SOSP: a stepwise optimal sparsity pursuit algorithm for practical compressed sensing

摘要: Compressed Sensing (CS), as a promising paradigm for acquiring signals, is playing an increasing important role in many real-world applications. One of the major components CS sparse signal recovery which greedy algorithm well-known its speed and performance. Unfortunately, classic algorithms, such OMP CoSaMP, real sparsity key prior information, but it blind. In another words, true not available practical Due to this disadvantage, performance these algorithms are significantly reduced. order avoid too much dependence on sparsity, paper proposed efficient reconstruction Sensing, termed stepwise optimal pursuit (SOSP). Differs from existing unique feature SOSP that assumption needed instead sparsity. Hence, limitations application can be tackled. Based arbitrary initial satisfying certain conditions, employs two variable step sizes hunt by comparing final residues. Since preserves ideas original innovates information thus applicable any effective requiring known Extensive experiments conducted demonstrate offers superior terms discarding