稀疏度自适应分段正交匹配追踪算法改进

Improved Algorithm for Sparsity Adaptive Stagewise Orthogonal Matching Pursuit

在基于压缩感知的信号重构问题中,有一类常见情况——未知信号稀疏度.针对此类情况,提出稀疏度自适应分段正交匹配追踪(Sparsity Adaptive Stagewise Orthogonal Matching Pursuit,SAStOMP)算法,该算法将自适应思想、变步长迭代思想与分段正交思想相结合,在未知信号稀疏度的情况下,自适应地选择支撑集原子的个数,最终实现信号的精确重构.仿真结果表明,针对长度为256位的原始信号,该算法重建效果优于正交匹配追踪算法、正则化正交匹配追踪算法和分段正交匹配追踪算法等.

Aiming at the reconstruction of unknown signal sparsity in compressed sensing,this paper proposes a new compression sensing signal reconstruction algorithm of SAStOMP (Sparsity Adaptive Stagewise Orthogonal Matching Pursuit).The algorithm combines the ideas of self-adaptation,variable step size iteration and piecewise orthogonal design,in the case of unknown signal sparsity,selects adaptively the number of atoms of the support set ,finally realizes the accurate reconstruction of signals.Simulation results show that the proposed algorithm is superior to OMP,ROMP and StOMP for the original signals of 256 digits.