Sparse signal processing is a rapidly growing field that has attracted considerable attention in the scientific community in the last years, especially in the field of compressed sensing.
Compressed Sensing (CS) is a signal processing technique for efficiently acquiring and reconstructing sparse signals. A signal is considered sparse if most of its information is contained within a few non-zero samples. Consequently, a signal reconstruction algorithm has to find a sparse vector that best represents the measured signal. Many algorithms to solve this problem are based on l1-norm optimization. However, for a real-time implementation exactly solving an l1-norm optimization problem is typically too complex to be implemented. Alternatively, approximate algorithms can be used instead. Such algorithms typically do not only have less computational complexity, but often also require less memory access operations and less precision for number representation and arithmetic operations.
The main goal of this Ph.D. project is to design and implement efficient novel reconstruction algorithms allowing for a reasonable performance/complexity trade-off. Potential candidates are iterative or approximate algorithms that often achieve low complexity. We will particularly focus on compressed sensing and sparsity-aware algorithms for system identification, parameter estimation, beamforming or direction finding. Our aim is to develop efficient algorithm designs and architectures for selected use-cases by considering the implementation in digital hardware during the algorithm development phase.