AIRLINK: Ad-hoc impulse radio, local instantaneous networks

 

Project description (2002)

Impulse radio, or Ultra Wideband (UWB) radio, is a promising new technology for wireless communications. Rather than modulating the information on a carrier (as with all current wireless systems such as AM/FM radio, TV, GSM, UMTS, WLAN, Bluetooth), the data is transmitted using a coded series of very narrow pulses --- each of duration of less than a nanosecond. The corresponding bandwidth is over a Gigahertz, covering the existing radio frequencies but at minute power densities.

Compared to existing solutions for wireless local area networks (WLAN) and personal area networks (PAN), the predicted advantages of UWB data communication are enormous, especially for short-range applications: devices are expected to be small and cheap, the technology allows for very high data rates and unprecedented user concentrations (there are bold claims of hundreds of devices each at 50 Mbps in the same 30 m area. it is very power efficient (transmission at $\mu$W rather than mW), it is inherently secure because of its low power and long spreading codes, and it works very well in an indoor environment with lots of multipath reflections. For this reason, it has the potential to become a very important technology for beyond-3G communication systems.

UWB technology has other significant, non-communication applications as well. The ultra-short pulses are similar to those used in radar, and can be used for a precise ranging or localization of receivers at cm accuracy. This is an enabling technology for context-aware and position-aware devices, with applications such as tagging luggage at airports or keeping track of high-value objects in offices or warehouses. The propagation of low frequency components through objects allow for motion detection through walls, and it is, e.g., possible to create a virtual ``security dome'' that gives an alarm when someone enters in a certain radius around a transceiver, without being triggered by people already in the dome. There are obvious automotive applications as well, e.g., vehicle collision avoidance. With more sophisticated multi-antenna receivers, it would be possible to do ``radar vision'', with applications such as looking through walls or debris in disaster areas, biomedical imaging similar to ultrasound echoscopy, and navigation aids for the blind.

In the past years, publications on this topic were mostly limited to a handful of research groups and companies, but the US military has been using the technology for decades. A new impulse came with the advent of high-speed digital electronics, enabling control over the pulse shape which is necessary to constrain the power spectrum and avoid spectral lines. Since two years, interest in UWB communication is growing, also in Europe, and it is now considered in CEPT and ETSI study groups and one of the IEEE 802 working groups, conferences are being organized, and an explosion of interest is starting to come up. The May 2002 issue of the Scientific American carries a feature article on the technology and its possibilities.

Problem statement

This research proposal addresses the question: How can a multi-user ad-hoc network be built using UWB technology. Our combined expertise allows us to work on all aspects: new hardware architectures and implementations, the physical layer with signal processing, modulation, coding and synchronization, and the link layer at which the multiple access scheme and the network architecture is defined. System and application aspects are not studied in this proposal, but are adopted from the CACTUS program.

The state-of-the-art of UWB communication can be considered immature, with demonstrated data rates only in the order of kbit/s, much below the theoretical limits. Technological and scientific challenges which we see are

• Hardware: Pulse timing and receiver synchronization (gating) have to be extremely accurate, or else the promised high data rates are not achievable. In the same vein, it may be desired to make the pulse repetition rate much higher than usual. Pulse shapes and repetition rates have to be designed so that the transmitted spectrum is as flat as possible, in compliance to proposed regulations. Other major challenges are the fast initial acquisition in the presence of many users, maintaining receiver lock, the development of low-power transceiver architectures, the design of broadband circuits and ultra-fast correlators in integrated-circuit (IC) technologies that are compatible with VLSI technology, partitoning into hardware-software, IC bonding and packaging techniques and the incorporation of adaptivity, e.g., in transmit power, data rates or receiver functionality. A further challenge is to build array antennas that can be mass produced at low cost, take little spac e, and maintain the essential properties of the signal.
• Channel modeling and measurements: Although some companies have carried out measurements, this has not been done at higher ($> 3$ GHz) frequencies, and no statistical channel model is available, although this is essential for system design. Channel measurements have to have a very high time-resolution, which requires sophisticated equipment. An unusual complication is that pulses transmitted through objects become distorted, and also the transceiver antennas may act as a differentiator.
• Signal processing: Advanced receivers should exploit the energy present in the multipath reflections. Traditionally, this requires on-line channel estimation, but this is not (yet) feasible at GHz rates. Coding and modulation schemes and algorithms have to be designed to shape the transmitted spectrum, facilitate acquisition, and provide efficient multi-user interference suppression at high data rates (mitigation of the cocktail party effect).
• Network: As far as we know no ad-hoc networks running over UWB have ev er been described. A new protocol that takes the characteristics of the technology into account is needed. This includes the augmented possibilities of ranging and location tracking, not found in existing protocols: this has great impact on the design of multi-hop packet forwarding (e.g, to minimize transmission power rather than delay).

Start:	Sept. 2002
End:	Sept. 2008
Sponsor:	BSIK Freeband Impulse

Partners:

• TU Delft: CAS, ELCA, WMC, IRCTR
• Industrial partner: TNO-FEL

Links

• Final report
• List of publications at CAS on UWB
• People involved in the project

Contact address

Mail:	prof.dr.ir. Alle-Jan van der Veen
	Delft University of Technology
	Fac. EWI/Electrical Engineering
	Mekelweg 4
	2628 CD Delft
	The Netherlands
Phone:	(+31 15) 2786240
Fax:	(+31 15) 2786190
E-mail:	allejan@cas.et.tudelft.nl