Compressed Sensing - Overview and Universal Algorithms
By Dr Dror Baron – NC State University
Traditional signal acquisition techniques sample band-limited analog signals above the Nyquist rate. Compressed sensing (CS) is based on the revelation that a sparse signal can be reconstructed from a small number of linear projections of the signal. Therefore, CS-based techniques can sample sparse signals at sub-Nyquist rates. Potential applications include broadband analog-to-digital conversion and new kinds of imaging devices.
The focus of our recent research has been on CS algorithms that can reconstruct signals despite not knowing their statistical properties. We will describe algorithms that are universal in the sense that they can adapt to unknown statistics. We will conclude the talk by briefly surveying recent work on a fast parallel algorithm for data compression, which can be used in applications involving high data rates including lossless compression of radio astronomy data.
