Description
Conventional radar processing is based on the established paradigm of “Sample then Compress”. So the signal is basically sampled at the Nyquist rate and then decimated (compressed) to represent the classical “point-like range/velocity map” of the radar systems. In application where is important to reduce velocity ambiguities fast frequency modulated waveforms are used with a dramatically increasing of the ADCs’ data rate and computational load for post processing algorithms. Compressive sensing is the most promising solution to overcome these problems without losing information. It is basically based on the paradigm of ”compress while collecting data” where the overall complexity of the hardware components can be reduced by undersampling the signal with certain prerequisites and not at the cost of the information to be represented. The use of high-performance delta-sigma Analog-to-Digital convertors also increases the data-rate of the processing. Compressive sensing can be used in such applications to reduce processing complexity without losing essential radar information. The scope of the master thesis is to develop suitable algorithms for radar compressive sensing for the automotive scenarios. The work shall be an exhaustive mixture of theory, simulation and practical activities.
References
Introduction to compressed sensing: http://statweb.stanford.edu/~markad/publications/ddek-chapter1-2011.pdf
An introduction to compressive sampling: http://dsp.rice.edu/sites/dsp.rice.edu/files/cs/CSintro.pdf
Sparsity and Compressed Sensing in Radar Imaging: http://ens.ewi.tudelft.nl/Education/courses/et4248/cs_potter.pdf
Compressed sensing in automotive applications: http://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=6869182
Your responsibilities
- Understanding of radar waveform and processing techniques
- Understanding of the principle of CS technique
- Simulate different CS techniques and analyze the results/trade offs
- Perform experiment with real data and in real scenario
We are looking for
- Currently studying towards your Master in the Electrical engineering department.
- Good knowledge in information theory and signal processing.
- Good knowledge of Matlab.
- Interest in radar sensors
- Knowledge in delta-sigma Analog-to-Digital convertors.
- Pro-active and hands-on
- At least 9 months availability
- Language: Good command of English
Contact
Dr. Francesco Laghezza (f.laghezza@tue.nl)
Paul van Zeijl (paul.van.zeijl@omniradar.com)