Robust recovery for radar applications by noise mitigated compressed sensing

Compressed sensing (CS) is an exciting, rapidly growing field that has attracted considerable attention in signal processing, statistics, and computer science. Especially, for radar applications it plays a major role for facilitating the signal processing and information recovery, since the information of interest in radar applications has usually its compressible or even sparse structure. In practice, however, a robust recovery in CS is a difficult task, which is strongly depending on the information of sparsity level and noise energy, something that we cannot know in advance and must be estimated or additionally provided. It would become worse if the related CS problem is ill-conditioned. In this paper, we will present our proposed noise mitigated compressed sensing (NMCS), which gives a promising framework for a robust CS recovery without specifying the signal sparsity or noise level.