A New Sparse Signal Reconstruction Algorithm for Compressed Sensing Based on l_p Norm

摘要

The theory of compressed sensing allows the reconstruction of sparse signals with much fewer measurements than traditionally believed. In this paper, a new reconstruction algorithm, which is based on the optimization of l_p-norm, is proposed. The new algorithm combines the Penalty Function (PF) method and the Sequential Quadratic Programming (SQP) method. It not only takes advantage of these two methods' merits but also overcomes the disadvantages of them. Also, signal reconstruction based on l_p-norm in compressed sensing reduces the redundancy among data to some extent and effectively decreases the number of measurements required to reconstruct the original signal. Simulation results show that the new algorithm offers improved signal reconstruction performance.