【摘要】 稀疏度自适应的匹配追踪算法(SAMP)是基于压缩感知理论的信号重建经典算法。针对稀疏度未知的信号重建,提出步长自适应的前向后向匹配追踪(AFBMP)算法,AFBMP算法在稀疏度自适应匹配追踪算法的框架下,前向搜索过程中采用对数型自适应变化的步长选择匹配原子,然后通过后向策略修正前向阶段造成的错误,删除支撑集中的部分错误原子,最终实现信号的精确逼近。实验表明AFBMP算法比SAMP算法能够更加高效地重建稀疏度未知的信号。更多还原

【Abstract】 Sparsity adaptive matching pursuit algorithm( SAMP) is a classical algorithm of signal reconstruction based on compressed sensing theory. Aiming at reconstructing signals with unknown sparsity,we presented an adaptive stepsize forward-backward matching pursuit algorithm( AFBMP). In the framework of SAMP,AFBMP selects matching atoms in its forward search process by using logarithmic adaptive changing steps,and then amends the mistakes caused in forward stage by backward strategy to delete part of the false atoms in support set,and finally realises accurate signal approaching. Experiments show that the AFBMP algorithm can reconstruct the signals with unknown sparsity more efficiently than SAMP algorithm.更多还原