Title | Compressive Covariance Sensing: Structure-Based Compressive Sensing Beyond Sparsity |
Publication Type | Journal Article |
Year of Publication | 2016 |
Authors | Romero, D, Ariananda, DD, Tian, Z, Leus, G |
Journal | IEEE Signal Processing Magazine |
Volume | 33 |
Issue | 1 |
Pages | 78--93 |
Date Published | 01/2016 |
Keywords | compass, compressed sensing |
Abstract | Compressed sensing deals with the reconstruction of signalsfrom sub-Nyquist samples by exploiting the sparsity of theirprojections onto known subspaces. In contrast, the presentarticle is concerned with the reconstruction of second-orderstatistics, such as covariance and power spectrum, even inthe absence of sparsity priors. The framework described hereleverages the statistical structure of random processes toenable signal compression and offers an alternative perspectiveat sparsity-agnostic inference. Capitalizing on parsimoniousrepresentations, we illustrate how compression and reconstructiontasks can be addressed in popular applications suchas power spectrum estimation, incoherent imaging, directionof arrival estimation, frequency estimation, and widebandspectrum sensing. |
PDF: