Research On Signal Reconstruction Algorithm Of Compressive Sense

Compressive sense (CS) is new sampling theory with a rapid development inrecent years. CS compresses the signal in the same time while it is sampling, thus,makes the compress progress and the sampling process into one. CS breaks theshackles of Nyquist law and saves a lot of storage, transmission, and computingresources.Firstly, this paper presents an overview of the theoretical framework of CS. Itincludes three parts: the sparse representation of signal, the measurement matrixdesign of signal, and the reconstruction of signal. And these three parts have beenstudied in this paper, of which the reconstruction of signal is the most important partand what the paper focus on. An improved design of measurement matrix is putforward in the paper. In the part of signal reconstruction, the principle of somecommon reconstruction algorithm has been presented, and their time complexity andreconstruction accuracy have also been compared and demonstrated. Secondly, thispaper also focus on the BP and BPDN algorithm, and through the improvement to thealgorithm, BP algorithm could be used in the signal which contains Gaussian pulsenoise, which extends its scope of application. Finally, this paper compares theimproved algorithm with the original algorithm through simulation and verification,the results shows that the proposed algorithm could effectively improve thereconstruction accuracy of the signal.