With fast development of science and technology, people gradually need more and more information, causing significant pressure on the sampling, transmission and storage of signals. Compressive sensing (CS) could fetch useful information efficiently carried in the signal, and consequently proposes an effective solution to relieve this pressure. As its sorts of advantages and wide application scenarios, compressive sensing, with highly bright future, has been involved in practical application and some relative technologies have been invented.In this paper, we firstly conclude the frame and main application field of compressive sensing and discuss the present problems in research. Based on distributed source coding (DSC) theory, we introduce distributed compressive sensing (DCS) theory, and analyze the sparsity presentation of signal, design of measurement matrix and reconstruction algorithms, respectively. For signal reconstruction, we gradually analyze advantage and disadvantage of three kinds of algorithms, which are (?)optimization,(?)optimization and matching pursuit (MP), including orthogonal matching pursuit (OMP) and stagewise orthogonal matching pursuit (StOMP). For signal reconstruction in DCS, we utilize the concept of side information in DSC and the first joint sparsity model (JSM), fully exploit both inter-and intra-signal correlation. Combined with orthogonal matching pursuit, we propose side information based orthogonal matching pursuit (SiOMP), and demonstrate the procedures of the algorithm. This algorithm treats one of the sources as side information, encodes and decodes the source utterly by compressive sensing. Correspondingly, utilizing the correlation with the side information, other sources could be successfully decoded with high probability and would reach a high decoding speed. At last, under different length and sparsity of signals, we give performance simulation results, and prove the stability and priority of the new algorithm. |