一种改进的稀疏自适应压缩感知重构算法

Improved Sparsity Adaptive Reconstruction Algorithm for Compressed Sensing

为了优化贪婪匹配追踪算法的性能,文中基于稀疏自适应两阶段回溯型贪婪算法-前后追踪算法,提出了一种改进的线性变步长前后追踪算法。该算法结合稀疏自适应追踪算法的分阶段、变步长的思想,将迭代过程分为两个阶段,采用线性变步长进行迭代,大步长较少运行时间,小步长提升重构精度,从而减少了运行开销的同时,提升了算法的重构精度,通过仿真实验对其进行了验证,线性变步长前后追踪算法能够明显减少算法的运行时间,且提升了重构精度。

For the performance optimization of the greedy matching pursuit algorithm, the linear forward-backward pursuit algorithm （LvsFBP） is proposed based on FBP, a two-stage sparsity adaptive variable step greedy algorithm. LvsFBP adjusts the iteration step size adaptively which reduces operational costs while enhancing algorithm reconstruction accuracy. The simulation results show that LvsFBP offers the better performance of signal reconstruction.