art performance in terms of scan range, bandwidth and polarization purity.

  • Rob Remis
    Imaging with Waves

    We present an overview of our current wave field imaging and inversion research. Effective inversion strategies for important applications in Magnetic Resonance Imaging (MRI), nano-optics, and subsurface monitoring will be discussed. In particular, dielectric shimming (shaping of the radio frequency field in MRI) as well as inversion algorithms that determine the dielectric properties of various tissue types based on measured MRI data will be considered, and state-of-the-art model-order reduction techniques for large-scale wave propagation problems will be discussed as well.


Signal Processing Seminar

Ehsan Tohidi

Ehsan is going to present his recent research.

Additional information ...


Signal Processing Seminar

Shahrzad Naghibzadeh

Additional information ...


Additional information ...


Signal Processing Seminar


Signal Processing Seminar

Christos Tzagkarakis

Additional information ...


MS3 seminar

Capabilities and Research Activities at the University of Oklahoma Advanced Radar Research Center

Prof. Nathan A. Goodman
The Advanced Radar Research Center (ARRC) at the University of Oklahoma

The Advanced Radar Research Center (ARRC) at the University of Oklahoma consists of a vibrant group of faculty and students from both engineering and meteorology, focused on solving challenging radar problems and preparing the next generation of students. Through the collaborative nature instilled in its members, the ARRC has proven effective at developing synergy between science and engineering in the field of radar. The ARRC resides in state-of-art Radar Innovations Laboratory, a one-of-a-kind and unrivalled facility for radar research, development, and education. This 35,000-sqft facility includes microwave labs, advanced fabrication capability, and two anechoic chambers.

Bio Prof. Goodman: Nathan A. Goodman received the B.S., M.S., and Ph.D. degrees in electrical engineering from the University of Kansas, Lawrence, in 1995, 1997, and 2002, respectively. From 1996 to 1998, he was an RF systems engineer for Texas Instruments, Dallas, TX., and from 2002 to 2011, he was a faculty member in the ECE Department of the University of Arizona, Tucson. He is now a Professor in the School of Electrical and Computer Engineering and Director of Research for the Advanced Radar Research Center at the University of Oklahoma, Norman.


MS3 seminar

MS3 Master Event

Come to learn about our group and current Master Thesis Projects...

Additional information ...


Signal Processing Seminar

Thomas Sherson

Additional information ...


MSc SS Thesis Presentation

Enhancement of the Spatial Resolution for the Temperature Sensing System of the 7 Tesla Magnetic Resonance Imaging Scanner

Tariq Saboerali

The MRI scanner with an ultrahigh magnetic field of 7T not only increases the image resolution but it also increases the Specific Absorption Rate (SAR) of the patient. In other words, the body temperature of the patient increases due to the absorption of heat produced by the 7T MRI scanner. This is dangerous for the health of the patient. In order to ensure that the SAR level of the patient does not exceed the acceptable limit, the body temperature of the patient should be monitored during the scan with a spatial resolution as small as possible. This way safety measures can be taken immediately if the body temperature increases. In order to monitor the temperature during the MRI scan, fiber optic sensors (FOS) can be used. The fiber optic sensors (FOS) are immune from electromagnetic interference and there is no electrical connection to the patient and thus it is safe to monitor the temperature during an MRI scan by using FOS [1]. However, the FOS may have a spatial resolution which is not acceptable for medical purposes. This study focuses on methods to increase the spatial resolution of an existing fiber optic temperature sensing system of a 7T MRI scanner. To increase the spatial resolution of the existing temperature sensing system two methods are evaluated, namely the total variation deconvolution method and the blind deconvolution method. This study shows that the total variation deconvolution method gives the best results for the input temperature estimate. The blind deconvolution method strongly depends on the initial guess of the impulse response of the temperature sensing system, which is difficult to find. Therefore the results of the input temperature and the impulse response are less reliable when using the blind deconvolution method. Also it is shown that the machine resolution gets worse when increasing the spatial resolution by interpolating the input temperature in the Fourier domain.


Signal Processing Seminar

Elvin Isufi

Additional information ...


Signal Processing Seminar

Jeroen van Gemert

Strong magnetic background fields are of great interest in Magnetic Resonance Imaging (MRI), since images with high spatial resolution can be obtained at reduced scanning times. Strong background fields may cause RF interference effects, however, and these effects can severely degrade the quality of an MR image. This problem can be partly resolved using various advanced and mostly expensive techniques, but there is also a cheap and practical solution, namely, dielectric pads. The design of such a pad is not trivial. Normally, finding the “optimal” pad for a specific region of interest involves evaluating many different pad designs using electromagnetic field simulations. This is a very time-consuming approach taking hours to days of computation time. We propose a nonlinear optimization method based on model order reduction that allows us to design high-permittivity pads in less than 30 seconds.

Additional information ...


Signal Processing Seminar

Dong Yang

Digital Active Noise Cancellation

Additional information ...


MS3 seminar

Dynamic rainfall monitoring using microwave links

Venkat Roy

We propose a sparsity-exploiting dynamic rainfall monitoring methodology using rain-induced attenuation measurements from microwave links. To estimate the rainfall intensity dynamically from a limited number of non-linear measurements, we exploit the physical properties of rainfall such as spatial sparsity and non-negativity along with the dynamics of the rainfall. We develop a dynamic state estimation algorithm, where the aforementioned spatial properties are utilized as prior information. To exploit spatial sparsity, we use a basis function to tailor the sparse representation of the rainfall intensity. The developed methodology is applied to dynamically monitor the rainfall field intensity in an area with a specified spatial resolution using less number of simulated non-linear measurements than pixels. The proposed methodology can be generalized for any dynamic field reconstruction, where the limited number of non-linear measurements are field intensities integrated over a linear path.

Additional information ...


Signal Processing Seminar

Sundeep Prabhakar Chepuri

Additional information ...


Signal Processing Seminar

Transmitter and Receiver Optimization for MIMO Radar Systems

Tuomas Aittomaki
Aalto University, Finland

The spatial and waveform diversity achieved in MIMO radars can be beneficial for target detection and target parameter estimation, especially for low-observability targets. Taking advantage of the diversity requires optimization of both the transmitter and the receiver end. For the transmitter, it is necessary to encode the transmitted waveforms to have minimal sidelobe and cross-correlation levels. Also by appropriate spatial coding, the channel estimation error can be reduced. Furthermore, the transmit power allocation can be optimized. For the receivers, mismatched filters can be optimized to reduce jamming and clutter as well as the sidelobe and the cross-correlation levels for any Doppler frequency.


MSc CE Thesis Presentation

A Real-Time Hybrid Neuron Network for Highly Parallel Cognitive Systems

G.J. Christiaanse

For comprehensive understanding of how neurons communicate with each other, new tools need to be developed that can accurately reproduce and mimic the behaviour of such neurons in real-time. By using current complex mathematical models, simulated neurons are able to accurately approximate the behaviour of biological neural tissue. This comes at the price of computing complexity, resulting in responses that lag behind, and thus cannot interface with biological neurons.

The proposed design in this thesis, models an Inferior Olivary Nucleus network on an FPGA device, with a maximised amount of simulated neurons for the given FPGA family type. To achieve both accuracy and real-time speed, a complex biophysically meaningful mathematical model has been analysed and scheduled on a highly pipelined, and parallel running architecture design, specified within a SystemC specification. This has contributed to the creation of hybrid neuron network that executes optimally scheduled floating-point operations that, together with open source IP, has resulted in cost-effective solutions, capable of simulating responses faster or on par with their biological counterparts.


Microelectronics Introduction Colloquium

Introduction 3 new Tenure Trackers

Masoud Babaie, Morteza Alavi, Faruk Uysal

On December 12 we organize the next Microelectronics Colloquium to introduce three new Assistant Professors (Tenure Trackers) of the Microelectronics department. They are happy to present a lecture about their research.

The colloquium start at 15.00 hrs. there will be a drink afterwards in the foyer.
Location: Theatre of Culture Builing (38) Mekelweg 10.
Please register online if you want to attend, latest December 5.

  • Masoud Babaie: Pushing The Limits of CMOS Circuits for Emerging Technologies
    Within the next few years, quantum processors, Fifth Generation (5G) cellular systems and the wireless Internet-of-Things (IoT) are expected to see significant deployment to realize more integration between the physical and digital worlds, promising enormous computation power, high data rate communications and enabling more objects to be remotely sensed and controlled.

    This talk will address some of the main challenges in the design and implementation of IoT devices, mm-wave 5G transceivers, and cryogenic CMOS controller for quantum computers. An overview of my past and ongoing research activities will be also presented, with emphasis on novel solutions to improve power efficiency and spectral purity of RF/mm-wave transceivers.

  • Morteza Alavi: Universal Transmitters for 5G
    Today, our daily activities are intertwined with the Internet. The ever-growing demand to swiftly get access to the data-cloud systems leads to huge data traffic. In order to seamlessly transmit and receive these gigantic data, _ 40 GB, agile radio-frequency (RF) transceivers are inevitable.

    These radios must be capable of supporting the current and future communication standards such as 5th generation of wireless mobile communications. The ultimate goal is that they can be implemented as universal radios whose modes of operation can be defined by their clients. To address these demands, RF transmitters are currently reinvented and are directed towards digital-intensive architecture. In this short presentation, we will briefly describe the strengths,possibilities, and challenges that exist for these advanced transmitters. First and foremost, the concept of RF-DAC based transmitters will be introduced. Next, the talk will review various RF-DAC based transmitters that have already been implemented at ELCA. Eventually, the presentation will concisely unveil the future directions of the research of these software-defined transmitters at ELCA.

  • Faruk Uysal: Distributed Radar Networks: Beyond a single radar
    The number of operational radar is rapidly increasing due to the growing demand of the remote sensing. Software defined radio and emerging single-chip radar technology make use of radars in every aspect of life such as autonomous driving, safety and security applications. With the increase of active transmitters, spectrum management and coexistence started to become a concern for some radar systems. In this talk, the previous applications of waveform, frequency agility will be reviewed to bring multi-functionality to the modern radar system. Finally, we will discuss the future research for distributed radar networks and how to fuse data from various radars to acquire different aspects of a target to be viewed simultaneously.


Signal Processing Seminar

Jiani Liu

Additional information ...


Signal Processing Seminar

Accurate ranging and localization capabilities of narrowband ISM band radios

Jac Romme

In this presentation, the results are presented of a feasibility study on the ranging and localization capabilities of narrowband radio. Starting point of the study was the phase-difference (PD) principle used by Atmel’s AVR2151 chipset. To quantify the ranging performance in indoor environment, multi-channel VNA-based channel measurements have been conducted. The analyses revealed that PD principle is sensitive to multipath, even in the presence of a line-of-sight. To improve the accuracy, two super-resolution-based ranging algorithms (using Matrix-Pencil and Music) have been evaluated, which are shown to be considerably more robust against multipath. Additionally, the channel measurements have been used to quantify the benefit of antenna/polarization diversity for ranging. The diversity gave an additional, significant improvement and resulted in ranging with 0.5 meter accuracy in combination with the super-resolution algorithms using only the 2.4 GHz ISM band.

Additional information ...


PhD Thesis Defence

Gigabit Wireless Transmission in Dispersive Environments

Seyran Khademi

The advent of the digital era has revolutionized many aspect of our society and has significantly improved the quality of our lives. Consequently, signal processing has gained a considerable attention as the science behind the digital life. Among different applications for signal processing theory and algorithms, wireless communications remains one of the attractive and popular ones due to the widespread use of mobile devices.

This thesis is dedicated to develop signal processing algorithms to design highspeed wireless transceivers that can perform in highly reflective and harsh environments. The start of this research work initiated as a collaboration between TU Delft and an industrial partner, on a research aimed at short range gigabit wireless link within a lithography machine. The underlying unique wireless environment, together with the challenging specifications of the communication link for mechatronic systems, made this a compelling research project.

The first part of this research work focuses on constructing a reliable propagation model for dispersive environments, based on actual measurements. In our opinion it is crucial to have decent models to build effective theory and applications upon it. We developed a statistical channel model for the 60 GHz band for the extreme case of a confined metal enclosure in order to evaluate and test the existing signal processing algorithms under such pessimistic ambient conditions. This unique experiment opened up new research challenges to look back to popular design paradigms and reevaluate them with respect to the proposed channel model with a delay spread in the order of miliseconds. The concept of orthogonal frequency division multiplexing (OFDM) transmission was revisited and a customized OFDM system was designed which meets the data rate requirements of the mechatronic system of interest. The effectiveness of the proposed OFDM design was examined via Matlab simulations using the measured and modeled channels. Interestingly, the performance of the OFDM system is not heavily affected by the frequency selectivity of the extreme propagation environment. The loss is mainly due to the time guard that is dedicated to avoid interference between consecutive OFDM blocks, suggesting the use of longer OFDM blocks to minimize the bandwidth loss.

The second part of this thesis is dedicated to multiple-input multiple-output (MIMO) systems versus the single-input single-output (SISO) system which was studied in the first part. The emphasis is on general challenges in high speed (wideband) communication systems rather than the specific wireless link within a mechatronic machine. Challenging research questions are posed regarding the design and implementation of MIMO systems. This part starts with a brief introduction to such systems and redefining our system model with respect to the MIMO setting and it continues by revisiting the timely problem of peak-to-average power-ratio (PAPR) reduction in OFDMsystems, which deals with stochastic (data-dependent) OFDM waveforms, and the proposal of an effective algorithm to handle this challenge within the MIMO context . The hard problem of antenna selection for MIMO system was considered at the end by investigating different linear precoding designs subject to the realistic hardware constraints including per antenna power constraints (rather than conventional total power constraint) and limited number of RF chains.

Additional information ...


Signal Processing Seminar

Bilinear Inverse Problems: Bad News and Good News

Urbashi Mitra
University of Southern California, USA

A number of important inverse problems in signal processing, such as blind deconvolution, matrix factorization, dictionary learning and blind source separation share the common characteristic of being bilinear inverse problems. In such problems, the observation model is a function of two inputs and conditioned on one input being known, the observation is a linear function of the other. We will review important applications and challenges.

A key question is that of identifiability: can one unambiguously recover the pair of inputs from the output? We shall consider both deterministic conditions for identifiability as well as probabilistic statements that result in new scaling laws under cone constraints. We provide additional results specific to blind deconvolution and show, surprisingly, that adding the sparsity structural constraint is insufficient for signal identifiability suggesting that other strategies such as coding are necessary to achieve identifiability. However, there is hope that additional structure can help in certain cases. To this end, we discuss a novel strategy that exploits low rank matrix factorization to estimate parameters of a time-varying wireless channel.

Biography

Urbashi Mitra received the B.S. and the M.S. degrees from the University of California at Berkeley and his Ph.D. from Princeton University. Dr. Mitra is currently a Dean’s Professor of Electrical Engineering at the University of Southern California.

Dr. Mitra is a Fellow of the IEEE. She is the inaugural Editor-in-Chief for the IEEE Transactions on Molecular, Biological and Multi-scale Communications. She is a member of the IEEE Information Theory Society's Board of Governors (2002-2007, 2012-2017) and the IEEE Signal Processing Society’s Technical Committee on Signal Processing for Communications and Networks (2012-2016).

Her research interests are in: wireless communications, communication and sensor networks, biological communication systems, detection and estimation and the interface of communication, sensing and control.

Additional information ...


Signal Processing Seminar

Jie Zhang

Additional information ...


Performance evaluation of a promising candidate waveform for 5G networks (GFDM)

Amirhossein Mohammadian

The advent of next generation wireless communication system (5G), arises new requirements that clearly go beyond higher data rates. Orthogonal Frequency Division Multiplexing (OFDM) is a widely adopted solution, but has some major drawbacks, including high out-of-band (OOB) emissions, high power consumption and Low spectral efficiency. As a result of OFDM limitation in addressing requirements of scenarios presumed for 5G, investigation of new waveforms is necessitated. Generalized Frequency Division Multiplexing (GFDM) is a flexible candidate of non-orthogonal waveform for next generation wireless networks. GFDM has attractive properties and as a result has recently received a great deal of attention.

In this presentation, after introducing the GFDM structure, I am going to discuss the following titles which are related to its performance evaluation.

  • Derivation of PAPR distribution for GFDM modulated signal
  • Spectral analysis of GFDM modulated signal under nonlinear behavior of power amplifier (PA)
  • Rate optimization of GFDM-based cognitive radio networks
One of the major drawbacks of every multicarrier system is its high peak-to-average-power ratio (PAPR). It is very important to accurately identify PAPR distribution in GFDM systems to work out some effective measures to curb PAPR. To fulfil this purpose, the distribution of GFDM PAPR is expressed in terms of complementary cumulative distribution function (CCDF).

One of the most critical elements in all communication systems is PA. Nonlinear behavior of PA causes the expansions of spectral contents, called spectral regrowth. The power leakage of GFDM on the adjacent channels, caused by nonlinear PA is investigated by spectrum analyzing.

Due to the limited spectral resources in wireless communication systems, cognitive radio (CR) appears to be a promising approach. A rate optimization problem for GFDM under interference constraints is formulated and solved.


Signal Processing Seminar

Jamal Amini

Additional information ...


Signal Processing Seminar

Swarm Exploration under Sparsity Constraints

Christoph Manss
German Aerospace Center Institute for Communications and Navigation (DLR)

Robotic exploration aims at reconstructing (or understanding) an unknown physical process autonomously and as efficiently as possible. The Curiosity Rover on Mars exemplifies well a typical application of a robotic platform for extraterrestrial exploration. In many situations, however, the exploration domain is either vast and a full coverage of it is time consuming for one agent. Another issue is the robustness of the exploration mission, where a single robot represents a single point of failure. This is the reason why our group focuses on multi-agent exploration systems, called swarms.

The multi-agent systems coordinate their actions by cooperatively collecting measurements and jointly processing the acquired data, thus they split the computational complexity of the estimation, and risk of whole system failure, among individual agents. This talk will address a specific class of models for exploration, where the exploration area of interest is considered to be sparse. Sparsity implies that the explored process can be accurately represented with only a few relevant components. This assumption leads to optimization algorithms employing specific non-smooth regularizations. With the help of compressive sensing methods such assumptions can significantly speed up estimation and exploration. Specifically, it will be shown how compressive sensing methods are used to estimate model parameters in a multi-agent setting and how optimal agent trajectories (with respect to the efficiency of the estimation of an unknown process) are computed.


MSc SS Thesis Presentation

Three dimensional Contrast Source Inversion-Electrical Properties Tomography (3D CSI-EPT)

Reijer Leijsen

Additional information ...


PhD Thesis Defence

Relative Space-Time Kinematics Of an Anchorless Network

Raj Thilak Rajan

Space and time awareness has been an integral quest of human evolution, and more so in the currently burgeoning era of wireless sensor networks (WSN), internet of things (IoT) and big data. The rapid advances in technology in recent times has led to affordable, miniaturized and low-power sensor nodes, enabling the feasibility of networks with numerous nodes. These nodes are typically equipped with diverse portfolio of sensors to measure various physical phenomenon, which are cooperatively communicated and processed for appropriate statistical inference. To ensure coherent sampling, efficient communication and prudent inference, the knowledge of position and time of the sampled data is imperative, and consequently accurate space-time estimation of the nodes is as valuable as the sampled data itself.

In this dissertation we address the space-time estimation of a specific class of WSNs, namely an anchorless network of asynchronous mobile nodes. As the terminology suggests, we consider a network of mobile nodes under non-relativistic motion, whose space-time kinematics are to be estimated. In addition, the term anchorless indicates no apriori information on the absolute position or time of any node within the network. This approach is a stark contrast to conventional anchored scenarios, e.g., GPS-based localization, where absolute space-time reference is known. Anchorless networks arise naturally when deployed in inaccessible regions, where an absolute space-time reference is non-existent or only intermittently available. Moreover, when a swarm of nodes is considered, imparting the absolute reference to all the nodes could be limited by communication resources. A few application scenarios include, for example, indoor localization, underwater networks, drone swarms and space-based satellite arrays. In such anchorless networks, it is paramount to understand the relative space-time kinematics, which is the primary theme of this dissertation.

Unfortunately, our understanding of relative kinematics in Euclidean space is inherently dependent on an absolute reference. For instance, consider the first-order relative spatial kinematics, i.e., relative velocity, which is rightly defined as the vector difference between absolute velocities of the respective nodes. However, in the absence of apriori information on any absolute velocities, a natural question arises if these relative velocities can be estimated using only pairwise distance measurements between the nodes. In addition to relative spatial estimation, the asynchronous clocks on-board each of these nodes must also be synchronized, in the absence of a known absolute time-reference. These are some of the fundamental challenges which are addressed in this dissertation.

Additional information ...


Signal Processing Seminar

Three easy pieces in statistical signal processing

Amir Leshem
Bar-Ilan University, Israel

I will describe some new results from estimation theory and networks dynamics. All the results are very simple and deal with very well known concepts in statistical signal processinfg.

The first will be the non existence of unbiased estimators in many constrained estimation problems. The second will describe a novel herding phenomenon in consensus based on actions The last will be novel bounds on the performance of LS estimators with finitely many samples and non Gaussian noise.

I will assume knowledge of basic estimation theory and will describe all the necessary probabilistic tools. Interestingly in all results advanced probability and measure theory play an important role.

Additional information ...


MSc CE Thesis Presentation

Design and Characterization of the Time Division Multiplexing concept on a dual clock Imaging DSP

Ioanna Kyriakou

The market demand for quality imaging on portable devices has increased heavily in the last few years. This increase on the demands has a direct effect on the complexity of the imaging-dedicated devices, making the Digital Signal Processors (DSP) development a key factor in the race of imaging quality. More complex devices have been developed to cope with the technical challenges in terms of computational capability, power consumption and area footprint of the devices, especially in the portable market. This challenge is the key to the success of the imaging processor in the market, forcing companies to develop new and creative ways to reduce area and power while keeping performance.

Experimental results showed that by applying the Time Division Multiplexing (TDM) concept, as a clock domain crossing technique, in an Imaging DSP (IDSP), the area that it occupies as well as the power that it consumes can be reduced. In this thesis, four new implementations of the IDSP are proposed and their functionality was proven to be correct. All these designs were synthesized and results about area and power were extracted and analyzed.

Additional information ...


MRI based Electrical Properties Tomography: Electromagnetic inversion

Patrick Fuchs

In magnetic resonance imaging (MRI) the interest in electric properties tomography (EPT) is growing. In current EPT applications the reconstruction is performed based on the Helmholtz equation which relies on the assumption of a homogeneous contrast. In this talk I will present new approaches to reconstruct the electrical properties that require less assumptions on the contrast. Two fundamentally new approaches are presented, one based on first order differentiation and one on the global integral field equations using a contrast-source variable. Reconstruction of both two- and three-dimensional simulations as well as the reconstruction of an in vivo measurement are performed to compare the five different methods. It can be concluded from this comparison that methods that are not based on the homogeneous contrast assumption perform much more accurate (overall) than the Helmholtz equation based method. These new methods provide new insight into the inversion problem in MRI, specifically for EPT and get us one step closer to accurate electric properties reconstruction from an MRI scan.

Additional information ...


DCSE and SIAM Seminar

Finite Difference Interpretation of Reduced Order Models and Applications to Wave Propagation

Vladimir L. Druskin
Schlumberger-Doll Research, Cambridge (USA)

In model order reduction, one approximates the response (transfer function) of a large scale dynamic system using a smaller system, known as reduced order model (ROM), that retains certain features of the larger problem.

For lossless wve propagation such a feature is energy conservation, that manifests in the Stieltjess-Markov property of (frequency-domain) transfer functions. In his seminal ork (1952) Mark Krein showed that the Stieltjes rational transfer functions can be equivalently presented by (Stieltjes) strings of point masses and weightless springs, i.e., described by a dynamic system with s.p.d. tri-diagonal matrix.

In turn, the Stieltjes strings give rise to interpretation of the ROMs via the second-order finite-difference approximation of the underlying PDE on judiciously chosen grids. A known application of such an approach is construction of optimal discrete perfectly matched layers (PMLs) for exterior wave problems.

In this talk I will present two more recent applications of the Stieltjes string technology. The first one is construction of so-called S-fraction multiscale reduced order model with sparse coarse cell approximation for hyperbolic problems, illustrated ith simulation examples for large scale 3D elastic anisotropic wave problems. The second application is direct nonlinear imaging algorithm via the data-driven discrete-time ROM with 2D examples from seismic exploration and ultrasonic medical imaging.

This is joint work with Alexander Mamonov, Andre Thaler and Mikhail Zaslavsky.

Biography

Vladimir L. Druskin is Scientific Advisor at Schlumberger-Doll Research and a SIAM Fellow. Throughout his career, Vladimir Druskin made fundamental contributions to inverse problems, scientific computing, and numerical analysis and their application to hydrocarbon exploration.


MSc SS Thesis Presentation

Accelerating Diffusion-Weighted Chemical Shift Imaging using Compressed Sensing with Parameter Mapping

Joost van der Kemp

Diffusion-weighted chemical shift imaging (DW-CSI) is a recently developed MRI modality that enables radiologists to reveal the diffusion properties of small molecules that act in metabolic reactions in-vivo. In order to extract this diffusion information from a patient, DW-CSI requires approximately one hour of scan time. This extensive scan time makes DW-CSI currently inapplicable for the clinical setting. This thesis describes the closely intertwined implementation of compressed sensing with parameter mapping (CS-PM) in the DW-CSI processing pipeline to accelerate its acquisition. The CS-PM algorithm enables DW-CSI to acquire less measurements (sample under Nyquist) and subsequently reconstruct the missing samples with the use of a custom designed, model-based sparsifying dictionary. As proof of concept, CS-PM was evaluated on the water signal of a non-water-suppressed scan. The results of the integration of CS-PM in the DW-CSI processing pipeline already indicate a feasible acceleration factor of 1.5 along with valuable insight to further improve the performance of DW-CSI in combination with CS-PM.

Additional information ...


Comparisson of State-of-the-Art Consensus Algorithms

Richard Heusdens

Abstract: Unlike the Telephone network or the Internet, many of the next generation networks are not engineered for the purpose of providing efficient communication between various networked entities. Examples of such networks are sensor networks, peer-to-peer networks, mobile networks of vehicles and social networks. These networks lack infrastructure; they exhibit unpredictable dynamics and they face stringent resource constraints. Therefore, algorithms operating within them need to be extremely simple, distributed, robust against networks dynamics, and efficient in resource utilization. Distributed consensus algorithms are built upon a gossip or rumor style unreliable, asynchronous information exchange protocol. Due to their immense simplicity and wide applicability, this class of algorithms is particularly suitable for the next generation networks. In this talk we will give an overview of state-of-the art consensus algorithms and compare their performance.

Slides

Here are the slides.

Additional information ...


Signal Processing Seminar

An Ideal-Theoretic Criterion for Localization of an Unknown Number of Sources

Matthew Morency

Source localization is among the most fundamental problems in statistical signal processing. Methods which rely on the orthogonality of the signal and noise subspaces, such as Pisarenko's method, MUSIC, and root-MUSIC are some of the most widely used algorithms to solve this problem. As a common feature, these methods require both a-priori knowledge of the number of sources, and an estimate of the noise subspace. In this paper, we propose a new localization criterion based on the algebraic structure of the noise subspace. An algorithm is proposed which adaptively learns the number of sources and estimates their locations.

Additional information ...


Optimizing Speech Intelligibility in Noisy Environments Using a Simple Model of Communication

Richard Hendriks

Modern communication technology allows a user to communicate from almost anywhere to almost anywhere. However, interfering sources in the environment of the talker at the far-end and the listener at the near-end often affect the ability of the parties to communicate. Conventional systems for noise reduction and speech intelligibility enhancement typically treat the processing at the far-end and the near-end as two independent problems. To quantify and optimize the intelligibility of noisy speech, we recently introduced a speech intelligibility model based on mutual information. This model takes the noise inherent in the speech production process into account at a fixed SNR (i.e., scaling independent). Such a constant production SNR has a significant effect on a power constrained communication system. That is, the usefulness of a particular communication channel saturates near the production SNR. The resulting intelligibility predictor resembles heuristically derived classical measures of intelligibility such as the AI and the SII. In this presentation we will review our proposed mutual information-based speech intelligibility model and demonstrate its potential. We will relate it to classical measures of intelligibility and will show its potential for speech intelligibility prediction. Furthermore, we will show that this model can be used to derive a multi-microphone processor that is jointly optimal with respect to interfering sources at the far-end as well as at the near-end.

Additional information ...


MSc SS Thesis Presentation

Informed phase restoration of amplitude spectra for audio signals

Iakov Chalegoua

In transform coding for audio, is it sufficient to store only the magnitude information?

Additional information ...


Special Celebrative Seminar: New model of Sino-Dutch R&D cooperation


Since the establishment of TU Delft's Beijing Research Centre (BRC) in 2012, 10 PhD researchers have been enrolled for this unique program in close cooperation with our Chinese Academic Partners. We are very pleased that the first two BRC PhD candidates will have their PhD thesis defence on September 19 2016, in the Aula of Delft University of Technology.

To celebrate this important milestone, we would like to invite you to join a special seminar after the defences, about the New model of Sino-Dutch R&D cooperation, to share the experiences, look to the future and raise the glass together.

Additional information ...


Multivariate signal processing for "big data" sensing

Gabril Vasile
GIPSA lab, Grenoble

Big datasensing brings a strong impact on many sensor-oriented application fields, suchas disaster control and monitor, healthcare services, and environmentprotection. This presentation aims at providing anoverview on several application-drivensignal processing schemes, which have been carried outin the field of environment protection at theGrenoble-Image-sPeech-Signal-AutomaticsLab.


MSc SS Thesis Presentation

Ultrasound Imaging Using a Single Element Transducer

Andrejs Fedjajes

Most of nowadays ultrasound systems visualize 3D space in 2D images. State of the art matrix transducers are very expensive and have not achieved the same popularity yet. We prove the possibility to image a 3D volume with a single element transducer. The core idea is to deliberately change the transducers spatial field and collect the knowledge about that change into a system matrix A. This is done by means of a static aberration mask and a calibration procedure. Opposite to the conventional beamforming, we formulate the imaging task as a least squares inverse problem. This comes at the price of computational resources needed due to the problem dimensionality. The project potentially can lead to low-cost ultrasound imagers as a part of growing industry of body area sensors.


MSc SS Thesis Presentation

Regularized Least Squares Imaging for High Resolution Ultrasound

Pim van der Meulen

Conventional ultrasound imaging by delay-and-sum beamforming is based on geometrical operations on the ultrasound measurements. We think that it makes more sense to try and estimate the scattering composition that best explains the measurement. We have investigated this approach by formulating a linear mathematical model, which allows for a large and flexible variety of techniques to estimate the image, which have been well established and investigated in mathematics. Such a formulation also makes it easier to incorporate prior information about an experiment, such as the sparsity of an image, or its statistical properties. In our work, we have focused on using these techniques to attain high-resolution images, obtained by finding the image that minimizes the squared error between the formulated model and the measurement (i.e. tries to 'explain' the measurement), which vastly improves the common delay-and-sum technique. In the coming presentation some of these exciting results will be shown.


MSc SS Thesis Presentation

Identification of room boundaries for sound field estimation

Mario Coutino Minguez

Echoes generated by the sound reflected off the walls of a room carry information about the geometry of the enclosure. Capitalization of this acoustic property could lead to improvements in current state-of- the-art methods for sound field estimation, where prior information can be used to improve the conditioning of the problem.

In this thesis, robust and computational efficient methods are developed for identifying first order reflections to estimate the room geometry using small microphone arrays.

Furthermore, as the estimation of such reflections becomes even more challenging in actual audio reproduction systems, this work aims to develop methods capable to deal with complications that might arise due to the employed drivers. This is done by considering the estimation problem in two different scenarios. Firstly, the first order reflections estimation problem is posed as a sorting problem. For this case, a set of echoes, received at different microphones, must be grouped accordingly to the wall which originated them. This problem is solved by using a greedy subspace-based algorithm. The proposed approach provides similar performance compared with the state-of-the-art method at a reduced computational cost. For the second scenario, instead of echoes, only raw microphones measurements are available. This instance of the problem is posed under an estimation theory framework, and solved by sequential minimization of a non-linear cost function based on the propagation of waves.

Experimental results, evaluated in simulated shoe-box shaped rooms, demonstrate the performance and applicability of the proposed methods for room geometry estimation.


MSc SS Thesis Presentation

Dielectric Shimming

Michiel Gerlach

Dielectric shimming is proven to be very useful in increasing the homogeneity of the B1+ field in high field MRI. Current optimization and design techniques for dielectric pad parameters are slow. The goal of this thesis is to find a fast and accurate pad design and optimization technique. Two new techniques are proposed. The first, a method that simply uses inspection by solving the forward problem in a relatively fast way. The other proposed technique follows a more analytical approach to find the optimal permittivity and conductivity of a pad in a couple of iterative steps with a Gauss-Newton method. This last technique uses a new proposed approach to predict the phase of the B1+ field in a direct fashion.

These techniques provide fast and accurate simulation results for a two-dimensional abdominal body slice placed in a 3T MRI scanner for different pad scenarios. From these results it can be concluded that both proposed techniques generate comparable pads, which are able to increase the homogeneity of the B1+ field.

A comparison between the two techniques is made. The Gauss-Newton method provides a fast, robust and accurate optimization technique for large scale problems, but offers less flexibility and insight to the data compared to the method via inspection.

The flexibility of the method via inspection and the insight it provides is shown for different scenarios (pad location, multiple pads, pad shape, pad thickness), where the effect of the optimal permittivity and conductivity on the homogeneity of the resulting B1+ field is simulated. Even the maximum allowed SAR can be incorporated in this pad optimization technique.


Multiple user Access Technology in Optical OFDM-PON Upstream Transmission

Han Dun

As bandwidth-hunger multimedia applications continue to fuel the rise in bandwidth demand in the multiple users access system, thus a PONs (passive optical networks)-based solution, which could provide high transmission capacity flexibly and cost-effectively, will be the core of the next-generation optical access network. Recently, Orthogonal Frequency Division Multiple (OFDM)-based PON has been highlighted as an attractive method to realize the next generation optical network due to its desirable feature such as higher spectral efficiency and resilience to linear dispersion.

This presentation will mainly focus on key technology on multi-user access OFDM (A)-PON specifically for the upstream transmission. Each ONU (optical network unit) shares a same ranging subcarrier, and the Gold sequences would be modulated in this subcarrier as synchronization codes (User ID). With Gold sequences modulated in a single carrier we could estimate the offsets both in time and frequency domain such as the transmission delay, different sampling clocks and optical wavelength drift. Each ONU would correct the time and frequency offsets with the estimated information, and keep the orthogonality between each subcarrier.


Energy harvesting wireless networks: A new frontier for communication and information theory

Aylin Yener
Pennsylvania State University

Wireless communication networks composed of devices that can harvest energy from nature will lead to the green future of wireless, as energy harvesting offers the possibility of perpetual network operation without adverse effects on the environment. By developing effective and robust communication techniques to be used under energy harvesting conditions, some of the communication devices in a heterogeneous network can even be taken off the grid. Energy harvesting brings new considerations to system level design of wireless communication networks, leading to new insights. These include randomness and intermittency of available energy, as well as additional system issues to be concerned about such as energy storage capacity and processing complexity. The goal of this talk is to furnish the audience with fundamental design principles of energy harvesting wireless communication networks which is an emerging area. The focus will be on identifying optimum transmission scheduling policies in various settings, and the ensuing algorithmic solutions. Time permitting we will also go into the information theory of energy harvesting communications, which brings in new challenges taking into account energy availability and storage at the channel use level.

Biography

Aylin Yener is a professor of Electrical Engineering at The Pennsylvania State University, University Park, PA since 2010, where she joined the faculty as an assistant professor in 2002. During the academic year 2008-2009, she was a Visiting Associate Professor with the Department of Electrical Engineering, Stanford University, CA. Her research interests are in information theory, communication theory and network science with recent emphasis on green communications, information security and networked systems.

She received the NSF CAREER award in 2003, the best paper award in Communication Theory in the IEEE International Conference on Communications in 2010, the Penn State Engineering Alumni Society (PSEAS) Outstanding Research Award in 2010, the IEEE Marconi Prize paper award in 2014, the PSEAS Premier Research Award in 2014, and the Leonard A. Doggett Award for Outstanding Writing in Electrical Engineering at Penn State in 2014. She is a fellow of the IEEE. Dr. Yener is an elected member of the board of governors of the IEEE Information Theory Society for the term 2015-2017.

Additional information ...


MSc TC Thesis Presentation

Chinese Digital Radio Receiver Design and Development with FM Interference Cancellation

Yiling Zhang

A Chinese Digital Radio (CDR) has been made and become effective in 2013 by SGAPPRFT (State General Administration of Press, Publication, Radio, Film and Television), which is an In-Band On-Channel (IBOC) digital audio broadcasting hybrided with analog FM signal in one FM channel. This project investigates the interferences to the CDR digital signals caused by both co-channel and adjacent-channel FM signals, and proposes solutions to combat such influences. A new acquisition algorithm is implemented to realize mode detection and synchronization for the hybrid signal. Two FM interference removal methods have been proposed and studied in simulation. An energy detector is introduced to select the best FM interference removal method in run time. The proposed solution has reached very good reception performance against co-channel FM interference in commonly used CDR modes. For the adjacent-channel FM interference, if we can detect and select the better removal method in real time, the performance degradation can be controlled to[LINK] a limited SNR increase under the AWGN channel.


Signal Processing Seminar

Jiani Liu

Additional information ...


Physical Foundations Underlying Green Information and Communication Technologies

Earl McCune

There are physical limitations on how much energy efficiency can be realized from any actual hardware used to implement any communications standard. Experience shows that in most instances the signals adopted by the standard committee place an additional ceiling on the achievable energy efficiency using that hardware. For example, there is hardware that is capable of providing more than 60% energy efficiency under ideal conditions, but for some standardized signals the maximum achievable efficiency drops to 7%. This drop in achievable efficiency is predictable, and such analyses should become part of standards committee deliberations. Such a low operating efficiency is not compatible with IoT, 5G, and other upcoming Standards objectives.

This presentation was originally given to the IEEE Green-ICT Initiative Steering Committee at the IEEE Board meeting series in New Jersey on June 16, 2016. It establishes the reasons why such efficiency ceilings occur and shows how to predict them. Further, recipes are provided on how it is physically possible to simultaneously achieve high bandwidth efficiency and optimum energy efficiency along with the PSD impacts that come with these more Green-optimized signal modulations.

Additional information ...


MSc SS Thesis Presentation

Electrical properties tomography for MRI

Patrick Fuchs

In Magnetic resonance imaging (MRI) an interest in electrical properties tomography (EPT) is growing. In current EPT the reconstruction is performed based on the Helmholtz equation which relies on the assumption of a homogeneous contrast. The goal of this thesis is to present new approaches to reconstruct the electrical properties that require less assumptions on the contrast. Two new approaches are presented, one based on first order differentiation and one on the global integral field equations using a contrast-source variable. In this thesis these methods are described alongside the existing Helmholz based approach, the contrast source inversion - EPT approach and the deconvolution approach. These last three approaches have already been published on, but are reviewed here for completeness.

Reconstruction of both two dimensional and three dimensional simulations as well as the reconstruction of an in-situ measurement are performed to compare the five different methods. It can be concluded from this comparison that all methods that are not based on the homogeneous contrast assumption perform much more accurate (overall) than the Helmholtz equation based method. Both contrast source inversion and the direct inversion method based on the global integral equations perform comparable, but the latter is a lot faster and offers almost the same range of flexibility regarding regularisation and preconditioning. The direct inversion method is a straight improvement on the deconvolution method, performing equally well regarding noise robustness, but offering better reconstructions in almost all cases due to the lack of apodisation step. The first order differential method provides a surprisingly robust, accurate and extremely fast way to get insight into the data, and shows that the inversion problem in MRI is actually very well behaved as far as inversion problems go. These new methods provide fresh insight into the inversion problem in MRI, specifically for EPT and get us one step closer to accurate electric properties reconstruction from an MRI scan.

Additional information ...


Signal Processing Seminar

Jie Zhang

Additional information ...


MSc SS Thesis Presentation

Acoustic Vector Sensor Based Source Localization

Krishnaprasad

MSc work done at Microflown

Additional information ...


Signal Processing Seminar

Yongchang Hu

Additional information ...


PhD Thesis Defence

Underwater Acoustic Localization and Packet Scheduling

Hamid Ramezani

Additional information ...


Computational neuroscience seminar

Short presentations with update of MSc research projects status.

Additional information ...


Signal Processing Seminar

Fundamentals of Graph Signal Processing and Applications to Diffusion Processes

Prof. Antonio G. Marques
King Juan Carlos University, Madrid

Coping with the challenges found at the intersection of Network Science and Big Data necessitates broadening the scope beyond classical temporal signal analysis and processing in order to also accommodate signals defined on graphs. Under the assumption that the signal properties are related to the topology of the graph where they are supported, the goal of graph signal processing (GSP) is to develop algorithms that fruitfully leverage this relational structure. Instrumental to that end is the so-termed graph-shift operator, a matrix capturing the graphs local topology and whose spectral decomposition is central to defining graph Fourier transforms.

In the last years a body of works has successfully implemented this approach, showing how several classical signal processing results can be gracefully generalized to the more irregular graph domain.

The talk consists of two parts. The first part introduces the field of GSP, motivates its usefulness via meaningful applications, and presents its foundational concepts, which have been derived over the past five years. The second part focuses on contemporary results, including optimal filter design, blind identification and network topology inference. For each of these results, the theoretical contribution will be first described and then the implications for distributed and dynamic processing will be discussed.

Biography

Antonio G. Marques received the Telecommunications Engineering degree and the Doctorate degree, both with highest honors, from the Carlos III University of Madrid, Spain, in 2003 and 2007, respectively. In 2007, he became a faculty of the Department of Signal Theory and Communications, King Juan Carlos University, Madrid, Spain, where he currently develops his research and teaching activities as an Associate Professor. From 2005 to 2015, he held different visiting positions at the University of Minnesota, Minneapolis. In 2015 and 2016 he was a Visiting Scholar at the University of Pennsylvania. His research interests lie in the areas of communication theory, signal processing, and networking. His current research focuses on stochastic resource allocation wireless networks and smart grids, nonlinear network optimization, and signal processing for graphs. Dr. Marques has served the IEEE and the EURASIP in a number of posts (currently, he is an Associate Editor of the IEEE Signal Process. Letters and of the EURASIP J. on Advances in Signal Process.), and his work has been awarded in several conferences and workshops.


Signal Processing Seminar

Andreas Koutrouvelis

Additional information ...


MSc Thesis Presentation

Quantization effects in the primal-dual method of multipliers (PDMM)

Daan Schellekens

Nowadays, large-scale networks of computing units, often characterized by the absence of central control, have become more commonplace in many applications. To facilitate data processing in these large-scale networks distributed signal processing is required. The iterative behaviour of distributed processing algorithms combined with the limited energy, computational power, and bandwidth, limitations imposed by such networks, place tight constraints on the transmission capacities of the individual nodes. For this reason quantization in distributed algorithms has become an interesting and popular topic.

Already considerable research has been performed into quantization effects for various distributed algorithms, such as alternating-direction method of multipliers (ADMM). However, for the primal-dual method of multipliers (PDMM), a recently proposed promising distributed algorithm, no research into the effects of quantization exists.

In this thesis the effects of subtractively dithered uniform quantization in PDMM are investigated for the synchronous distributed averaging problem. This specific averaging problem, which is often considered as the canonical distributed problem, was selected to start the research from a natural point. As such, the theory developed in this thesis can form a foundation for further research into quantization effects in PDMM in general.

The quantization effects are discussed by considering the convergence rate of the algorithm. This is done by deriving expressions for the mean squared error (MSE) that include quantization noise. Also the required bitrate for quantized PDMM is considered. It was concluded that for practical applications quantization in PDMM can be applied with a near-fixed-rate quantizer, such that significant bitrate reduction can be achieved, without compromising the rate of convergence.

Additional information ...


PhD Thesis Defense

Covariance matching techniques for radio astronomy calibration and imaging

Millad Sardarabadi

The search for the answer to one of the most fundamental scientific questions, How was the universe formed?, requires us to study very weak radio signals from the early universe. In the last eighty years, radio astronomers have been able to use radio frequency observations for significant discoveries such as quasars, supermassive Black Holes and the Cosmic Microwave Background radiation. Radio astronomers use a radio telescope to study the cosmos. A radio telescope usually consists of an array of radio receivers (antennas) and supporting hardware/software to produce synthesized images of the sky. While the earlier generation of the radio telescopes such as the Westerbork Synthesis Radio Telescope (WSRT), the Very Large Array (VLA) and the Giant Meterwave Radio Telescope (GMRT) consisted of 14-45 receivers separated a few kilometers (3-25 km basedlines), the next generation of radio telescopes such as LOFAR and SKA have thousands of receivers which cover distances of over 1000 km. This massive increase in the number of receivers and the geometric dimensions is a consequence of the required (high) resolution and sensitivity for modern scientific studies and while it is necessary, it does not guarantee the desired results without the appropriate data and signal processing.

The main challenges in radio astronomy can be divided in three closely related problems: mitigation of manmade radio frequency interference, calibration and image formation. The main goal of this thesis is to investigate howthe signal processing formalism can be used to systematically model and analyze these three problems and what signal processing tools are needed for addressing them.

The number of RFI free bands is diminishing rapidly as a consequence of the increased number of wireless services and applications. The shift towardswideband digital systems has created new problems which are not sufficiently addressed by currently implemented RFI detection and mitigation systems. For this class of continuously present wideband RFI, the use of array processing techniques such as spatial filtering could provide access to frequency bands otherwise unusable by astronomers. Such a spatial filtering can be achieved by estimating and removing the subspace that the interfering signal is occupying. Many signal processing algorithms use the eigenvalue decomposition (EVD) for estimating the signal subspace. However the use of EVD is limited to systems where the noise is white or known from calibration. This requirement is a limiting factor for applying these techniques to uncalibrated arrays with unknown noise models. In these situations a more generic approaches which allows for combined RFI filtering and noise power calibration is preferred. In this thesis factor analysis (FA) is proposed as suitable substitution for EVD.

FA is a technique that allows for the decomposition of the signal into a lowrank part corresponding to the signal and a diagonal part which represents the covariance of the noise on the receivers. Because the diagonal elements can be different this technique can be used when the noise is not white and forms a generalization of the EVD. In RFI mitigation applications the signal part of the data is dominated by RFI and changes more rapidly than the noise. Estimating the noise covariance which is shared by several measurements jointly allows for a more accurate estimation. As a result extensions to the classical FA are proposed to improve the estimates for the diagonal part of the decomposition in a joint fashion. Even a diagonal noise structure can be limiting in some applications. For example the contribution of theMilkyWay affects the short baselines which can be modeled by using a nondiagonal covariance matrix. The FA model can be extended for this type of signals. An extension to FA called Extended FA (EFA) is used to allow for capturing such structures into the model. Similar to JFA we can also estimate the parameters in EFA jointly, and the resulting method is denoted by Joint EFA (JEFA). Using nonlinear optimization techniques combined with Krylov subspace based solvers an scalable algorithm is developed. The statistical efficiency of this algorithm is shown by comparing its results to the CramrRao bound and its application in RFI mitigation has been demonstrated on measurements from the WSRT and LOFAR.

Antenna gain calibration is an essential step in producing accurate images. Using common array processing data models, the gain calibration is formulated as a nonlinear covariance matching problem. In this thesiswe showthat the matrices involve in this estimation problem are highly structured and that the systemof equations involving these matrices can be efficiently solved using Krylov subspace based solvers (similar to JEFA). The resulting calibration algorithm is scalable and requires a low number of iterations in order to converge which makes it an attractive alternative to currently available techniques.

Both classical and parametric based image formations consist of two steps. First a dirty image is constructed from the measurements and then an improved estimate is found by performing a deconvolution step. When the number of pixels on the image becomes large, the deconvolution step becomes an illposed problem. In this thesis we showthat image values are bounded frombelowby a nonnegativity constraint and above by the dirty image. Using beamforming techniques, we show that tighter upper bounds can be constructed using the MVDR beamformer. These bounds allow us to regularize the deconvolution problem by a set of inequality constraints. Following a signal processing model, the image formation is then formulated as a parameter estimation problem with inequality constraints. This optimization problem can be solved using an active set algorithm. We show that, with the right initialization, the active set steps are very similar to sequential source removing techniques such as CLEAN. This connection between classical approaches and parametric imaging techniques provides the necessary theoretical basis for further analysis and allows for improving both methods.

Based on the results presented in this thesis we can conclude that signal processing methodologies can provide new solutions to the radio astronomical problems and also shed light on the inner working of the classical techniques. Hence, a signal processing approach is extremely beneficial in tackling the problems that the next generation of radio telescopes will face.

Additional information ...


SP Seminar

Signal Processing for Radio Astronomy

Yves Wiaux, Stefan Wijnholds, Amir Leshem

Yves Wiaux: "Astronomical imaging, in every sense of the word: scalable optimisation algorithms in radio-interferometry"

Amir Leshem: "Detection of transient sources"

Stefan Wijnholds: "Blind calibration of aperture arrays"


Signal Processing Seminar

Yan Xie

Additional information ...


Computational neuroscience seminar

Short presentations with update of MSc research projects status.

Additional information ...


Signal Processing Seminar

Seyran Khademi

Additional information ...


Computational neuroscience seminar

Short presentations with update of MSc research projects status.

Additional information ...


Signal Processing Seminar

Christos Tzagkarakis

Additional information ...


Signal Processing Seminar

Shahrzad Naghibzadeh

Additional information ...


Signal Processing Seminar

Daan Schellekens

Additional information ...


PhD Thesis Defence

Cognitive Radio-based Home Area Networks

Mohd Adib Sarijari

A future home area network (HAN) is envisaged to consist of a large number of devices that support various applications such as smart grid, security and safety systems, voice call, and video streaming. Most of these home devices are communicating based on various wireless networking technologies such as WiFi, ZigBee and Bluetooth which typically operate in the already congested ISM licensed-free frequency band. As these devices are located in a small physical space (i.e., limited by the size of the house), they might interfere with one another which causes a severe limitation to the quality-of-service (QoS) such as throughput. These issues are further aggravated in dense cities where the HAN also receives interference from neighboring HANs. Cognitive radio (CR) is seen as one of the most promising technologies to solve these problems and at the same time fulfill the HANs communication needs. CR technology enables the HAN devices to intelligently exploit idle spectrum including licensed spectrum for their communications, avoiding from being interfered as well as causing interference to others (in particular the incumbent user). We study these problems and the appropriateness of CR as a candidate solution.

We start by designing a new communication system for HAN based on CR technology and clustered network topology, called TD-CRHAN. TD-CRHAN aims at sustainably and efficiently supports the ever-rising throughput demand as well as solving the interference issue in HAN. In the TD-CRHAN, the achievable throughput is optimized to be just equal or slightly higher than the total networks throughput demand, instead of being maximized. We then mathematically model the proposed TD-CRHAN where in the model, general expressions of the cooperative spectrum sensing performance parameters are considered. This allows us to analyze the performance of TD-CRHANfor amore realistic scenario where the incumbent user signal-to-noise-ratio (SNR) is not the same at different sensing devices. We illustrate promising results, numerically and through simulation, confirming the performances of the proposed design.

As a cognitive radio based network also imposes additional overhead in energy consumption due to the spectrum sensing, we then propose an energy efficient cooperative spectrum sensing (CSS) scheme. The scheme is designed based on the proposed TD-CRHAN. In this scheme we also ensure that the throughput demand is kept satisfied efficiently. From the difference in sensing devices incumbent user SNR (that is previously considered), we select the optimal sensing devices for CSS with the corresponding sensing time and detection probability which can be varied from one sensing device to another. We then evaluate the proposed CSS scheme and exhibit the gains obtained in energy- and throughput-efficiency.

Finally, we present a sensing device grouping and scheduling scheme for multichannel CSS. In addition to the energy- and throughput-efficiency, this scheme addresses the fairness in spectrum sensing load distribution among the available sensing devices in a HAN. In this work, we consider the fairness objective as to maximize the lifetime of each sensing device to its expected lifetime. In the proposed scheme, we determine the optimal number of channels that should be used for the network and the selected channels. We also determine the optimal number of devices in each sensing group and which devices. Subsequently, we optimally schedule the formed sensing groups to sense the selected channels. The results from the performance analysis verify our claims.

Additional information ...


Signal Processing Seminar

Jörn Zimmerling

Additional information ...


Signal Processing Seminar


PhD Thesis Defence

Array of single-photon detectors applied to PET imaging

Chockalingam Veerappan

Additional information ...


High-accuracy Positioning in Multipath Channels: Location-Awareness for 5G Networks

Klaus Witrisal
Graz University of Technology

Location-awareness is the capability of a mobile network to employ position information for the sake of operating more efficiently. It is foreseen that fifth-generation (5G) wireless networks will be able to exploit much more accurate position information than any previous generation of wireless networks. This comes naturally due to the increased bandwidth and the application of multi-antenna techniques, which will eventually turn multipath propagation from an enemy to a friend.

This talk will first highlight the impact of multipath propagation on the accuracy of wireless range and position estimation. The important influence of bandwidth and the benefit of MIMO processing will be analyzed. Next it will be shown how one can make use of multipath to benefit from improved positioning accuracy, robustness, and a relaxed need for infrastructure nodes. Analytical performance bounds, their experimental validation, and algorithms derived thereof will be discussed. Finally, it will be shown that a multipath-enabled positioning system is a showcase example of a cognitive dynamic system that can optimize the information gained from each measurement, exploiting its memory of past measurements to plan future measurements. The environment map it collects can be used to predict the propagation characteristics, yielding location-awareness for positioning and communications. It is concluded that the use of position information may become as important to 5G networks as other new, disruptive technologies such as massive MIMO and mm-wave.

Biography

Klaus Witrisal received the Dipl.-Ing. degree in electrical engineering from Graz University of Technology, Graz, Austria, in 1997, the Ph.D. degree (cum laude) from Delft University of Technology, Delft, The Netherlands, in 2002, and the Habilitation from Graz University of Technology in 2009. He is currently an Associate Professor at the Signal Processing and Speech Communication Laboratory (SPSC) of Graz University of Technology, Graz, Austria, where he has been participating in various national and European research projects focused on UWB communications and positioning. He is co-chair of the Technical Working Group "Indoor'' of the COST Action IC1004 "Cooperative Smart Radio Communications for Green Smart Environments.'' His research interests are in signal processing for wideband and UWB wireless communications, propagation channel modeling, and positioning. Prof. Witrisal served as a leading chair for the IEEE Workshop on Advances in Network Localization and Navigation (ANLN) at the IEEE Intern. Conf. on Communications (ICC) 2013 - 2016, as a TPC (co)-chair of the Workshop on Positioning, Navigation and Communication (WPNC) 2011, 2014, and 2015, and as a co-organizer of the Workshop on Localization in UHF RFID at the IEEE 5th Annual Intern. Conf. on RFID, 2011. He is an associate editor of IEEE Communications Letters since 2013.


Signal Processing Seminar

Thomas Sherson

Additional information ...


Signal Processing Seminar

Elvin Isufi

Additional information ...


Signal Processing Seminar

Negin Bakhshi Zanjani


Signal Processing Seminar

Rob Remis

Additional information ...


Signal Processing Seminar

Jeroen van Gemert

Additional information ...


Signal Processing Seminar

An Algebraic Approach to Rank-Constrained Semidefinite Programs

Matthew Morency


PhD Thesis Defence

Sparse Sensing for Statistical Inference: Theory, Algorithms, and Applications

Sundeep Prabhakar Chepuri

In today's society we are flooded with massive volumes of data of the order of a billion gigabytes on a daily basis from pervasive sensors. It is becoming increasingly challenging to locally store and transport the acquired data to a central location for signal/data processing (i.e., for inference). Consequently, most of the data is discarded blindly, causing serious performance loss. It is evident that there is an urgent need for developing unconventional sensing mechanisms to extract as much information as possible yet collecting fewer data. Thus, reducing the costs of sensing as well as the related memory and bandwidth requirements.

The first aim of this thesis is to develop theory and algorithms for data reduction. We develop a data reduction tool called sparse sensing, which consists of a deterministic and structured sensing function (guided by a sparse vector) that is optimally designed to achieve a desired inference performance with the reduced number of data samples. The first part of this thesis is dedicated to the development of sparse sensing models and convex programs to efficiently design sparse sensing functions.

Sparse sensing offers a number of advantages over compressed sensing (a state-of-the-art data reduction method for sparse signal recovery). One of the major differences is that in sparse sensing the underlying signals need not be sparse. This allows us to consider general signal processing tasks (not just sparse signal recovery) under the proposed sparse sensing framework. Specifically, we focus on fundamental statistical inference tasks, like estimation, filtering, and detection. In essence, we present topics that transform classical (e.g., random or uniform) sensing methods to low-cost data acquisition mechanisms tailored for specific inference tasks. The developed framework can be applied to sensor selection, sensor placement, or sensor scheduling, for example.

In the second part of this thesis, we focus on some applications related to distributed sampling using sensor networks. Recent advances in wireless sensor technology have enabled the usage of sensors to connect almost everything as a network. Sensor networks can be used as a spatial sampling device, that is, to faithfully represent distributed signals (e.g., a spatially varying phenomenon such as a temperature field). On top of that, the distributed signals can exist in space and time, where the temporal sampling is achieved using the sensor's analog-to-digital converters, for example. Each sensor has an independent sample clock, and its stability essentially determines the alignment of the temporal sampling grid across the sensors. Due to imperfection in the oscillators, the sample clocks drift from each other, resulting in the misalignment of the temporal sampling grids. To overcome this issue, we devise a mechanism to distribute the sample clock wirelessly. Specifically, we perform wireless clock synchronization based on the time-of-flight measurements of broadcast messages. In addition, clock synchronization also plays a central role in other time-based sensor network applications such as localization. Localization is increasingly gaining popularity in many applications, especially for monitoring environments beyond human reach, e.g., using robots or drones with several sensor units mounted on it. Consequently we now have to localize more than one sensor or even localize the whole sensing platform. Therefore, we extend the classical localization paradigm to localize a (rigid) sensing platform by exploiting the knowledge of the sensor placement on the platform. In particular, we develop algorithms for rigid body localization, i.e., for estimating the position and orientation of the rigid platform using distance measurements.

Additional information ...


Signal Processing Seminar

Censor, Sketch, and Validate for Learning from Large-Scale Data

Georgios Giannakis
Univ. of Minnesota

We live in an era of data deluge. Pervasive sensors collect massive amounts of information on every bit of our lives, churning out enormous streams of raw data in various formats. Mining information from unprecedented volumes of data promises to limit the spread of epidemics and diseases, identify trends in financial markets, learn the dynamics of emergent social-computational systems, and also protect critical infrastructure including the smart grid and the Internets backbone network.

While Big Data can be definitely perceived as a big blessing, big challenges also arise with large-scale datasets. This talk will put forth novel algorithms and present analysis of their performance in extracting computationally affordable yet informative subsets of massive datasets. Extraction will effected through innovative tools, namely adaptive censoring, random subset sampling (a.k.a. sketching), and validation. The impact of these tools will be demonstrated in machine learning tasks as fundamental as (non)linear regression, classification, and clustering of high-dimensional, large-scale, and dynamic datasets.

Biography

Georgios B. Giannakis received his Diploma in Electrical Engr. from the Ntl. Tech. Univ. of Athens, Greece, 1981. From 1982 to 1986 he was with the Univ. of Southern California (USC), where he received his MSc. in Electrical Engineering, 1983, MSc. in Mathematics, 1986, and Ph.D. in Electrical Engr., 1986. Since 1999 he has been a professor with the Univ. of Minnesota, where he now holds an ADC Chair in Wireless Telecommunications in the ECE Department, and serves as director of the Digital Technology Center.

His general interests span the areas of communications, networking and statistical signal processing subjects on which he has published more than 380 journal papers, 650 conference papers, 20 book chapters, two edited books and two research monographs (h-index 115). Current research focuses on big data analytics, wireless cognitive radios, network science with applications to social, brain, and power networks with renewables.

He is the (co-) inventor of 22 patents issued, and the (co-) recipient of 8 best paper awards from the IEEE Signal Processing (SP) and Communications Societies, including the G. Marconi Prize Paper Award in Wireless Communications. He also received Technical Achievement Awards from the SP Society (2000), from EURASIP (2005), a Young Faculty Teaching Award, the G. W. Taylor Award for Distinguished Research from the U. of Minnesota, and the IEEE Fourier Technical Field Award (2015). He is a Fellow of IEEE and EURASIP, and has served the IEEE in a number of posts including that of a Distinguished Lecturer for the IEEE-SP Society.

Additional information ...


Signal Processing Seminar

Nambur Ramamohan Krishnaprasad

Additional information ...


Signal Processing Seminar

Jing Han

Additional information ...


MSc Thesis Presentation

Multi-Domain SystemC Model of a Neural Interface

Kiki Wirianto

Neural networks have been investigated by researchers for several decades. Microelectrodes and neural interface are used to obtain the informations contained in the neuronal networks activity, which can be used to control neural prosthetic devices. This field has developed rapidly and the current research is focusing on multi-channel implementation of neural interface system to monitor the activity of a large number of neurons simultaneously.

Area and safety are two main constraints in the design of neural interface electronics. The chip area constrain is important to minimize the severity of the surgery and limit the displacement of the brain caused by the implanted device. The safety constrain is critical in avoiding the damage to the brain tissue. Both constrains create a limitation on the power consumption of the neural interface system. This limited power budget needs to be utilized carefully to implement a design with low noise and high data rate with as few computational resources as possible. An efficient design allows a large number of channels to be implemented within the allowed power budget.

This thesis proposes behavioral models of the electrode and the neural interface front-end, a part which precondition the neural signals before they are further processed and stored. The functionality of the proposed models are verified and, together with a power estimation model, they are used to perform a system study to investigate the trade-offs between neural interface design parameters.

Additional information ...


Signal Processing Seminar

Venkat Roy

Additional information ...


PhD Thesis Defence

Blind Beamforming Techniques for Global Tracking Systems

Mu Zhou

In lower frequency bands, existing communication systems face unprecedented demands to accommodate more users in new applications. These growing demands exceed the designed system capacity and thus call for innovative solutions while keeping compatibility to the current setup to reduce the cost of users. For example, in the automatic identification system (AIS), satellite receivers are being used for expanding the service coverage of ship tracking to the global range, and similarly in the automatic dependent surveillance-broadcast (ADS-B) system for aircraft tracking. These systems are narrowband and originally designed in the last century, but they will continue to run for at least another couples of years without major updating of the user-side equipment.

The new application of AIS considered in this thesis is Satellite AIS. The satellite runs in the low-earth orbit (LEO). On the satellite, receiving AIS signals becomes much more difficult than before: one has to combat in-cell and inter-cell interfering sources from the system itself. Interference suppression is the main topic of this thesis.

Narrowband spatial beamforming techniques for antenna arrays are candidate solutions to this challenge. This thesis tries to develop new beamforming techniques with a simple structure and a low computational complexity. With these techniques, this thesis establishes a framework of multiuser reception for Satellite AIS.

As a basic tool for the proposed algorithms in this thesis, a signed URV algorithm (SURV) is proposed for the basic problem of principal subspace computation and tracking as a replacement of the singular value decomposition (SVD). The updating and downdating of SURV is direct and simple. SURV has no issue of numerical stability unlike previous algorithms in linear algebra and shows consistent performance in both stationary and nonstationary cases.

New blind beamforming techniques are proposed for separating overlapping packets in nonstationary scenarios. The connections between subspace intersection, oblique projection, the generalized SVD (GSVD), the generalized eigenvalue decomposition (GEVD), and SURV are exposed. Simulation and experimental results of the proposed algorithms are shown.

In the remaining part of the thesis, the work developing the software simulation model and constructing the hardware platform is presented. The outputs of the work are used for the verification and validation of the proposed algorithms in this thesis.

Additional information ...


Signal Processing Seminar

Efficient two-way relaying schemes for amplify and forward relays with multiple antennas

Martin Haardt
TU Ilmenau, Germany

Biography

Martin Haardt is a Full Professor in the Department of Electrical Engineering and Information Technology and Head of the Communications Research Laboratory at Ilmenau University of Technology, Germany, since 2001.

He received his Ph.D. degree from Munich University of Technology in 1996 after which he joined Siemens Mobile Networks.

Martin Haardt has received the 2009 Best Paper Award from the IEEE Signal Processing Society, the Vodafone (formerly Mannesmann Mobilfunk) Innovations-Award for outstanding research in mobile communications, the ITG best paper award from the Association of Electrical Engineering, Electronics, and Information Technology (VDE), and the Rohde & Schwarz Outstanding Dissertation Award.

His research interests include wireless communications, array signal processing, high-resolution parameter estimation, as well as numerical linear and multi-linear algebra.

Additional information ...


Signal Processing Seminar

From Tensor Decomposition to Coupled Matrix/Tensor Decompositions in Array Processing

Lieven Delathauwer
KU Leuven

Additional information ...


Signal Processing Seminar

Something on GSVD

Soren Holdt Jensen
Aalborg University (Denmark)

GSVD, oblique projections, and source separation

Additional information ...


MSc ME Thesis Presentation

Sensor Selection and Bit Allocation in WSNs with Realistic Digital Communication Channels

Hongrun Zhang

For energy management in wireless sensor networks, only the sensors with most informative measurements are activated to operate. How to select sensors that make good tradeoff between performance and energy consumption is what many researchers are focusing on. Existing solutions assume analog data model, i.e., the data from sensors collected by a center node, called fusion center, are analog measurements. In practical application, due to limitations of energy of sensors and bandwidth of wireless channel, original measurements are usually compressed before being transmitted to the fusion center. In addition, transmitted signals are usually distorted by wireless channel effects, therefore it is possible that the received data are corrupted with errors. In this thesis, we consider two compressive techniques: one-bit quantization and multi-bit quantization. In one-bit quantization, an indicator message is generated in a sensor according to whether the original measurement is larger than a threshold or not. In multi-bit quantization, the original measurements are quantized to multiple bits and only the most significant bits are reserved. The indicators or the most significant bits are then transmitted through realistic wireless channel to the fusion center for it to process. By these ways, the transmitted signals are digital, and they may flip into opposite values by the effects of wireless channels. For one-bit quantization case, we develop a sensor selection approach, based on convex programming. For multi-bit quantization, we extend the sensor selection to bit allocation and propose a novel algorithm to determine the number of bits to transmit for each sensors, which is also based on convex programming. In both cases we consider the effects of wireless channels, which are characterized as bit error rate. Particularly, for the multi-bit quantization, numerical results show that the bit allocation can further reduce the cost that we defined compared with existing solutions where transmitted data are assumed to be analog.


Signal Processing Seminar

Sundeep Prabhakar Chepuri

Data reduction tools like censoring or sketching for (large-scale) linear inverse problems in presence of outliers and/or bounded data uncertainties

Additional information ...


Signal Processing Seminar

Jac Romme

Additional information ...


MSc SS Thesis Presentation

Speech Based Onset Estimation for Multisensor Localization

Rodolfo Solera Urcuyo

This work presents a study of a current problem in the field of audio processing: Source and receiver localization. Currently, this problem requires that either the onset time of the sources or the internal delay of the receivers are known. The algorithms studied here, take advantage of the structure of the time matrix, which contains the TOA of all the receivers with respect to all the sources, and finds the solution to the locations when the onset times are known. The problem here is then approached from a time difference of arrival (TDOA) perspective, which inherently cancels the onset times by subtracting the time of arrival (TOA) of a source at every receiver.

An alternative approach is also proposed, which uses speech signals as calibration signals in order to estimate the onset times. Such an approach is based on an algorithm which uses artificial calibration signals to calculate the onsets. Those signals are known a-priori, which implies that an additional device which produces those signals is needed. Once both internal delays and onset times are known, the locations of both sources and receivers can be estimated using a current algorithm which is also described here


MSc ME Thesis Presentation

A 1 GSa/s Deep Cryogenic, Reconfigurable Soft-core FPGA ADC for Quantum Computing Applications

Stefan Visser

This project proposes a solution using a FPGA to create a soft-core ADC architecture. Except for some small resistors on the PCB, the ADC can be completely integrated into the reconfigurable hardware blocks of an FPGA. Therefore the ADC can be easily interfaced with the remainder of the digital circuitry, it can be scaled to the required sampling rate or resolution and it even allows ADCs with different specifications in one system.

This approach allows calibration to each new environment the system is operating in, i.e. changes in voltage, temperature or chip can be calibrated out. We aim to show the effectiveness of our calibration techniques by operating the ADC both at room temperature and in a deep cryogenic environment at 4 Kelvin.


Signal Processing Seminar

Schur Subspace Estimator

Alle-Jan van der Veen

The Schur subspace estimator (SSE) can replace the SVD and GSVD, and is easily updated allowing sliding window tracking of subspaces.

It is based on a generalization of the Schur algorithm. Recall that the Schur algorithm establishes the stability of a polynomial (roots inside unit circle) without explicitly computing the roots. Likewise, the SSE partitions the space into a dominant and a minor subspace w.r.t. a threshold, without computing the SVD.

(Joint work with Mu Zhou, who will defend his PhD thesis on Dec. 4.)

Additional information ...


Signal Processing Seminar

Compressive Sensing-based Video Coding and Range Imaging Platforms

George Tzagkarakis
EONOS Investment Technologies (Paris, Fr.) and ICS-FO.R.T.H (Crete, Greece)

The framework of compressive sensing (CS), acting simultaneously as a sensing and compression protocol, has revolutionized the design of low-complexity onboard remote imaging devices with reduced power and processing requirements. This is achieved by reducing radically the sampling rates dictated by the well-established Shannon-Nyquist theory.

In this presentation, the core properties of CS will be exploited towards improving the performance of existing solutions in two distinct application areas of high industrial interest.

First, we will focus on the design of efficient lightweight video codecs to cope with the growing compression ratios required by modern remote imaging applications. To this end, appropriate CS modules are embedded in the typical MPEGx and M/JPEG systems to exploit frame sparsity in an appropriate residual and/or transform domain, with the main computational burden being put at the decoder, where increased computational and power resources are usually available.

The second application concerns the design of a CS-based methodology for range imaging using time-of-flight cameras, which aims at reducing dramatically the number of necessary frames for reconstructing the depth map of a scene. The use of CS in this case is motivated naturally by the inherent spatial sparsity of objects located in space.

For both application scenarios, we illustrate the effectiveness of CS in improving significantly the performance of previous approaches, thus making it a very promising solution for future high-performance systems.

About the presenter

Dr. G. Tzagkarakis received the Ph.D. and M.Sc. degrees (first in class, Honors) from the Computer Science Department ‐ University of Crete (UoC), Greece, and a B.Sc. degree in Mathematics from the Department of Mathematics (UoC) (first in class, Honors).

Since 2000 he has been also collaborating with the Wave Propagation Group of the Institute of Applied and Computational Mathematics (IACM)‐FO.R.T.H., while in 2002‐2010 he was affiliated as a research assistant in the Telecommunications and Networks Lab (TNL) of the Institute of Computer Science (ICS)‐FO.R.T.H. In the period 2010‐2012 he was a member of the Cosmology and Statistics Lab of CEA/Saclay as a Marie Curie post‐doctoral researcher focusing on the design and implementation of compressive sensing algorithms for remote imaging in areal and terrestrial surveillance systems.

From 2012, he holds a research associate position in EONOS Investment Technologies, working on the development of statistical signal processing algorithms with applications in computational finance and econometrics, while in parallel he is a research collaborator with the Signal Processing Lab (SPL) at ICS‐FO.R.T.H.

His research interests lie in the fields of statistical signal and image processing, with emphasis in non‐Gaussian heavy‐tailed modeling, compressive sensing and sparse representations, with applications in signal processing, image and video processing, and computational finance. Among his topics of interest are also distributed signal processing for sensor networks, information theory, image classification and retrieval, and inverse problems in underwater acoustics.

Additional information ...


MSc SS Thesis Presentation

Automated Detection of Central Apnea in Preterm Infants

Marina Nano

In 2010, an estimated 14.9 million babies were born preterm, which amounted to 11.1% ofalllivebirths worldwide, ranging from about 5% in several European countries to 18% insomeAfrican countries. The rate of preterm births has increased remarkably. Prematurity of birth canpredispose neonates to undesirable cessations of breathing, a conditiontermed as Apnea ofPrematurity. The prevalence of this condition poses problems, becausewhen untreated orinadequately treated Apnea of Prematurity, may impair development.

This thesis investigates theautomated central apnea detection in preterm infants based onraw waveform analysis of one-lead ECG and chest impedance signals. For this purpose,18 novel features and 34 features ofexisting research that characterize different aspectsof chest impedance and ECG signals wereextracted for automated apnea classification.Features aim to extract information regardingrespiratory and cardiac regularity, estimatedfrom chest impedance and ECG signals. Thesefeatures are indicators of some propertiesof cardio-respiratory physiology, which is notindependent of the presence of apnea andthus can be in turn used to classify apnea.

Theobjective is to find the most discriminative subset of features from one-lead ECG and chestimpedance signals that can be usedby a machine-learning approach to study and accuratelydetect central apnea. This wasachieved by applying feature selection algorithms in order toremove redundant or irrelevant features without incurring much loss of information.

In thisthesis, nine hours of continuously recorded data of ten very low-birth-weight infants (birth weight< 1,500 gr) undergoing continuous cardiopulmonary monitoring in the NICU at Maxima MedischCentrumfrom 2008 were included in the analysis. The dataset was annotated by twoneonatologists.Results from this work indicate that the analysis of chest impedance and ECGsignals witha support vector machine can automatically detect Apnea of Prematurity.

Additional information ...


MSc SS Thesis Presentation

Blind Segmentation of Time-Series

Vana Panagiotou

Change-point detection is an indispensable tool for a wide variety of applications whichhas beenextensively studied in the literature over the years. However, the development ofwireless devices andminiature sensors that allows continuous recording of data poses newchallenges that cannot beadequately addressed by the vast majority of existing methods.

In this work, we aim to balancestatistical accuracy with computational efficiency, by developing a hierarchical two-level algorithmthat can significantly reduce the computationalburden in the expense of a negligible loss of detectionaccuracy. Our choice is motivatedby the idea that if a simple test was used to quickly select somepotential change-pointsin the first level, then the second level which consists of a computationallymore expensive algorithm, would be applied only to a subset of data, leading to a significant run-timeimprovement. In addition, in order to alleviate the difficulties arising in high-dimensionaldata, we use adata selection technique which gives more importance to data that are moreuseful for detectingchanges than to others.

Using these ideas, we compute a detectionmeasure which is given as theweighted sum of individual dissimilarity measures and wepresent techniques that can speed up somestandard change-point detection methods.Experimental results on both artificial and real-world datademonstrate the effectivenessof developed approaches and provide a useful insight about thesuitability of some of thestate-of-the-art methods for detecting changes in many different scenarios.

Additional information ...


Distinguished Lecture IEEE COMSOC

A Stochastic Analysis of Network MIMO Systems

Wei Yu
University of Toronto

Network MIMO, where base-stations (BSs) cooperatively transmit and receive to/from the users, promises to significantly alleviate the inter-cell interference problem in wireless cellular networks; but its analytical performance characterization is still a difficult open problem.

In this talk, we describe a stochastic geometry analysis of a network MIMO system, where the multiple-antenna BSs are distributed according to a Poisson point process and cooperate using zero-forcing beamforming to serve multiple users. We obtain tractable and accurate approximations of the signal power and inter-cluster interference power distributions, and derive a computationally efficient expression for the achievable per-BS ergodic sum rate. The analysis enables us to obtain the optimal number of users to schedule. Further, it allows us to quantify the performance improvement of network MIMO systems as a function of the cooperating cluster size.

In particular, due to the zero-forcing penalty across a distributed set of BSs and the inevitable out-of-cluster interference that always exists, the per-BS ergodic sum rate of a network MIMO system does not approach that of an isolated cell even at unrealistically large cluster sizes.

Finally, we illustrate the benefit of user-centric clustering for cell-edge users, and remark on a comparison between massive MIMO and network MIMO systems.

About the presenter

Professor Wei Yu (S97-M02-SM08-F14) received the B.A.Sc. degree in Computer Engineering and Mathematics from the University of Waterloo, Waterloo, Ontario, Canada in 1997 and M.S. and Ph.D. degrees in Electrical Engineering from Stanford University, Stanford, CA, in 1998 and 2002, respectively. Since 2002, he has been with the Electrical and Computer Engineering Department at the University of Toronto, Toronto, Ontario, Canada, where he is now Professor and holds a Canada Research Chair (Tier 1) in Information Theory and Wireless Communications. His main research interests include information theory, optimization, wireless communications and broadband access networks.

Additional information ...


CAS lunch

Let's introduce several new members of the CAS group!


MSc ME Thesis Presentation

Highly Accurate Synchronization Over Ethernet

Jeroen Somers

Additional information ...


Audio Signal Processing Seminar


Audio Signal Processing Seminar


Audio Signal Processing Seminar


Audio Signal Processing Seminar


PhD Thesis Defence

Thermal-Aware Design and Runtime Management of 3D Stacked Multiprocessors

Sumeet Kumar

This dissertation presents architectural techniques to enable the realization of efficient, high-performance chip multiprocessors, and facilitate runtime temperature management to ensure their dependable operation. Most importantly, it provides new insights into the complex thermal behaviour of 3D ICs, and illustrates how the design space of stacked die architectures can be effectively explored in order to maximize performance in the dark silicon era. This dissertation consists of two main themes, architecture and temperature.

The thesis addresses the following questions.

  • How can the performance and efficiency of on-chip memory operations in multiprocessors be improved?
  • How do the physical design parameters in Nagatas equation affect the thermal behavior of 3D Integrated Circuits?
  • How can the knowledge of thermal behaviour be effectively leveraged in the design of 3D stacked multiprocessors?
  • How can the architecture and operating parameters be efficiently adapted at runtime to mitigate the severity of thermal issues, and improve execution performance?

Additional information ...


Audio Signal Processing Seminar

Fast room-mode reproduction in box-shaped rooms: the rigid walls case.

Jorge Martinez-Castaneda

Additional information ...


MSc TC Thesis Presentation

GSVD based blind-beamforming technique for suppression of partially overlapping Bluetooth data packets from WiFi signals

Shailja Shukla

WLAN 802.11 (WiFi) and WPAN 802.15 (Bluetooth) operate in the same 2.4GHz ISM band. OFDM WLANs are designed to provide very high data rates, resilience to multipath and extended operating range. One of the main barriers to actually achieving the high data rates is the interference from the Bluetooth systems which is one of the main sources of interference in the 2.4GHz unlicensed band. This thesis investigates the interference problem and proposes a novel subspace-based method to mitigate the Bluetooth interference inWiFi signals using spatial filtering with an array of antennas. We assume continuous transmission of WLAN packets and partially overlapping, unsynchronous and intermittent Bluetooth interfering packets. The proposed method estimates the target (WiFi) and interfering (Bluetooth) signal subspaces and uses this subspace information to estimate beamformers for interference suppression.

Results show that through the proposed subspace based algorithm the interference of Bluetooth transmission for 802.11 as target model can be reduced and the throughput of 802.11 can be significantly improved at the expense of additional computational complexity.

Additional information ...


MSc TC Thesis Presentation

Compressive Power Spectrum Estimation

Ruijie Zhang

Power spectrum estimation of a wide-sense stationary signal using multi-coset (non-uniform) sampling


MSc TC Thesis Presentation

Sparse Arrays: Vector Sensors and Design Algorithms

Shilpa Rao

Direction-of-arrival (DOA) estimation of acoustic sources is of great interest in a number of applications. Acoustic vector sensors (AVSs) provide an edge over traditional scalar sensors since they measure the acoustic velocity field in addition to the acoustic pressure. It is known that a uniform linear array (ULA) of M conventional scalar sensors can identify up to M-1 DOAs. However, using second-order statistics, the class of sparse scalar sensor arrays have been shown to identify more source DOAs than the number of sensors. In this thesis, we extend these results using sparse AVS arrays. We first assume that the sources are quasi-stationary and use the Khatri-Rao subspace approach to estimate the source DOAs. In addition, a spatial-velocity smoothing technique is proposed to estimate the DOAs of stationary sources. For both scenarios, we show that the number of source DOAs that can be identified is significantly greater than the number of physical vector sensors. The second problem considered in this thesis is sensor selection for non-linear models. It is often necessary to guarantee a certain estimation accuracy by choosing the best subset of the available set of sensors. A non-linear measurement model in additive Gaussian noise is considered. To solve the sensor selection problem, which is inherently combinatorial, a greedy algorithm based on submodular cost functions is developed. The proposed low-complexity greedy algorithm is computationally attractive as compared to existing sensor selection solvers for non-linear models. The submodular cost ensures optimality of the greedy algorithm. Such a sensor selection can be applied, for example, to design sparse AVS arrays that also ensure a certain quality of the DOA estimates next to their identifiability.

Additional information ...


Doppler-resilient Orthogonal Signal Division Multiplexing for Underwater Acoustic Communication

Tadashi Ebihara
University of Tsukuba, Japan

D-OSDM multiplexes several data vectors in addition to a pilot vector, and preserves orthogonality among them even after propagation through doubly spread channels, under the assumption that the channel can be modeled by a basis expansion model (BEM). We describe the signal processing steps at the transmitter and the receiver for D-OSDM, and evaluate its performance by both simulations and experiments. To generate a doubly spread channel, a test-tank with a wave generator is employed. The obtained results suggest that D-OSDM can provide low-power and high-quality UWA communications in channels with large delay and Doppler spreads. Overall, it was found that D-OSDM can become a powerful communication tool for underwater operations.

Additional information ...


MSc ME Thesis Presentation

Implementation of in-situ monitoring techniques for power reduction in smart sensors

Guozhi Xu

Nowadays, smart sensors are widely used in a variety of application domains, such as telecommunication, health care, cars, mobile phones, smart cities. Because of limited battery capacity, low-power design is required for smart sensors. Low-voltage operation is a key leverage to reduce power consumption in smart sensors. However, uncertainties due to process, voltage and temperature variations or random fluctuations gain in relevance when operating in the near-threshold range. Hence, monitoring of the actual silicon behavior is crucial to lowering supply voltage while preserving reliable operation. An interrupt-based in-situ monitoring approach is proposed in this thesis. This approach uses an interrupt service routine to stimulate the critical paths on the chip. By monitoring the timing of the exercised paths, a warning signal is generated to steer the control of reliable supply voltage levels. This approach is developed, validated and applied on an ARM-based processor. Finally, a 10 mV supply voltage margin is achieved based on measurements in the near-threshold range. In addition, reliable operation is verified by running different self-checking codes over multiple dies while varying the environmental conditions.


MSc TC Thesis Presentation

Software Defined Radio Receiver Design Developement for China Digital Radio

Yun Wang

(Thesis work done at NXP, Eindhoven)


MSc SS Thesis Presentation

Distributed Convex Optimization

He Ming Zhang

The Primal-Dual Method of Multipliers (PDMM) is a new algorithm that solves convex optimization problems in a distributed manner.This study focuses on the convergence behavior of the PDMM. For a deeper understanding, the PDMM algorithm was applied to distributed averaging and distributed dictionary learning problems. The results were compared to those of other state-of-the-art algorithms. The experiments show that the PDMM algorithm not only has a fast convergence rate but also robust performance against transmission failures in the network.Furthermore, on the basis of these experiments, the convergence rate of the PDMM was analyzed. Different attempts at proving the linear convergence rate were carried out. As a result, the linear convergence rate has been proven under certain conditions.


MSc SS Thesis Presentation

Heart Rate Variability Analysis based on Instantaneous Frequency Estimation

Di Feng

Additional information ...


MSc ME Thesis Presentation

Low power digital baseband architecture for wireless sensor nodes

Yuteng Hao

This thesis presents a digital baseband design for an upcoming wireless standards: IEEE 802.11ah. It is a branch of Wi-Fi (IEEE 802.11) standards. Compared with the previous Wi-Fi standards, this new standard has larger coverage range and consumes less energy. It is particularly suited for energy-constrained sensor applications. In contrast to the Digital Baseband (DBB)s of other Wi-Fi standards, this design consumes much less power. The basic modulation method of the system is Orthogonal Frequency Division Multiplexing (OFDM) and the detailed algorithms are explored. To prove the robustness of the system, some error tests for the system are performed. A gate-level hardware design and the synthesis netlist are also presented to prove the low-power design. Based on the synthesis results, a series of optimization is done to lower the power consumption. The DBB has been implemented in 40nm Low-power CMOS process to prove the concept. It includes the key blocks of this system. Measurement results show that the DBB for IEEE 802.11ah is suitable for low power applications. The power consumption of this DBB is around 200 - 400 uW, which is hundreds times less than that of the traditional 802.11 baseband.

Thesis work performed at Holst Centre

Additional information ...


3rd International Workshop on Compressed Sensing Theory and its Applications to Radar, Sonar and Remote Sensing

Following the success of the previous two editions of the workshop on compressive sensing applied to radar, we are pleased to announce the third one in this series. The 3rd Int. Workshop on Compressed Sensing Theory and its Applications to Radar, Sonar and Remote Sensing (CoSeRa 2015) will be held in Pisa (Italy) on 22-24 June 2015. The aim of CoSeRa is to bring experts of Compressive Sensing (CS) and radar/sonar/EO/IR signal processing and remote sensing together to explore the state-of-the-art in development of CS techniques for different areas of applications and to turn out its advantages or possible drawbacks compared to classical solutions.

Topics include but are not limited to:

  • Compressive sensing theory
  • Mathematical aspects of Compressive sensing in imaging systems
  • Sparsity of Radar/SAR/Sonar/IR signals
  • Applications of sparse sensing in Radar/SAR/Sonar/IR signal processing
  • Compressive sensing for SAR tomography (TomoSAR)
  • Compressive sensing for SAR Interferometry (InSAR)
  • Compressive sensing for Inverse SAR (ISAR)
  • Target detection based on compressive sensing
  • Compressive sensing for slow GMTI
  • Co-prime sampling in radar/sonar/EO/IR systems
  • Co-prime array processing in radar/sonar/EO/IR systems
  • Nested sampling in radar/sonar/EO/IR systems
  • Netsted array processing in radar/sonar/EO/IR systems
  • Sparse sensing in synthetic aperture imaging systems

Important Dates

  • Full five-page paper submission: February 2, 2015
  • Notification of acceptance: March 16, 2015
  • Final camera-ready papers and author registration: April 20, 2015

All accepted and presented papers will be referenced by IEEEXplore

Additional information ...


Signal Processing Seminar

Speaker Tracking Using Recursive EM Algorithms

Sharon Gannot
Bar Ilan University (Israel)

The problem of localizing and tracking a known number of concurrent speakers in noisy and reverberant enclosures is addressed in this talk. We formulate the localization task as a maximum likelihood (ML) parameter estimation problem, and solve it by utilizing the expectation-maximization (EM) procedure.

For the tracking scenario, we propose to adapt two recursive EM (REM) variants. The first, based on Titterington's scheme, is a Newton-based recursion. In this work we also extend Titterington's method to deal with constrained maximization, encountered in the problem at hand. The second is based on Cappe and Moulines' scheme. We discuss the similarities and dissimilarities of these two variants. The applicability of the proposed methods to localization and tracking problems is demonstrated using both simulated data and recordings from our acoustic lab.

Additional information ...


Signal Processing Seminar

A Simple Model of Speech Communication and its Application to Intelligibility Enhancement

Bastiaan Kleijn

Modern technology allows speech communication from anywhere to anywhere. With phone booths a relic of the past, speech intelligibility has become a common problem, particularly when the listener side is noisy. We will show that it is possible to enhance intelligibility in a noisy listener environment using a formal, information-theory based approach. The new paradigm leads to a family of intelligibility-enhancement algorithms, some of which resemble existing heuristically-derived methods.

Additional information ...


PhD Thesis Defence

Distributed Speech Enhancement in Wireless Acoustic Sensor Networks

Yuan Zeng

In digital speech communication applications like hands-free mobile telephony, hearing aids and human-to-computer communication systems, the recorded speech signals are typically corrupted by background noise. As a result, their quality and intelligibility can get severely degraded. Traditional noise reduction approaches process signals recorded by microphone arrays using centralized beamforming technologies. Recent advances in micro-electromechanical systems and wireless communications enable the development of wireless sensor networks (WSNs), where low-cost, low-power and multi-functional wireless sensing devices are connected via wireless links. Compared with conventional localized and regularly arranged microphone arrays, wireless sensor nodes can be randomly placed in environments and thus cover a larger spatial field and yield more information on the observed signals. This thesis explores some problems on multi-microphone speech enhancement for wireless acoustic sensor networks (WASNs), such as distributed noise reduction processing, clock synchronization and privacy preservation.

First, we develop a distributed delay-and-sum beamformer (DDSB) for speech enhancement in WASNs. Due to limited power of each wireless device, signal processing algorithms with low computational complexity and low communication cost are preferred in WASNs. Distributed signal processing allows that each node only communicates with its neighboring nodes and perform local processing, where communication load and computational complexity are distributed over all nodes in the network. Without central processor and network topology constraint, the DDSB algorithm estimates the desired speech signal via local processing and local communication. The DDSB algorithm is based on an iterative scheme. More specifically, in each iteration, pairs of neighboring nodes update their estimates according to the principle of traditional delay-and-sum (DSB) beamformer. The estimation of the DDSB converges asymptotically to the optimal solution of the centralized beamformer. However, experimental study indicates that the noise reduction performance of the DDSB is at the expense of a higher communication cost, which can be a serious drawback in practical applications.

Therefore, in the second part of this thesis, a clique-based distributed beamformer (CbDB) has been proposed to reduce communication costs of the original DDSB algorithm. In the CbDB, nodes in two neighboring non-overlapping cliques update their estimates simultaneously per iteration. Since each non-overlapping clique consists of multiple nodes, the CbDB allows more nodes to update their estimates and leads to lower communication costs than the original DDSB algorithm. Furthermore, theoretical and experimental studies have shown that the CbDB converges to the centralized beamformer and is more robust for sensor nodes failures in WASNs.

In the third part of this thesis, we propose a privacy preserving minimum variance distortionless response (MVDR) beamformer for speech enhancement in WASNs. Different wireless devices in WASNs generally belong to different users. We consider a scenario where a user joins the WASN and estimates his desired source via the WASN, but wants to keep his source of interest private. To introduce a distributed MVDR beamformer in such scenario, a distributed approach is first proposed for recursively estimation of the inverse of the correlation matrix in randomly connected WASNs. This distributed approach is based on the fact that using the Sherman- Morrison formula, estimation of the inverse of the correlation matrix can be seen as a consensus problem. By hiding the steering vector, the privacy preserving MVDR beamformer can reach the same noise reduction performance as its centralized version.

In the final part of this thesis, we investigate clock synchronization problems for multi-microphone speech enhancement in WASNs. Each wireless device in WASNs is equipped with an independent clock oscillator, and therefore clock differences are inevitable. However, clock differences between capturing devices will cause signal drift and lead to severe performance degradation of multi-microphone noise reduction algorithms. We provide theoretical analysis of the effect of clock synchronization problems on beamforming technologies and evaluate the use of three different clock synchronization algorithms in the context of multi-microphone noise reduction. Our experimental study shows that the achieved accuracy of the three clock synchronization algorithms enables sufficient accuracy of clock synchronization for the MVDR beamformer in ideal scenarios. However, in practical scenarios with measurement uncertainty or noise, the output of the MVDR beamformer with time-stamp based clock synchronization algorithms gets degraded, while the accuracy of signal based clock synchronization algorithms is still enough for the MVDR beamformer, albeit at a much higher communication cost.

Additional information ...


Signal Processing Seminar

Andrea Pizzo

Additional information ...


Signal Processing Seminar

Richard Hendriks

Additional information ...


Signal Processing Seminar

Shahrzad Naghibzadeh

Additional information ...


Signal Processing Seminar

Yan Xie

Additional information ...


Signal Processing Seminar

Maximum Likelihood Self-Estimation for Path-Loss Exponent in Wireless Networks

Yongchang Hu

Additional information ...


Signal Processing Seminar

Thomas Sherson

Additional information ...


Signal Processing Seminar

Jeroen van Gemert

Additional information ...


Signal Processing Seminar

Jia Yan

Additional information ...


QuTech Seminar

Designs for quantum information hybrid devices and systems

Prof. Kae Nemoto
National Institute of Informatics, Quantum Information Sciences, Tokyo, Japan

There have been many architectures for quantum computer and quantum information devices proposed, yet we face a gap between these proof-of-principle idea and feasible quantum devices. We focus on an integrated cavity device based on a single diamond NV center to identify the problems and obstacles by integrating necessary elements to perform computational tasks.

Additional information ...


PhD Thesis Defence

Compressive Power Spectral Analysis

Dyonisius Dony Ariananda

The main focus of this thesis is on the reconstruction of the second-order statistics (including correlations and power spectrum) from digital samples produced by compressive sampling a.k.a. sub-Nyquist-rate sampling. Note that it has been known that compressive sampling offers substantial assistance in sampling rate reduction, which is important when we deal with signals having a very large bandwidth.

What interests us is that there are applications where the sampling still needs to be done at sub-Nyquist rate (due to the high bandwidth of the signal of interest) but where the second-order statistics (instead of the original signal) are of interest. One application is, for instance, spectrum sensing for a cognitive radio network, which is a network where unlicensed radio systems opportunistically search for a currently unoccupied frequency band in the licensed spectrum and then borrow these discovered white spaces to establish a communication link. This spectrum sensing is continuously performed by these unlicensed systems since they have to monitor when the actual owners of the borrowed bands (called licensed users) become suddenly active, in which case the unlicensed radios have to vacate the spectrum. In this application, sampling the signal at sub-Nyquist rate is of interest since the spectral range that has to be sensed is generally very wide. However, note that the unlicensed radio systems are never interested in the original signal of the licensed users occupying the bands to be monitored. This implies that a power spectrum plot describing which frequency bands are occupied together with the amount of power in the occupied bands is more than enough and any efforts to reconstruct the original signal in this application will be overkill. We show that power spectrum reconstruction of WSS signals below the Nyquist rate is possible without any additional constraints on the original signal or the power spectrum.

Other applications are the estimation of directions-of-arrival from more sources than antennas.

Additional information ...


Signal processing mini-symposium

Roberto Lpez Valcarce, Visa Koivunen, Yonina Eldar
University of Vigo, Vigo, Spain. Aalto University, Helsinki, Finland. Technion, Haifa, Israel.

Facilitator: Prof. dr. ir. Geert Leus, Circuits and Systems group

1. Carving the Multicarrier Spectrum

Roberto Lpez Valcarce University of Vigo, Vigo, Spain

Multicarrier modulation has become the format of choice in modern high-speed wireless and wireline systems, due to its many well-known qualities. Nevertheless, the large IFFT sidelobes result in substantial leakage across subcarriers with the ensuing adjacent channel interference. The usual approach of deactivating a number of guard subcarriers at the edges of the signal spectrum is very inefficient in terms of data rate. In order for OFDM to be adopted by future high-performance systems, e.g., 5G, a number of enhancements will become necessary to overcome this and other drawbacks. The leakage problem is also of concern in wideband OFDM-based cognitive systems, in which deep notches must be sculpted in the spectrum to avoid interfering to narrowband licensed users. Judiciously modulating (a few) cancellation subcarriers in order to reduce leakage is an appealing alternative, which originally incurred in high online implementation complexity. We will review this Active Interference Cancellation approach and present efficient designs recently developed in our group, with extensions to linear symbol precoding.

Roberto Lopez-Valcarce received the Ph.D. degree in electrical engineering from the University of Iowa, Iowa City, in 2000. He was a Postdoctoral Fellow of the Spanish Ministry of Science and Technology from 2001 to 2006, with the Signal Theory and Communications Department, University of Vigo, Spain, where he currently is an Associate Professor. His main research interests lie in the areas of adaptive signal processing, digital communications, and sensor networks, having coauthored over 50 papers in leading international journals. He holds several patents in collaboration with industry. Roberto was the recipient of a 2005 Best Paper Award of the IEEE Signal Processing Society. He served as an Associate Editor of the IEEE TRANSACTIONS ON SIGNAL PROCESSING from 2008 to 2011, and as a member of the IEEE Signal Processing for Communications and Networking Technical Committee from 2011 to 2013.

2. Optimal Array Signal Processing in the Face of Non-Idealities

Visa Koivunen Aalto University, Helsinki, Finland

In this talk we describe techniques that facilitate applying high performance array processing algorithms using real-world sensor arrays with nonidealities. The theoretical background is in wavefield modeling that allows one to develop computationally-efficient and asymptotically-optimal array processing methods regardless of the array geometry (conformal arrays). Wavefield modeling also facilitates incorporating array nonidealities into array processing methods and performance bounds. Parameter estimation and beamforming in the azimuth-elevation-polarimetric domain will be addressed. We acquire a realistic array steering vector model by taking into account array nonidealities such as mutual coupling, mounting platform reflections, cross-polarization effects, errors in element positions as well as individual directional beampatterns. This facilitates achieving optimal or close-to-optimal performance and retaining high-resolution capability despite the nonidealities. Moreover, tighter performance bounds may be established for parameter estimation. We describe how the various approaches can be applied in practice in the context of high-resolution direction finding as well as beamforming so that problems related to beamsteering, SOI and interference cancellation are mitigated. This is joint work with Dr. Mario Costa.

Visa Koivunen received his Ph.D in electrical engineering from the University of Oulu, Finland. He was visiting researcher at the Univ of Pennsylvania in 1992-1995. Since 1999 he has been full professor of signal processing at Aalto University (Helsinki Univ of Technology), Finland where he currently holds the Academy Professor position. He has been an adjunct professor at Penn and visiting fellow at Nokia Research Center. He has spent multiple research visits and sabbaticals terms at Princeton University. His research interests include statistical, communication and array signal processing. Dr. Koivunen is an IEEE Fellow and 2015 IEEE SPS Distinguished Lecturer. He received the 2007 IEEE Signal Processing Society best paper award.

3. Sub-Nyquist Sampling: Bounds, Algorithms and Hardware

Yonina Eldar Technion, Haifa, Israel

The famous Shannon-Nyquist theorem has become a landmark in the development of digital signal processing. However, in many modern applications, the signal bandwidths have increased tremendously, while the acquisition capabilities have not scaled sufficiently fast. Consequently, conversion to digital has become a serious bottleneck. Furthermore, the resulting high rate digital data requires storage, communication and processing at very high rates which is computationally expensive and requires large amounts of power. In this talk, we present a framework for sampling and processing a wide class of wideband analog signals at rates far below Nyquist. We refer to this methodology as sampling:combination of compression and sampling, performed simultaneously.

Using the Cramer-Rao bound we develop a generic low-rate sampling architecture that is optimal in a mean-squared error sense, and can be applied to a wide variety of wideband inputs. The resulting system can be readily implemented in hardware, and is easily modified to incorporate correlations between signals. We consider application of these ideas to a variety of problems including low rate ultrasound imaging, radar detection, ultra wideband communication, and cognitive radio, and show several demos of real-time sub-Nyquist prototypes.

Yonina Eldar received the B.Sc. degree in physics and the B.Sc. degree in electrical engineering both from Tel-Aviv University (TAU), Tel-Aviv, Israel, in 1995 and 1996, respectively, and the Ph.D. degree in electrical engineering and computer science from the Massachusetts Institute of Technology (MIT), Cambridge, in 2002. She is currently a Professor in the Department of Electrical Engineering at the TechnionIsrael Institute of Technology, Haifa. She is also a Research Affiliate with the Research Laboratory of Electronics at MIT and a Visiting Professor at Stanford University, Stanford. Dr. Eldar was a Horev Fellow of the Leaders in Science and Technology program at the Technion and an Alon Fellow. In 2004, she was awarded the Wolf Foundation Krill Prize for Excellence in Scientific Research, in 2005 the Andre and Bella Meyer lectureship, in 2007 the Henry Taub Prize for Excellence in Research, in 2008 the Hershel Rich Innovation Award, the Award for Women with Distinguished Contributions, the Muriel & David Jacknow Award for Excellence in Teaching, and the Technion Outstanding Lecture Award, in 2009 the Technion's Award for Excellence in Teaching, in 2010 the Michael Bruno Memorial Award from the Rothschild Foundation, and in 2011 the Weizmann Prize for Exact Sciences. In 2012 she was elected to the Young Israel Academy of Science and to the Israel Committee for Higher Education, and elected an IEEE Fellow. In 2013 she received the Technion's Award for Excellence in Teaching, the Hershel Rich Innovation Award, and the IEEE Signal Processing Technical Achievement Award. In 2014 she was awarded the IEEE/AESS Fred Nathanson Memorial Radar Award. She received several best paper awards together with her research students and colleagues. She is the Editor in Chief of Foundations and Trends in Signal Processing. In the past, she was a Signal Processing Society Distinguished Lecturer, a member of several Signal Processing technical committees, and an associate editor for several IEEE and SIAM journals.

Additional information ...


MEST Symposium

THE SILICON CRYSTAL BALL

Symposium on silicon technology -where speakers from industry, academia and from leading researchcenterswithinNetherlands and from abroad will cover the latest advancements and challenges in silicon technology.

Speakers

  • P. de Jager( ASML) Lithography beyond EUV
  • E. Vreugdenhil (ASML) 3D-NAND Flash: vertical stacking of new thin-film gate-all-around transistors
  • M. Pelgrom (PelgromConsulting) Statistical design has the future
  • Z. Tokei (IMEC) Wiring in 3D
  • F. Rosenboom (TU Eindhoven) Plasma etching for continued semiconductor scaling
  • S. Hamdioui (TU Delft) Computing for Data-Intensive Applications: Beyond CMOS and beyond Von Neumann
  • J. Dorgelo (Marvell) Terabit NAND Flash comes with advanced error correction

Open to all

It is FREE for allMsc, PhD, PD and Professors in Micro-electronics, Computer engineering and Telecommunications. Don't forget to REGISTERatwww.mest-delft.nl

Additional information ...


Signal Processing Seminar

Sparse covariance sensing using vector-sensor array

Shilpa Rao

In directional-of-arrival (DOA) estimation of sources, vector sensors provide an edge over scalar sensors since they measure directional information in addition to the acoustic pressure. We study the covariance sensing problem for a sparse vector-sensor array using two techniques: using a quasi-stationary assumption on the sources, and using spatial smoothing in the correlation domain.

Additional information ...


Sensor Data Fusion @ Fraunhofer FKIE: Surveillance and Protection for Defence and Security Applications

Dr. Wolfgang Koch from FKIE, Bonn
Fraunhofer FKIE, Bonn, Germany

Advanced algorithms to effectively exploit data streams from heterogeneous sources and optimally manage available sensor and unmanned platforms are of crucial importance. The talk will provide an overview of both, methodological work and advanced applications at Fraunhofer FKIE. We will place emphasis on exact track-to-track fusion, multistatic exploration and passive surveillance, aspects of resources management, and fusion tasks with unmanned aerial vehicles.


Signal Processing Seminar

A Krylov Subspace Approach to Modeling Wave Propagation in Open Domain

Jörn Zimmerling

Simulating electromagnetic or acoustic wave propagation in complex open structures is extremely important in many areas of science and engineering. In a wide range of applications, ranging from photonics and plasmonics to seismic exploration, efficient wave field solvers are required in various design and optimization frameworks. In this talk, a Krylov subspace projection methodology is presented to efficiently solve wave propagation problems on unbounded domains. To model the extension of the computational domain to infinity, an optimal, frequency independent complex scaling method is introduced, that allows us to simulate wave propagation on unbounded domains provided we compute the propagating waves via a stability-corrected wave function. In our Krylov subspace framework, this wave function is approximated by polynomial or rational functions, which are obtained via Krylov subspace projection on Polynomial, Extended and Rational Krylov subspaces. In this talk we compare the convergence within these three Krylov subspaces. Further we show how symmetry relations in the finite difference approximation of wave equations can be used to efficiently construct Polynomial and Extended Krylov subspaces. Numerical examples illustrate the performance of the method and show that our Krylov resonance expansions significantly outperform conventional solution methods.

Additional information ...


PhD Thesis Defence

Sensor management for surveillance and tracking. An operational perspective. March 5, 12.00 Aula, Senaatszaal. Promotor A. Yarovoy, co- promotor, H. Driessen

Fotios Katsilieris

Defence, March 5, 12.00 Aula, Senaatszaal. Sensor management for surveillance and tracking. An operational perspective. In the literature, several approaches to sensor (including radar) management can be found. These can be roughly grouped into: a) rule-based or heuristics; b) task-based; c) information-driven; and d) risk/threat-based. These approaches are compared in this dissertation and it is found that there is not a single approach that is both Bayes-optimal and takes into account explicitly the user requirements in different operational contexts. In order to overcome the challenges with the existing approaches, this dissertation proposes managing the uncertainty in higher-level quantities (as per the JDL model) that are directly of interest to an operator and directly related to the operational goal of a radar system. The proposed approach is motivated by the threat assessment process, which is an integral part of defence missions. Accordingly, a prominent example of a commonly used higher-level quantity is the threat-level of a target. The key advantage of the proposed approach is that it results in Bayes-optimal sensor control that also takes into account the operational context in a model-based manner. In other words: a) a radar operator can select the aspects of threat that are relevant to the operational context at hand; and b) external information about the arrival of targets and other scenario parameters can be included when defining the models used in the signal processing algorithms, leading to context-adaptive sensor management.

Additional information ...


Signal Processing Seminar

Novel Symbol- and Frame-level Precoding Algorithms and Connections to Multicasting

Symeon Chatzinotas
Interdisciplinary Centre for Security, Reliability and Trust - University of Luxembourg

Abstract: This seminar focuses on the latest advancements in MU-MISO downlink precoding. Two recent applications which target novel precoding scenarios are reviewed and connection to PHY-layer multicasting are revealed. More specifically, the first part addresses the concept of frame-based precoding which originates in practical standards dictating the grouping of multiple users in a single Forward Error Corrected frame. The second part reviews the field of constructive interference precoding, where the precoding takes place on a symbol-by-symbol basis taking into consideration both the user channel vector and its desired symbol. Finally, a number of promising but challenging open research topics are proposed.

Bio: Dr Symeon Chatzinotas (MEng, MSc, PhD, SMIEEE) received the M.Eng. in Telecommunications from Aristotle University of Thessaloniki, Greece and the M.Sc. and Ph.D. in Electronic Engineering from University of Surrey, UK in 2003, 2006 and 2009 respectively. He is currently a Research Scientist with the research group SIGCOM in the Interdisciplinary Centre for Security, Reliability and Trust, University of Luxembourg, managing H2020, ESA and FNR projects. In the past, he has worked in numerous R&D projects for the Institute of Informatics & Telecommunications, National Center for Scientific Research Demokritos, the Institute of Telematics and Informatics, Center of Research and Technology Hellas and Mobile Communications Research Group, Center of Communication Systems Research, University of Surrey. He has authored more than 120 technical papers in refereed international journals, conferences and scientific books. His research interests are on multiuser information theory, cooperative/cognitive communications and wireless networks optimization. Dr Chatzinotas is the co-recipient of the 2014 Distinguished Contributions to Satellite Communications Award, Satellite and Space Communications Technical Committee, IEEE Communications Society. He is currently co-editing a book on "Cooperative and Cognitive Satellite Systems" to appear in 2015 by Elsevier and he is co-organizing the First International Workshop on Cognitive Radios and Networks for Spectrum Coexistence of Satellite and Terrestrial Systems (CogRaN-Sat) in conjunction with the IEEE ICC 2015, 8-12 June 2015, London, UK.


Signal Processing Seminar

A fast speech analysis method performing simultaneous high resolution voiced unvoiced detection and glottal closure/opening instant estimation.

Andreas Koutrouvelis

Additional information ...


MEST event

TU Delft in ISSCC 2015

Program:

9:00 Prof. Kofi Makinwa Welcome
9:10 A. Carimatto A 67,392 SPAD PVTB-Compensated Multi-Channel Digital SiPMwith 432 column-Parallel 48ps 17b TDCs for Endoscopic Time-of-Flight PET
9:50 M. Shahmohammadi A 1/f Noise Up-conversion Reduction Technique Applied to Class-D and Class-F Oscillators
10:15 R. Quan A 4600um2 1.5oC (3s) 0.9kS/s Thermal-Diffusivity Temperature Sensor with VCO-Based Readout
10:40 Break
10:55 L. Xu A 110dB SNR ADC with +/-30V Input Common-Mode Range and 8uV offset for Current Sensing Applications
11:35 Y. He A 0.05-mm2 1-V Capacitance-to-Digital Converter Based on Period Modulation
12:00 H.Jiang A 30-ppm <80-nJ Ring-Down-Based Readout Circuit for Resonant Sensors

There will be free pizza from 12:45 to 13:15


Signal Processing Seminar

Signal Processing Tools for Radio Astronomy

Millad Sardarabadi

Millad will give an overview of "Signal Processing Tools for Radio Astronomy" that he have developed at Delft, he will cover the problems that arose and remaining challenges to be solved

Additional information ...


Signal Processing Seminar

Dynamic rainfall monitoring using attenuation measurements from telecommunication links.

Venkat Roy

Venkat will present his work in this round of the Signal Processing Seminar.

Additional information ...


Signal Processing Seminar

Acoustic localization and ad-hoc beamforming

Nikolay Gaubitch, Jorge Martinez-Castaneda

A demonstration of a combined acoustic localization and beamforming technology for multi-microphone speech enhancement applications will be given. A short discussion on the theory behind the technology will follow.


Signal Processing Seminar

Computational challenges in Mass Spectral Imaging

Raf Van de Plas
3ME DCSC

Mass spectral imaging (MSI) is a novel imaging modality capable of concurrently measuring the spatial distribution of thousands of molecular species throughout an organic tissue section. This technique, also known as imaging mass spectrometry, has been gaining considerable attention in recent years with applications ranging from medicine to plant science and from material science to forensics. In this talk, we introduce the nature of MSI data with a specific focus on its high-dimensional aspects. We illustrate some of the low-level signal processing challenges using wavelet analysis and matrix factorization examples. High-level biological interpretation challenges are demonstrated using recent work on the automated anatomical interpretation of ion images.


Signal Processing Seminar

Fiber-Optic Communication using Nonlinear Fourier Transforms

Sander Wahls
3ME DCSC

When a signal travels through optical fiber, it evolves in a complicated way that is approximately described by the nonlinear Schroedinger equation. However, it turns out that the evolution of a solution to the nonlinear Schroedinger equation becomes simple when it is considered in the so-called nonlinear Fourier domain. This fact has recently started to attract attention in fiber-optic communications, where the idea has risen to encode information in the nonlinear Fourier domain instead of the time or the conventional Fourier domain. In this talk, the advantages of this concept will be explained. Recent results as well as open questions will also be discussed.


Signal Processing Seminar

Compressive and Sparse Sensing for Statistical Inference

Geert Leus, Alle-Jan van der Veen

Geert Leus with an introduction by Alle-Jan van der Veen. Geert will talk about relevant research topics within our group.

Additional information ...


Signal Processing Seminar

Andrea's Talk

Andrea Pizzo

Additional information ...


MSc CE Thesis Presentation

Profiling of Polyhedral Process Networks

Wouter van Teijlingen

High-level synthesis (HLS) is a design method to raise the level of abstraction in the design of digital circuits. We use HLS to map sequential C-code to Polyhedral Process Networks (PPN), which are implemented in hardware. Designers need feedback on performance limitations as soon as possible, as going through the complete design flow to derive PPNs is time-consuming. The result is that only a few design points can be evaluated in a given amount of time.

In this work, we leverage previous research, and present cprof. Additionally, it estimates the performance of sequential C-code, when implemented as a PPN in hardware. Cprof estimates the execution finish time and the degree of parallelism of a PPN. Cprof provides assistance in Design Space Exploration (DSE), and Hierarchical Program Analysis (HPA) is used to profile programs with inter-procedural behavior.

We verified that cprof is capable of profiling and optimizing sequential C programs, which are realized in hardware as PPNs. We have also shown that on average, cprof overestimates the execution finish time of PolyBench/C benchmarks implemented in hardware by 0.44%. Cprof helps increasing engineering productivity by assisting in DSE, and risk is reduced by making design limitations explicit at an early stage in the design process. The result is that the hardware design flow looks like a regular software design flow, and no special hardware skills are required to analyze and optimize a design that is implemented as a PPN in hardware.

Additional information ...


Signal Processing Seminar

MCMC methods: Brief review and application to constrained blind deconvolution

Georg Kail

A brief and basic review of Markov chain Monte Carlo (MCMC) methods and an application where I have used them

Additional information ...


PhD Thesis Defence

A GPS inspired terrain referenced navigation algorithm

Daniela Vaman

Terrain Referenced Navigation (TRN) refers to a form of localization in which measurements of distances to the terrain surface are matched with a digital elevation map allowing a vehicle to estimate its own position within the map. The main goal of this dissertation is to improve TRN performance through better signal processing. More specifically, the project aims to explore opportunities in the field of TRN by using digital signal processing techniques that were originally developed for the acquisition and tracking of GPS signals.

A typical TRN system uses speed, heading and time to establish the relative horizontal position between subsequent elevation measurements. Thus, any error in speed, heading or time will cause an error in the resulting relative position. If the speed or heading error contains a bias, this will cause a gradual reduction in the correlation. To prevent that a reduction in correlation causes the estimated position to drift away, the idea behind the research described in this thesis is the use of arrays of terrain elevation measurements with intentional (positive and negative) offsets in speed and heading in a tracking-loop configuration. It is well known that such a concept works well for optimized signals such as the ones used in GPS.

To further explore the viability of this idea for a signal defined by a series of terrain elevation measurements, an analysis of similarities and differences with the GPS signal is performed. In accordance to the GPS receiver approach, a novel correlation algorithm for TRN is proposed and implemented. The basic rationale for the algorithm is to use terrain correlation to acquire and track the speed and heading of the host vehicle, while the position advances are calculated using these estimates together with the previously determined position. The novelty of the approach consists in the implementation of a tracking scheme based on the DLL concept. To answer feasibility-related questions, the algorithm is first evaluated in a purely theoretical framework. Based on this analysis it is concluded that the concept seems feasible and promising, but additional considerations in the design are required to compensate for the differences between the GPS and TRN signals. Enhancements are brought to the initial design resulting in the development of an adaptive tracking scheme, in which the tracking loops are configured based on an analysis of the terrain signal.

Next, an in-depth sensitivity analysis is carried out to understand how sensor measurement errors (in speed, heading and terrain height) impact the algorithm performance. The analysis is performed using exclusively simulated data. It is shown that sensitivity to speed and heading errors is dependent on terrain features and it is possible to assess the degree of sensitivity by analysing the terrain signal. By combining this information with the expected error characteristic of the navigation sensors, the performance of the algorithm can be predicted. The sensitivity to terrain measurement errors depends on the ratio between the terrain signal strength and the measurement errors. It is shown that this ratio can be predicted up to a certain extent and a method to improve the ratio is proposed and discussed.

The developed capabilities are validated with recorded sensor data from flight tests. Two different types of recorded sensor data are used: radar and lidar based datasets.

Additional information ...


Signal Processing Seminar

An experimental setup for UWB indoor positioning system

Yan Xie

Additional information ...


TU Delft best graduate ceremony

Final competition

Jörn Zimmerling

The best graduates from each faculty (for EWI: Jorn Zimmerling, member of CAS) will try to convince the jury that they are the Best Graduate 2014 by shortly presenting their graduation project. Together with a professional film crew, the study associations of the faculties made short movies about why their each lecturer deserves the Best Lecturer TU Delft 2014 title. Once again, it promises to be a festive celebration of the people who make our education special. The best graduates and best lecturers count on your support!

Additional information ...


Signal Processing Seminar

Near-field beamforming for ad-hoc microphone arrays

Jorge Martinez-Castaneda, Nikolay Gaubitch

We will give a short talk and demonstration of an autolocalizing beamformer for ad-hoc networks that we developed.

Additional information ...


Signal Processing Seminar

Differential Received Signal Strength-Based Localization with Unknown Path-Loss Model

Yongchang Hu

We present and investigate a new decorrelated model for the DRSS-based localization, on which all the proposed estimators in this paper are based. Then when assuming that the PLE is known, three different kinds of estimators for the DRSS-based localization are introduced: the advanced best linear unbiased estimator (A-BLUE), the semidefinite programing (SDP)-based estimators (SDPE) and the exact estimators (EE). The simulation results show that all the proposed DRSS-based estimators are able to outperform a jointly least squares RSS-based estimator (JLS-RSS) which estimates the unknown transmit power and the target location altogether. Furthermore, they are also shown having a better performance than a recent weighted least squares DRSS-based estimator (WLS) which particularly requires a perfect knowledge of the variance of the shadowing effect. Besides besting the existing methods, the three proposed estimators have their own advantages from different perspectives: the A-BLUE has the lowest computational complexity; the EE holds the best accuracy in a small shadowing environment; the SDPE yields the best performance to endure a large shadowing effect and possesses a very good tolerance, e.g., of using an imperfect PLE. Some more theoretical analyses and comparative discussions are also presented. Finally, based on the studies above, we propose a SDP-based modified block coordinate descent estimator (SDP-MBCDE) to deal with the case when the PLE also becomes unknown. The SDP-MBCDE jointly estimates the unknown PLE and the target location iteratively and, with an increasing iteration number, its performance approaches that of the SDPE using a perfectly-known PLE. In a nutshell, we present a thorough study on the DRSS-based localization based on our proposed decorrelated model. To meet different practical demands when encountering different situations, different proposed estimators are provided as options. Additionally, all the proposed estimators and the corresponding analyses are expected of any help, not only to the DRSS-based localization, but also to the source localization in general or some similar optimization problems.

Additional information ...


Signal Processing Seminar

Like-Tex : Simplify the writing of your PhD thesis (and all your other scientific documents)

Seyran Khademi, Andrea Simonetto, Jorge Martinez-Castaneda

During her PhD the candidate is required to write high-quality scientific documents. To achieve this, Latex and its ecosystem of tools constitute a very powerful yet difficult to learn tool.

In this special session of the SP Seminar we will cover the use of Latex and some of its related tools. Getting to know these tools will allow the user to considerably simplify the workload inherent in writing high-quality scientific documents. In this session we will focus on writing the PhD thesis, but the tips and ideas we give can of course be used for writing documents at any level. Anyone interested is therefore welcome.

The presentation will be given during lunch time, including pizza and mexican burritos. So if you are interested please fill in your name in this doodle so that we can estimate how much food we need to order:

http://doodle.com/sdyswx7idgktuydd

See you all there!

Seyran, Andrea and Jorge.


SP mini symposium

Distributed time-varying optimization

Andrea Simonetto

Abstract: We devise a distributed asynchronous stochastic epsilon-gradient-based algorithm to enable a network of computing and communicating nodes to solve a constrained discrete-time time-varying stochastic convex optimization problem.

Additional information ...


SP Mini symposium

Where am I? An experiment in Indoor Localization

Prof. K.V.S. Hari
Department of ECE, Indian Institute of Science, Bangalore

Abstract: Indoor Localization of people (or objects), where GPS is not available, is an interesting problem with several applications. Several solutions exist, of which, positioning based on WiFi, Video-Tag identification, Ultra Wideband signals and Inertial sensors, are a few examples. In this talk, we consider a scenario where a First-Responder team enters a building after a disaster and the position of each member of the team needs to be known to the control center outside the disaster-affected building. Specifically, we will discuss how an Inertial Navigation System (INS) embedded in a shoe can be designed, to address this problem.

Brief Bio: K.V.S. Hari received the B.E., M.Tech and PhD(1990) degrees from Osmania University, IIT Delhi, University of California at San Diego, respectively. Since 1992, he has been a Faculty Member at the Department of ECE, Indian Institute of Science (IISc), Bangalore, where he is currently a Professor and coordinates the activities of the Statistical Signal Processing Lab in the department. Currently, he is also an Affiliated Professor in the Department of Signal Processing, KTH-Royal Institute of Technology, Stockholm. His current research and development interests include MIMO Wireless Communication, Sparse signal Processing, Indoor Localization and Assistive technologies for the Elderly.


International Radar Conference 2014, Lille, France

Alexander Yarovoy, François Le Chevalier, Fotios Katsilieris, Nikita Petrov, Alexey Narykov, Oleg Krasnov

The French SEE Society (Socit de l'Electricit, de l'Electronique, et des Technologies de l'Information et de la Communication) organises RADAR 2014 in Lille, from 13 to 17 of October 2014. The conference will be organized in the frame of the international relations set up between the Institution of Engineering and Technology (IET), the Institute of Electrical and Electronics Engineers (IEEE), the Chinese Institute of Electronics (CIE), the Institution of Engineers Australia (IEAust) and the SEE.

Additional information ...


Signal Processing Seminar

Joint antenna selection and precoding with quadratic sparsity inducing regularizer for multi-user MIMO systems

Seyran Khademi

Additional information ...


MEST Colloquium

Electronics in Nano-Era: Are we Facing a Reliability Wall?

Said Hamdioui

The talk will address technology scaling and its impact on different aspects of IC and electronics, and in particular the emerging reliability bottlenecks. First the basics of scaling will be covered, together with its impact on integration density, performance and power. The technology outlook will be analyzed in order to extract the challenges with respect to design, test and reliability both for near and long terms. IC realization process will be (re) defined while considering the technology trends and business pressure. Possible ways for the realization of future systems will be discussed.


MSc ME Thesis Presentation Menno Vastenholt

A Sub-GHz UWB Correlation Receiver for Wireless Biomedical Communication

Menno Vastenholt


Receivers Topology Optimization of the Combined Active and WiFi-based Passive Radar Network

Presentation for the EuRAD14 conference

Inna Ivashko

This paper focuses on the accuracy analysis of the combined active and WiFi-based passive radar network. The Cramer-Rao Lower Bound is used as an accuracy metric. It is shown that localization performance of the active radar network can be improved with exploitation of the signals from passive bistatic WiFi radars. This makes reasonable to use information from passive and active radars simultaneously in order to enhance system localization capability. Sparsity-based algorithm is applied to find optimum geometry of the WiFi receivers at the fixed positions of the WiFi access points and active radars.

Additional information ...


Signal Processing Seminar

Indoor granular presence sensing with an ultrasonic circular array sensor

Shahrzad Naghibzadeh

Shahrzad will talk about a paper she is preparing based on her Msc. work.

Additional information ...


MSc CE Thesis Presentation

Guaranteed Quality ECG Signal Compression Algorithm

Dongni Fan

The aim of the project is to develop an ECG signal compression algorithm that has a high compression ratio while guaranteeing signal quality.

An electrocardiography (ECG) signal is a representation of cardiac activity and has an need to be compressed to reduce data storage requirements. Previous ECG signal compression techniques have shown steady improvement on compression ratio. However, these techniques generally lack quality considerations, so their applications are limited. We present a discrete cosine transform (DCT) based compression scheme and use beat detection which considerably improve the compression ratio. The quality of the compressed signal is configurable, and the accuracy of the signal is maintained given a signal quality requirement.

The algorithm is implemented in a software/hardware solution. Some parts need to be done in the software. As a proof of concept, we have chosen the filter to be implemented in hardware. Mathworks HDL coder was used for generating RTL code and testbenches. Results show that our algorithm is capable of maintaining the specified quality, has a better compression ratio compared to previous work and is also capable to compress noisy ECG signals.

Additional information ...


EI Colloquium

Analog-to-digital converters

Jesper Steensgaard
Linear Technology, Milpitas, CA

Analog-to-digital converters have traditionally been a weak link in mixed-mode signal chains. As such, logarithmic and programmable-gain amplifiers have been used to effectively increase their overall dynamic range. In recent years, however, ADC performance has dramatically improved, making it difficult to design amplifiers and references capable of matching their performance. This talk will discuss the challenges of designing a circuit capable of driving a 20-bit SAR ADC with better than 1-part-per-million accuracy.

Biography

Jesper Steensgaard, obtained his MSEE and then his PhD from the Technical University of Denmark in 1999. He has 20+ years of experience in the design of high-resolution data converters. His early work focused on delta-sigma data converters, including mismatch-shaping binary-weighted-element DACs and continuous-time delta-sigma ADCs. Recently, Jesper developed a family of high-resolution low-power SAR ADCs, including the LTC2378-20, which combine the best features of delta-sigma ADCs (precision, low noise) and SAR ADCs (speed, low power, ease of use).


MSc CE Thesis Presentation

CacheBalancer: A communication latency and utilization aware resource manager

Jurrien de Klerk

As the number of processors increases in today's many-core processors, new issues regarding memory management arises. Performance of many-core processors, including large numbers of processors, is often limited by the communication latency due to transfer of data from one node to another. Conventional dynamic memory allocators are unaware of the communication costs, and do not consider what data is send between nodes due to memory allocation. Existing proposals that address this issue result in a limited number of utilized memory resources, potentially leading to over utilized memory resources.

This work introduces a technique for dynamic memory allocation, where state-of-the-art is improved to overcome the limited utilization of the memory resources. The proposed memory allocation method measures the utilization of the different memory resources and uses this information to determine which memory section should be assigned to a requesting task.

This work demonstrates that the proposed memory allocation scheme can reduce memory access latency up to 63.4%, by avoiding the allocation of memory that maps to over utilized resources. The memory allocation scheme is implemented in a run-time manager called Cache Balancer. In addition, the Cache Balancer includes a task mapping algorithm that combines information on the tasks themselves with information communication costs, to map task such that memory throughput is improved. The task mapping algorithm showed a further reduction of memory access latency of 14.5%.

Additional information ...


Quantum imaging seminar

Quantum Nonlinear Optics in Integrated Devices

Marco Liscidini
University of Pavia, Italy

There is a burgeoning interest in the study of parametric fluorescence in integrated devices to obtain compact and efficient sources of non-classical states of light, which are necessary toward the full integration of quantum optical devices. As in the case of quantum linear optics, the use of microstructures gives the opportunity for investigating new phenomena that would be hardly observable in larger bulk-crystal sources. Yet, integrated sources present also several challenges, such as their efficient and fast characterization, which is typically quite demanding because of their low external brightness. In this talk, I will review all these points, starting with the demonstration of silicon microring resonators as CMOS compatible room-temperature sources of time-energy entangled photon pairs.

About the speaker

Marco Liscidini received the Ph.D degree in physics from the University of Pavia (Italy) 2006, working in the group of Prof. Lucio Andreani, with a dissertation entitled "Nonlinear optical properties of planar microcavities and photonic crystal slabs". From 2007 to 2009, he was Post-Doctoral Fellow in the group of Prof. John E. Sipe at the Department of Physics of the University of Toronto, Canada. From 2009 to 2013 is research scientist at the University of Pavia. He is currently tenure-track Assistant Professor at the Department of Physics of the University of Pavia. Since October 2011 is professor of Photonics at the Department of Physics of the University of Pavia. His research activity is focused on the theoretical study and modeling of light-matter interaction in micro- and nanostructures. He works in several areas of photonics, including classical and quantum nonlinear optics, spontaneous emission, plasmon and QW-exciton polaritons, optical sensing and bio-sensing, and photovoltaic effects. He is coauthor of more than 50 papers in peer-reviewed journals. His theoretical research activity is in strong collaboration with experimental groups and in the framework of national, European, and Canadian research programs.


Signal Processing Seminar

Sparse Sensing for Statistical Inference Tasks

Sundeep Prabhakar Chepuri

Additional information ...


Signal Processing Seminar

Recent advances

Amir Leshem
Bar-Ilan University (Israel)

Amir is visiting Delft this week and will talk about his research progress. Amir has been a postdoc and visiting professor at CAS in the past, for in total several years.

Additional information ...


DIMES colloquium

Probabilistic Design for Reliability in Electronics and Photonics: Role, Attributes, Challenges

Prof.dr. Ephraim Suhir
Portland State University, USA

The recently suggested probabilistic design for reliability (PDfR) concept is based on:1) highly focused and highly cost-effective failure oriented accelerated testing (FOAT),aimed at understanding the physics of the anticipated failures and at quantifying, on the probabilistic basis, the outcome of FOATs conducted for the most vulnerable element(s) of the product of interest and the most likely and meaningful combination of possible stressors (the principle of superposition does not work in reliability engineering), and 2) simple and physically meaningful predictive modeling (PM), both analytical and computer-aided, aimed at bridging the gap between what one "sees" as a result of FOAT and what he/she will supposedly "get" in the field. FOAT and PM based sensitivity analysis (SA) algorithms are developed as by-products.

The PDfR concept is based on the recognition of the fact that nobody and nothing is perfect, and that the difference between a highly reliable and insufficiently reliable product is merely in the level of its probability of failure. If this probability (evaluated for the anticipated loading conditions and the given time in operation) is not acceptable, then such a SA can be effectively employed to determine what could be possibly changed, in terms of materials, geometries, application restrictions, etc., to improve the situation.

The PDfR analysis enables one also to check if the product is not "over-engineered", i.e., is not superfluously robust: if it is, it might be too costly: although the operational reliability cannot be low, it does not have to be higher than necessary either, but has to be adequate for the given product and application. This means that when both reliability and cost-effectiveness are imperative, ability to quantify reliability is a must. In this seminar the major PDfR concepts will be illustrated by case studies and practical examples. Although some advanced and subtle PDfR predictive modeling techniques have been recently developed for quantifying and assuring reliability of electronic and photonic products, especially those intended for aerospace applications, the practical examples addressed employ more or less elementary analytical models.

Biography

Prof. Dr. E. Suhir is Fellow of ASME, IEEE, American Physical Society (APS), Institute of Physics (UK), Society of Optical Engineers (SPIE), International Microelectronics and Packaging Society (IMAPS), Society of Plastics Engineers (SPE), Foreign Full Member (Academician) of the NAE, Ukraine, and Fulbright Scholar in Information Technologies. He has authored above 300 publications (patents, books, book chapters, papers) and received numerous professional awards, including 2004 ASME Worcester Read Warner Medal for outstanding contributions to the permanent literature of engineering and laying a foundation of a new discipline Structural Analysis in Electronics and Photonics Systems. Dr. Suhir is the third Russian American, after Steven Timoshenko and Igor Sikorsky, who received this prestigious award. Dr. Suhir is co-founder of the ASME Journal of Electronic Packaging and served as its Technical Editor for eight years (1994-2002).

Additional information ...


MEST welcome drink

Meet and greet your friends and colleagues with a FREE Drink to say Hallo !!!

Organized by MEST student association

Additional information ...


MSc CE Thesis Presentation

Multi-chip dataflow architecture for massive scale biophysically accurate neuron simulation

Jacco Hofmann

The ability to simulate brain neurons in real-time using biophysically-meaningful models is a critical pre-requisite grasping human brain behavior. By simulating neurons behavior, it is possible, for example, to reduce the need for in-vivo experimentation, to improve artificial intelligence and to replace damaged brain parts in patients.

A biophysically accurate but complex neuron model, which can be used for such applications, is the Hodgkin-Huxley (HH) model. State of the art simulators are capable of simulating, in real-time, tens of neurons, at most. The currently most advanced simulator is able to simulate 96 HH neurons in real-time. This simulator is limited by its exponential growth in communication costs.

To overcome this problem, in this thesis, we propose a new system architecture, which massively increases the amount of neurons which is possible to simulate. By localizing communications, the communication cost is reduced from an exponential to a linear growth with the number of simulated neurons As a result, the proposed system allows the simulation of over 3000 to 19200 cells (depending on the connectivity scheme). To further increase the number of simulated neurons, the proposed system is designed in such a way that it is possible to implement it over multiple chips. Experimental results have shown that it is possible to use up to 8 chips and still keeping the communication costs linear with the number of simulated neurons. The systems is very flexible and allows to tune, during run-time, various parameters, including the presence of connections between neurons, eliminating (or reducing) resynthesis costs, which turn into much faster experimentation cycles. All parts of the system are generated automatically, based on the neuron connectivity scheme.

A powerful simulator that incorporates latencies for on and off chip communication, as well as calculation latencies, can be used to find the right configuration for a particular task. As a result, the resulting highly adaptive and configurable system allows for biophysically accurate simulation of massive amounts of cells.


MSc ME Thesis Presentation

Long-range 3D Range Detector Based on Time-correlated Single-photon Counting

Dali Zhang

Three-dimensional (3D) range detectors enabling 3D computer vision is now popular in automotive industry. With their participation, automobile safety has been further enhanced, autonomous driving has become realizable. Time-correlated single-photon counting (TCSPC) technique utilizing complementary metal-oxide semiconductor (CMOS) single photon detectors (SPDs) and time-to-digital converters (TDCs) embodies the proper participant of automotive 3D vision, with low power consumption, low cost, high speed, high robustness, small size, and portability.

In this thesis, a TCSPC 3D range detector for automotive application was studied and modeled. The model covered all main components of a TCSPC system, including the TCSPC range detection process, the signal, and the noise. It was designed to predict the behavior of TCSPC systems and help future designers optimize the performance in accordance with the targeted application.

To verify the model, a experimental setup was designed, implemented, and characterized. The setup consists of a data acquisition system, data processing procedures, and an optical-mechanical system. Measurements performed using the setup have confirmed that the model was designed correctly. For further exploration, range detection from 0.2 m to 60 m were carried out.


CMOS-based implantable electronics for bioscientific and medical applications

Takashi Tokuda
NAIST, Japan

CMOS-based implantable device technology is attracting a lot of interest because of its potential for next-generation bioscientific and medical applications. In this presentation, circuit design, device packaging, and functional demonstration of some CMOS-based implantable devices are presented. An implantable imaging device for in vivo (in a living body) optical brain imaging, implementation of light source for neural stimulation in optogenetics, and flexible neural stimulator for retinal prosthesis will be mainly described.

Additional information ...


MSc SS Thesis Presentation

Speech Production Modelling and Analysis

Andreas Koutrouvelis

The first part of the present thesis reviews the speech production mechanism and several models of the glottal flow derivative waveform and of the vocal tract filter. The source filter model is investigated in depth, since it is the most important "ingredient" of linear prediction analysis. We also review seven linear prediction (LP) methods based on the same general LP optimization framework. Moreover, we examine the importance of pre-emphasis and glottal-cancellation prior to LP.

The second part of the thesis, provides an experimental evaluation of the LP methods combined with several pre-emphasis and glottal-cancellation techniques in the context of two general application areas. The first area consists of applications which aim to estimate the true glottal flow or glottal flow derivative signal. The second area consists of applications which aim to find a sparse residual. In particular, five factors are investigated: the sparsity of the residual using the Gini index, the estimation accuracy of the glottal flow derivative using the signal to noise ratio (SNR), the estimation accuracy of the vocal tract spectral magnitude using the log spectral distortion distance (LSD) metric, and the probability of obtaining a stable linear prediction filter. All these factors are evaluated for clean and reverberated speech signals. The sparse linear prediction methods and the iteratively reweighted least squares method combined with the second order pre-emphasis filter give the most accurate glottal flow derivative estimates, the most accurate vocal tract estimates and the sparsest residuals in most cases. Finally, we compare several linear prediction methods in the context of the speech dereverberation method proposed in [1, 2]. This method enhances the reverberated residual obtained via the autocorrelation method. In the context of this application, we show that the sparse linear prediction method and the weighted linear prediction method combined with a second-order pre-emphasis filter perform better than the autocorrelation method.


MSc TC Thesis Presentation

Network Coding in Underwater Communications

Elvin Isufi


MSc TC Thesis Presentation

Indoor Granularity Presence Sensing and Control Messaging with an Ultrasonic Circular Array

Shahrzad Naghibzadeh


MSc ME Thesis Presentation

Development of a Multichannel TCSPC System in a Spartan 6 FPGA

Harald Homulle

For the master project work was carried out for the development of a fluorescence lifetime imaging probe for fluorescence guided surgery. For this project a prototype was designed. The work on the prototype was divided into three main parts, hardware, firmware / software, and system / optics. In this thesis the firmware / software of the system are described. An overview of the system is given and the performance is evaluated.

Additional information ...


PhD Thesis Defence

Multichannel Digital Silicon Photomultipliers for Time-of-Flight PET

Shingo Mandai

This thesis discusses the potential of CMOS based SiPMs, especially for TOF PET applications, in a systematic and comprehensive fashion. CMOS based SPADs are still need to be designed carefully to improve fill factor, TDCs be improved from the point of the area and power consumption, and the necessity of high voltage for SPADs be handled efficiently. Thus, this thesis also aims to design and integrate various circuits in the SiPM to realize the high integrations utilizing the biggest advantage of the CMOS technology.

Additional information ...


PhD Thesis Defence

Compressive Sampling for Wireless Communications

Shahzad Gishkori

Wireless communications is undergoing massive development in all forms of its manifestations. In the field of short-range communications, technologies like ultra-wideband (UWB) systems are promising very high data rates, fine timing resolution and coexistence with other physical layer standards. Along with these benefits, the promise of low cost and low complexity devices makes UWB systems a highly sought-after option. The main reason for these benefits is the utilization of a very large bandwidth. However, these benefits come at a price, that is the high sampling rate required to receive such signals. According to the Nyquist sampling theorem, a signal can be fully determined if sampled at twice its maximum frequency. This means that the UWB signals may require a sampling rate in the order of Giga samples per second. At the receiver, the sampling is carried out by an analog-to-digital converter (ADC). The power consumption of an ADC is proportional to its sampling rate. A very high sampling rate means stressing the ADC in terms of power consumption. This can put the whole idea of low cost and low complexity UWB systems in jeopardy. Therefore, using subsampling methods is indispensable. In this regard, we propose the utilization of compressive sampling (CS) for UWB systems. CS promises a reasonable reconstruction performance of the complete signal from very few compressed samples, given the sparsity of the signal. In this thesis, we concentrate on impulse radio (IR) UWB systems. IR-UWB systems are known to be sparse, meaning, a large part of the received signal has zero or insignificant components. We exploit this time domain sparsity and reduce the sampling rate much below the Nyquist rate but still develop efficient detectors.

We propose CS based energy detectors for IR-UWB pulse position modulation (PPM) systems in multipath fading environments. We use the principles of generalized maximum likelihood to propose detectors which require the reconstruction of the original signal from compressed samples and detectors which skip this reconstruction step and carry out detection on the compressed samples directly, thereby further reducing the complexity. We provide exact theoretical expressions for the bit error probability (BEP) to assess the performance of our proposed detectors. These expressions are further verified by numerical simulations.

We also propose CS based differential detectors for IR-UWB signals. These detectors work on consecutive symbols. We develop detectors with separate reconstruction and detection stages as well as detectors that perform these steps jointly. We further present detectors which do not need reconstruction at all and can work directly on the compressed samples. However, this can put some limitations on the overall flexibility of the detector in terms of the measurement process. To assess the performance of all these detectors, we also provide maximum a posteriori (MAP) based detectors. We provide numerical simulations to display the detection results.

We extend the CS based classical differential detectors to the case of multiple symbol differential detectors. To keep the implementation complexity at its minimum, we work only with compressed samples directly. We use the principles of the generalized likelihood ratio test (GLRT) to eliminate the limitations on such detectors, in terms of the measurement process. Apart from focusing on compressed detectors which contain full timing information, we also propose detectors which need such information at symbol level only. This effectively results in low cost and low complexity detectors.

Finally, we present some work on the theoretical aspects of CS. We develop algorithms which exploit the block sparse structure of the signal. This block sparsity is combined with varying block sizes and signal coefficients having smooth transitions. Such signals are often encountered in a wide range of engineering and biological fields.

Additional information ...


Signal processing seminar

Relaying Techniques for Multi-Hop Noncoherent UWB Communications

Dr. Vincenzo Lottici
University of Pisa

Ultra-wideband impulse radio has been attracting an interest as a strong candidate for short-range high-rate indoor connectivity, low-rate communications with high-resolution ranging, and location-aware wireless sensor networks. On top of this communication format, multi-hop relaying techniques can effectively contribute to extend the coverage and boost the system performance, especially when the energy available to each relay node of the network is a critical issue. This talk discusses a few novel cooperative approaches for both amplify-and-forward (AF) and decode and forward (DF) relaying. Firstly, we focus on a non-coherent setup employing a double-differential encoding scheme at the source node and a single differential demodulation at the relay and destination. The log-likelihood ratio based decision rule is derived at the destination node, and a semi-analytical power allocation strategy is presented by evaluating a closed-form expression for the effective signal to noise ratio (SNR) at the destination. Secondly, we focus on developing a single differential encoded DF non-cooperative relaying scheme. To favor simple receiver structures, differential noncoherent detection is employed which enables effective energy capture without any channel estimation. Putting emphasis on the general case of multi-hop relaying, we illustrate an original algorithm for the joint power allocation and path selection (JPAPS), minimizing an approximate expression of the overall bit error rate (BER). After deriving a closed-form power allocation strategy, the optimal path selection is reduced to a shortest path problem on a connected graph, which can be solved without any topology information with complexity O(N3), N being the number of available relays of the network. An approximate scheme is also presented, which reduces the complexity to O(N2), while showing a negligible performance loss. For benchmarking purposes, an exhaustive-search based multi-hop DF cooperative strategy is taken into consideration. Simulation results for various network setups corroborate the effectiveness of the proposed low-complexity JPAPS algorithm, which favorably compares to existing AF and DF relaying methods.
CV. Since 1993, Vincenzo Lottici has been with the Department of Information Engineering of the University of Pisa, where he is currently Professor in Communication Systems. He participated in several international and national research projects, and as TPC member, in numerous IEEE conferences in wireless communications and signal processing, He recently joined the Editorial Board of EURASIP Advances on Signal Processing. His research interests include the broad area of signal processing for communications, with emphasis on synchronization, channel estimation, dynamic resource allocation, cognitive radio and compressive sensing.

Additional information ...


Signal Processing Seminar

Noncoherent Decision-Feedback Equalization in Massive MIMO Systems

Dr. Robert Fischer

We discuss noncoherent receivers in multi-user massive MIMO uplink systems. On the one hand, sorted decision-feedback differential detection (DFDD) is attractive for the detection of a particular user over a transmission burst. On the other hand, we present a noncoherent approach to decision-feedback equalization (DFE) over the users. Thereby, the contradicting principles of DFE, where interference of already detected symbols is canceled using actual channel knowledge, and noncoherent reception, where the symbols are detected without any channel-state information, are combined. Based on an analysis of the statistics of the interference terms in autocorrelation-based noncoherent receivers, DFE is proposed and optimized. Moreover, we discuss a joint user/temporal detection with optimized sorting and an iterative scheme. Numerical results quantifying the performance of the noncoherent schemes relative to coherent BLAST, taking the non-perfect channel knowledge due to finite-length training sequences into account, are presented. CV. Since 2011 Dr. Robert Fischer has been full professor for communications and signal theory at the University of Ulm, Germany. Currently, he teaches undergraduate and graduate courses on signals and systems and on digital communications. His research concentrates on fast digital transmission including single- and multicarrier modulation techniques. Current interests are information theory, coded modulation, digital communications and signal processing, and especially precoding and shaping techniques for high-rate transmission schemes.

Additional information ...


Signal Processing Seminar

Channel Estimation and Hybrid Precoding for Millimeter Wave Cellular Systems

Geert Leus

Millimeter wave (mmWave) cellular systems will enable gigabit-per-second data rates thanks to the large bandwidth available at mmWave frequencies. To realize sufficient link margin, mmWave systems will employ directional beamforming with large antenna arrays at both the transmitter and receiver. Due to the high cost and power consumption of gigasample mixed-signal devices, mmWave precoding will likely be divided among the analog and digital domains. The large number of antennas and the presence of analog beamforming requires the development of mmWave-specific channel estimation and precoding algorithms. This talk discusses an adaptive algorithm to estimate the mmWave channel parameters that exploits the poor scattering nature of the channel. To enable the efficient operation of this algorithm, a novel hierarchical multi-resolution codebook is designed to construct training beamforming vectors with different beamwidths. Using the estimated channel, a new hybrid analog/digital precoding algorithm that is proposed that overcomes the hardware constraints on the analog-only beamforming, and approaches the performance of digital solutions. Simulation results show that the proposed low-complexity channel estimation algorithm achieves comparable precoding gains compared to exhaustive channel training algorithms. The results also illustrate that the proposed channel estimation and precoding algorithms can approach the coverage probability achieved by perfect channel knowledge even in the presence of interference.

Additional information ...


Signal Processing Seminar

Sparsity-Aware Sensor Selection: Centralized and Distributed Algorithms

Hadi Jamali-Rad

Additional information ...


MSc TC Thesis Presentation

Modeling of wave propagation in open domains: A Krylov subspace approach

Jorn Zimmerling

Simulating electromagnetic or acoustic wave propagation in complex open structures is extremely important in many areas of science and engineering. In a wide range of applications, ranging from photonics and plasmonics to seismic exploration, efficient wave field solvers are required in various design and optimization frameworks.

In this talk, a Krylov subspace projection methodology is presented to efficiently solve wave propagation problems on unbounded domains. To model the extension of the computational domain to infinity, an optimal complex scaling method is introduced. Traditionally, complex scaling has been used to simulate open quantum systems. Here, an optimized complex scaling method is implemented that allows us to simulate wave propagation on unbounded domains provided we compute the propagating waves via a stability-corrected wave function. In our Krylov subspace framework, this wave function is approximated by polynomial or rational functions, which are obtained via Krylov subspace projection. We show that the field approximations are actually expansions in terms of approximate open resonance modes of the system and we present a novel and highly efficient Krylov subspace implementation for media exhibiting second-order relaxation effects. Numerical examples for one-, two-, and three-dimensional problems illustrate the performance of the method and show that our Krylov resonance expansions significantly outperform conventional solution methods.


Signal Processing Seminar

Venkat Roy


ISCAS 2014

2014 IEEE International Symposium on Circuits and Systems

Welcome from the General Chairs of the Organising Committee

On behalf of the Organising Committee we welcome you to Melbourne, ranked by the Economist Intelligence Unit in 2011, 2012 and 2013 as the most liveable City in the world, to Australia, and to the 2014 IEEE International Symposium on Circuits and Systems.

ISCAS2014 is sponsored by the Institute of Electrical and Electronic Engineers Circuits and Systems Society (IEEE CASS), and generously supported by the State Government of Victoria and the Melbourne Convention Bureau.

As you all know, ISCAS is the flagship annual conference of IEEE CASS, and it is well established as the worlds premier networking forum in the fields of theory, design and implementation of circuits and systems. As a result of the release of its 2012 Vision and Mission (see ieee-cas.org), the CASS goal is to develop ISCAS also as the leading forum for pioneering circuits and systems contributions to humanitys grand challenges.

Accordingly, the special theme of ISCAS 2014 is nano/bio circuits and systems applied to enhancing living and lifestyles, particularly in relation to the multidisciplinary grand challenges in healthcare and wellbeing, the environment and climate change.

Keynotes

ISCAS2014 has four keynote presentations, two of which address crucial aspects of high priority grand challenges, in health and in sustainability, while the other two describe frontier work at the extreme ends, in terms of scale, of circuits and systems engineering new devices that promise to sustain the remarkable advances in semiconductors that we have enjoyed for over 60 years, and design methods for systems of systems, which are relevant to so many grand challenge problems:

Dr Donald E. Ingber from Harvard University on Monday will present Microengineered Human Organs On Chips, describing advances he and his team have made in the engineering of microfluidic Organs-on-Chipsmicrochips lined by living human cells created with microfabrication techniques that recapitulate organ-level structure and functions as a way to replace animal testing for drug development and mechanistic discovery.

Professor Iven Mareels from The University of Melbourne in his talk on Wednesday, titled Circuits and Systems for Modern Irrigation Management, describes work over 15 years on circuits and systems research, development and commercialisation of an internet-of-things dedicated to smart irrigation water management.

Professor Victor Zhirnov from the Semiconductor Research Corporation, in Scaling Limits of Nanoionic Devices, elaborates how recognition that crystal defects could be used as controllable entities, rather than being seen as imperfections, leads to the possibility that nanoionic resistive switching devices may be scalable down to ~ 1nm and thus may offer a promising path to replace the foundation of todays computing technologies.

Dr. Stephan C. Stilkerich from Airbus Group will present Model Based Engineering of Highly Mobile Systems of Systems: Safe Aeroplanes; Safer Automobiles, with an introduction and post-talk discussion moderated by Dr Graham Hellestrand from Embedded Systems Technology. This keynote deals with front-line approaches to engineering electronic systems and their software, that are required to perform real-time control critical for the safe operation of airplanes and cars, including while operating in dense traffic and simultaneously reducing environmental impact.

Technical Program Regular Sessions

The technical program consists of tutorials, lecture papers, poster papers and demonstrations accepted based on peer review of the submission from regular open calls. We have retained many of the ISCAS features that have evolved in recent years, and added new features, some in response to ISCAS feedback, to continue to improve attendees experience of the event.

We are very pleased to report that ISCAS2014 will be first time that the new CASS Conference App will be made available to all attendees, and we look forward to your feedback to improve it. The CAS Society has supported the development of the Conference App, through Conference4Me, to facilitate the navigation of the conference agenda and venues, secure access to proceedings, micro-blogging, live discussion and ranking of papers, providing feedback to organizers and general improvement of attendees experience at CAS conferences.

Lecture papers follow the traditional ISCAS format. There are nine lecture sessions over three days, with session having 11 parallel streams. Sessions are 90 minutes with up to five papers, allowing 18 minute for each including introduction, presentation and discussion.

The Demonstration session and Poster sessions are held over 3 hours commencing at the 3pm coffee break on Monday, Tuesday and Wednesday. The Demonstration session is Monday only. There are no competing parallel lecture sessions during the first 90 minutes of each days Posters/Demonstrations, allowing increased attention to them from all attendees.

We have increased the length of the lunch break to 90 minutes. This will allow more time for the CASS side meetings, particularly the annual meetings of the 15 CASS Technical Committees, which are playing an increasingly important role in leadership of the Society. The longer break will also provide a more relaxed walk to the nearby restaurants for the lunch break, and we hope it will facilitate a greater level of networking.

Following the ISCAS2013 lead we continue the trial of offering free attendance at Tutorial and CAS-FEST sessions for all ISCAS2014 registrants. We have also expanded both the tutorial program and CAS-FEST. CASS goals in these moves are both to widen the reach of and to increase participation in the tutorial program and CAS-FEST. We will greatly appreciate feedback from attendees on the value you perceive in these offerings.

Tutorials

ISCAS2014 commences on Sunday with 19 half-day and 1 full-day Tutorial sessions.

We have included two Tutorial sessions on Technology Management in response to feedback from CASS industry members:

T19 Interfacing Organisations: How to successfully manage organizational interfaces by Felix Lustenberger; and T20 - Managing Technology Professionals by Tuna B. Tarim: Transitioning from Individual Contributor to Management. Felix and Tuna are CASS members and also leaders of IEEEs Technology Management Council, which was recently approved to transition to an IEEE Society.

Also in response to feedback, from the Women in CAS (WiCAS) and Young Professionals Program (formerly GoLD) groups, is a tutorial on career development, social skills, collaboration and networking:

T7 Engineering Networks that Work: Design Tools for Your Career by Dr Margaret Collins Margaret is a Cardiff-based research consultant, professional coach and trainer with extensive experience in helping people achieve their career goals. Come ready to get involved this is an active workshop session!

A third initiative in the Tutorials is a full day introduction to Memristive devices, circuits, systems and applications, the topic of this years CAS-FEST. This will cover all aspects of this emerging technology, namely: theory, practical nanodevices, physical switching mechanisms, circuits and emerging applications:

T21 If its Pinched its a memristor (AM), Professor Leon Chua T22 ReRAM Memristive Devices: Electrochemical Systems at the Atomic Scale (AM), Dr Ilia Valov T23 Analog and Mixed-Signal Applications of Memristive Devices (PM), Professor Dmitri Strukov T24 Integrating memristive devices in CMOS neuromorphic computing architectures (PM), Professor Giacomo Indiveri The aim of these sessions is to provide sufficient introduction to enable a typical ISCAS attendee to appreciate the state of the art material that will be presented in the CAS-FEST sessions.

CAS-FEST

Since its inception in 2010, the Circuits and Systems Forum on Emerging and Selected Topics has progressively become more closely integrated with ISCAS. This years topic was again selected from an open call and the winning proposal, from members of the Nonlinear Circuits and Systems (NCAS) Technical Committee, has taken still further this level of integration with ISCAS. This includes the presentation of invited introductory tutorials in the regular ISCAS Tutorials program (see above), the inclusion of three Special Sessions in the regular ISCAS Lecture Papers program, a full day of CAS-FEST Special Sessions on Wednesday, and the highlight full day of CAS-FEST Keynote talks on Thursday.

With this additional integration, CAS-FEST 2014 will bring together leading experts and provide a thorough coverage of the field of memristors, from an introduction to those unfamiliar with the field, through solidifying existing knowledge, to highlighting developments at the forefront of the field, and pointing to future challenges and promising directions for research. We hope that this coordinated approach will result in a landmark event in the development of the field.

Social Events and Awards Dinner

We are planning the now standard set of ISCAS evening events, with the Welcome Reception on Sunday evening soon after the conclusion of Tutorials, the WiCAS/YPP (formerly GoLD) event on early Monday evening, the Awards Dinner on Tuesday evening, and the Closing Reception immediately following the last session on Wednesday. Watch out for the Australian twists!

We hope that you will have a rewarding and enjoyable time in Melbourne at ISCAS2014 and look forward to meeting as many of you as we can!

Professor Jugdutt (Jack) Singh & Dr David Skellern General Co-Chairs, ISCAS 2014

Additional information ...


Signal Processing Seminar

Rocio Arroyo-Valles


MSc ME Thesis Presentation

Physical design of a 3D router

Milovan Vasic

With the use of multi-core architectures, the Network-on-Chip (NoC) became an important research topic. The most important benefit of a NoC compared to a communication bus is that it is scalable. The heart of the NoC is the router, which provides the communication between different computational units. This component is highly suitable to be a 3D component, which means that the connection can go into a vertical direction. This way the NoC is extended, with the same area footprint. This thesis describes the physical design of the 3D router, where various design problems are solved. An existing router architecture is used as a start-point. One of the problems which this thesis is trying to solve, is the reduction of the number of needed data lines. This is especially useful for the vertical data lines, which are implemented with Through Silicon Vias (TSVs). A TSV has a large footprint compared to a transistor, which means it takes up a lot of chip area. This results in increased cost. The reduction of the data lines is accomplished by the serialization of the data. It is determined that the best serialization ratio is 4. The 3D router is also adjusted for asynchronous operation. This is accomplished with the use of FIFOs, two-flop synchronizers and Gray Encoders.


Joint clock synchronization and delay estimation under quadratic clock model in wireless networks

"Joint clock synchronization and delay estimation under quadratic clock model in wireless networks"

Yan Xie

Additional information ...


Conferences

IEEE ICASSP'2014

IEEE ICASSP is the world's largest and most comprehensive technical conference focused on signal processing and its applications. The series is sponsored by the IEEE Signal Processing Society and has been held annually since 1976. The conference features world-class speakers, tutorials, exhibits, a Show and Tell event, and over 120 lecture and poster sessions.

Additional information ...


Signal Processing Seminar

Compressive Cyclic Spectrum Reconstruction

Dyonisius Dony Ariananda

Additional information ...


Signal Processing Seminar

GPU-Accelerated Adaptive Unstructured Road Detection Using Close Range Stereo Vision

Bugra Ozutemiz, PhD candidate, Middle-East Technical University

Detection of road regions is not a trivial problem especially in unstructured and/or off-road domains since traversable regions of these environments do not have common properties. Even the properties and appearance of these environments can change on the run. Hence, an algorithm working under unstructured conditions should have a continuous adaptation capability. To achieve this, a novel unstructured road detection algorithm that can continuously learn the road region is proposed in this work. The algorithm gathers close-range stereovision data using a simple roughness threshold and uses this information to estimate the road region in the far field. The proposed approach simplifies over the approaches in the literature by changing offline supervised learning and pose estimation of the vehicle and sensor with a simple heuristic coming from the nature of the problem: roughness (or smoothness) of the terrain. Thanks to the parallel nature of the algorithm, it is also implemented on a GPU with CUDA and a real-time running performance is achieved even on a low-performance graphics card. The experiments show that the algorithm gives excellent road detection results even under fast-changing light conditions and a running frequency of 50 Hz is achieved even in the worst case.


Signal Processing Seminar

Jorge Martinez-Castaneda

Additional information ...


MSc TC Thesis Presentation

Joint Angle-Frequency Estimation for Multiple Signals with Circular Arrays

Joost Geelhoed
CAS, TNO

(Work carried out at TNO Defence, Safety and Security--The Hague)

In electronic warfare information about radio signals is gathered. Parameters as the directionof- arrival (DOA) and the frequency can be estimated from sampled data received on antenna arrays. The objective of this thesis is to design a joint angle-frequency estimation (JAFE) algorithm for a circular uniform antenna array.

A 1-D and 2-D JAFE algorithm is presented. Both algorithms use phase-mode excitation and ESPRIT. The 1-D algorithm with spatial smoothing is introduced. With this algorithm it possible to estimate signals with similar frequencies, when the elevation is 90 degrees. Simulations demonstrate that when two signals are coherent and a spatial smoothing factor of two is applied the mean of the azimuth estimation is the true mean. It is also demonstrated that when two signals have the same DOA a temporal smoothing factor of two is necessary and that more temporal smoothing reduces the standard deviation of the azimuth estimation. It is shown that the phase-mode excitation technique introduces a systematic error that is considerably high for few antenna elements and an even number of elements. It is demonstrated that interpolation can reduce this error in case a UCA of 5 elements or a UCA of 12 elements.


Signal Processing Seminar

Tamas Keviczky
3ME DCSC

Additional information ...


Signal Processing Seminar

OFDM Analysis: an Industry Point of View

Earl McCune

With more than 40 years of experience in wireless communications technology and its associated hardware construction, this presenter finds the current flurry of interest in orthogonal frequency division modulation (OFDM) very interesting. It is surprising that the extremely revolutionary nature of this signal class is actually poorly understood, particularly by the academic community. On the industrial side, there is a general misunderstanding of the economic consequences of building hardware needed to generate and receive an OFDM signal. Here the motivation and physical principles of OFDM are built from fundamental concepts, from which the signal characteristics are evaluated and compared with conventional QAM. Conclusions on the comparative economics of OFDM are drawn.


EUCAP14 conference on antennas and propagation

Abstract submission: 13 Oct 2013

Conference dates: 6-8 April 2014

Application areas:

  • Fundamental research
  • Satcom on-the-move terminal antennas
  • Navigation, localisation, positioning and tracking
  • Cellular mobile communications (includes: base station, handheld devices)
  • Machine to machine, internet on devices
  • Wireless networks (includes: WLAN, indoor communication)
  • High data-rate transfer and backbone networks
  • RFID and sensor networks
  • Biomedical (includes: human body interaction, on-body antennas, electromagnetic exposure and interactions)
  • Satellite communications
  • Satellite passive and active remote sensing
  • RADAR
  • Radio astronomy
  • Signal and image processing
  • Defense and security
  • Short-range Giga-bit communications
  • Commercial software

Additional information ...


Signal Processing Seminar

Andreas Loukas
EWI, Embedded Systems group

Additional information ...


Signal Processing Seminar

Millad Sardarabadi


Signal Processing Seminar

Speech reinforcement in noisy reverberant environments using a perceptual distortion measure

Joao Crespo

Additional information ...


Signal Processing Seminar

On indoor localization based on acoustic signals and channel fingerprinting

Yongwei (Enki) Wang
NWPU, China

Additional information ...


MSc CE Thesis Presentation

Modeling of Olivocerebellar Neurons using SystemC and High-Level Synthesis

M.F. van Eijk

Neuro-scientific experiments often require a heavy amount of computational power to achieve a more efficient research process. Speeding up neuron models, used in such experiments, facilitates faster testing of scientific hypotheses and faster model refinement in order to better replicate the biological-cell behavior in question. To be able to simulate realistic behavior, high-detail neuron models need to be built and typically require maximal computing power. Hardware acceleration can be used to execute such simulations in a realistic amount of time. However, building a hardware implementation is very time-consuming and error-prone. Recent trends in RTL design have resulted in tools which can convert a high-level behavioral hardware description to an RTL description targeting an FPGA, which makes them ideal tools for complex FPGA designs, such as high-detail neuron models. In this thesis, a model of the Inferior-Olivary Nucleus (ION) network has been implemented with SystemC RTL and mapped onto an FPGA using a High-Level Synthesis (HLS) tool-flow. A shared-bus architecture has been used as proof of concept to interconnect the various cells in the modeled network. SystemC Transaction Level Modeling (TLM) facilitates fast network-interconnect modeling and verifying of model functionality. We have thus developed a SystemC TLM model that can predict trends of our RTL implementation. With this TLM model we were able to quickly model large network sizes and assess the models scalability with respect to utilized resources and performance. The complete network model has, subsequently, also been synthesized. Xilinx Vivado HLS has been used to convert the SystemC implementation to an RTL description mapped on a Virtex 7 (XC7VX550T) FPGA device. The resulting design achieved a speed-up of x6 compared to a reference C model, making it possible to simulate a network of 48 cells in real-time. Because HLS tools are used, the model can be easily modified to accommodate last-minute changes and models updates by the neuroscientific community.


Signal Processing Seminar

Current and future research lines in audio signal processing

Richard Heusdens


Signal Processing Seminar

AIS ship identification signal separation using GSVD and the Signed URV algorithm

Alle-Jan van der Veen

Separating partially overlapping data packets using subspace intersection.


Signal Processing Seminar

Low-complexity computer simulation of multichannel room impulse responses

Jorge Martinez-Castaneda
CAS

Additional information ...


Signal Processing Seminar

Rocio Arroyo-Valles
CAS

Additional information ...


Signal Processing Seminar

Relative velocity estimation using Multidimensional Scaling

Raj Thilak Rajan
CAS

Additional information ...


Signal Processing Seminar

Tracking Position and Orientation of a Mobile Rigid Body

Sundeep Prabhakar Chepuri
CAS

Additional information ...


PhD Thesis Defence

Spectrum sensing for cognitive sensor networks

Sina Maleki

Additional information ...


Signal Processing Seminar

Resource allocation for mobile ad hoc networks

Philippe Ciblat
TELECOM ParisTech

Additional information ...


Signal Processing Seminar

ICA and IVA: Theory, Connections, and Applications to Medical Imaging

Tulay Adali
University of Maryland at Baltimore County

Data-driven methods are based on a simple generative model and hence can minimize the assumptions on the nature of data. They have emerged as promising alternatives to the traditional model-based approaches in many applications where the underlying dynamics are hard to characterize. Independent component analysis (ICA), in particular, has been a popular data-driven approach and an active area of research. Starting from a simple linear mixing model and imposing the constraint of statistical independence on the underlying components, ICA can recover the linearly mixed components subject to only a scaling and permutation ambiguity. It has been successfully applied to numerous data analysis problems in areas as diverse as biomedicine, communications, finance, geophysics, and remote sensing.

This talk reviews the fundamentals and properties of ICA, and provides a unified view of two main approaches for achieving ICA, those that make use of non-Gaussianity and sample dependence. Then, the generalization of ICA for analysis of multiple datasets, independent vector analysis (IVA), is introduced and the connections between ICA and IVA are highlighted, in particular in the way both approaches make use of signal diversity. Several key problems for achieving a successful decomposition, such as matrix optimization and density matching are discussed as well, along with examples of their application to medical image analysis.

Additional information ...


Signal Processing Seminar

Yongchang Hu
CAS

Additional information ...


Signal Processing Seminar

Constrained Imaging with Active-set methods

Millad Mouri Sardarabadi
CAS

Can radio astronomy imaging be improved by incorporating constraints such as non-negativity and known upper bounds?


Signal Processing Seminar

Secure Signal Processing: Challenges and Opportunities

Zekeriya Erkin
TU Delft ISP Lab

Additional information ...


Short Course

Network Information Theory: Cooperation and Interference

Prof.dr. M.C. Gastpar
EPFL

Information Theory has proved to be a relevant and successful foundation for the understanding of point-to-point digital communication systems. In today's networked world, information flows in more complex patterns: We can think of networks of sensors or cameras that are wirelessly connected or of cooperating base stations in mobile telephony systems. Over the past decade, significant progress has been made towards extending information theory to such communication systems. In this short course, we survey both classical results and recent progress in Network Information Theory.

Registration:

This course is intended for interested PhD/MSc students with some background in information theory. They should register before Monday, September 2, 2013, via e-mail to J.H.Weber@tudelft.nl

Credit:

Upon completion of the class, including homework, PhD students in the TU Delft Graduate School will earn 2 credits for their doctoral education program. Other students can receive a certificate.

Outline:

The short course will be taught mostly on the blackboard, along with some slides. It contains four modules:
  • Review of Basic Information Theory
  • Broadcast and Multiple Access
  • Techniques for Relaying and Cooperation
  • Managing Interference
Two short homework sets will be provided to the students.

Textbook:

Recommended (but not necessary) is the book by Abbas El Gamal and Young- Han Kim, Network Information Theory, Cambridge University Press, 2013.

Additional information ...


MSc TC Thesis Presentation

Compressive power spectral density estimation with non-uniform sampling

Fernando de la Hucha Arce


MSc TC Thesis Presentation

Photo-acoustic imaging

Victor Sloev

A new non-invasive imaging modality is based on transmitting a laser pulse on the skin; the created heat will cause an acoustic pulse that is detected with an ultrasonic array. How can an image be formed?


Signal Processing Seminar

Optimal power allocation for energy harvesting cognitive radio networks

Ning
CAS


Signal Processing Seminar

Location tracking

Raj Thilak Rajan
CAS


Signal Processing Seminar

Underwater localization

Hamid Ramezani
CAS


PhD Thesis Defence

Wireless Transceiver Design For High Velocity Scenarios

Tao Xu

This thesis is dedicated to transceiver designs for high data-rate wireless communication systems with rapidly moving terminals. The challenges are two-fold. On the one hand, more spectral bandwidth of the transmitted signals is required by future wireless systems to obtain higher transmission rates, which can result in the frequency selectivity of the communication channels. On the other hand, Doppler effects emerge when high mobile speeds are present, which can result in the time selectivity of the communication channels. Therefore, it is likely that future wireless communication systems operate in doubly-selective channels, which impose many difficulties on transceiver designs. In this thesis, we investigate these challenges in the four scenarios, and propose a number of corresponding solutions.

  1. OFDM over Narrowband Channels;
  2. OFDM over Wideband Channels;
  3. Multi-Rate Transmissions over Wideband Channels;
  4. Robust Multi-band Transmissions over Wideband Channels.


MSc Thesis Presentation

Underwater Ultra-Wideband Fingerprinting-Based Localization

Siavash Shakeri

In this work a new location fingerprinting-based localization algorithm is proposed for an underwater medium by utilizing ultra-wideband (UWB) signals. In many conventional underwater systems, localization is accomplished by utilizing acoustic waves. On the other hand, electromagnetic waves havent been employed for underwater localization due to the high attenuation of the signal in water. However, it is possible to use UWB signals for short-range underwater localization. In this work, the feasibility of performing localization for an underwater medium is illustrated by utilizing a location-based fingerprinting approach. Existing algorithms for an indoor environment are evaluated in this project for an underwater medium. These algorithms are based on a neural networks or maximum likelihood estimator. Further, we also consider a classical k-nearest neighbors (KNN) approach. In addition, by employing the concept of compressive sampling, we propose a sparsity-based localization approach for which we define a system model exploiting the spatial sparsity. Moreover, a recently proposed grid mismatching algorithm is also adapted to the current localization framework and its performance is evaluated. Finally, the performance of the proposed methods is compared with the existing fingerprinting-based localization approaches.