Compressive Sensing is a novel field in digital signal processing that is concerned with the efficient sampling and reconstruction of compressible signals. In the field of CS compressibility is often defined by a sparse respresentation of a given signal in an appropriate basis (Fourier, Wavelets, etc.). Thus, reconstruction of signals is focused on under-determined equations with sparse signals. This very general idea opens a broad range of applications of which two are focused here:
Fields of Application:
- Application of Compressive Sensing for data detection in wireless digital communications
- Efficient signal aquisition and processing in invasive neural implants
Work tasks (at present):
- Design and study of CS detection concepts such as:
- L1 / L2 Optimization
- Greedy algorithms such as Orthogonal Matching Pursuit (OMP), Orthogonal Least Squares (OLS), CoSaMP
- Belief Propagation (AMP, GAMP)
- Application of CS to communication systems / sensor communication
- Impact of channel coding
- Communication specific sparsity structure
- Activity detection (false alarms, missed detections)
- Cross-Layer PHY/MAC design issues
- CS in neuro sciences
- Appropriate bases for neural signals (action potentials, local field potential)
- Correlation modelling and exploitation in CS
- Hardware efficient CS approaches
- Theoretical studies on performance limits and complexity
Projects