project description

Data reduction and image formation for future radio telescopes (DRIFT)

Themes: Signal processing for communication

The future SKA 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--can compressive sampling and advanced algebraic techniques help?
The future SKA 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.

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

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

Expected results

  • Compressive covariance sampling: the objective is to enable non-uniform time-domain sampling at the stations, and thereby reduce the amount of sampled and transported data. This will result in a flexible sampling scheme with a trade-off between data rate (power consumption) and total integration time.
  • Image formation: the objective is to derive new algebraic algorithms that are more efficient than existing algorithms, produce accurate wide-field images, allow flexible implementation of most currently proposed CS-based imaging schemes, and enable massively parallel processing.

For more information, see the project homepage.

Project data

Starting date: May 2014
Closing date: May 2019
Funding: 250 kE; related to group 250 kE
Sponsor: NWO
Partners: ASTRON, IBM

Publication list