基于压缩感知的自适应匹配追踪算法优化  

Modified adaptive matching pursuit algorithm based on compressive sensing

作　　者：吕伟杰[1] 陈霞[1] 刘红珍[1] 

机构地区：[1]天津大学电气与自动化工程学院,天津300072

出　　处：《系统工程与电子技术》2015年第5期1201-1205,共5页Systems Engineering and Electronics

摘　　要：针对基于压缩感知的稀疏自适应匹配追踪(sparsity adaptive matching pursuit,SAMP)算法运行效率低的问题,给出了一种优化的自适应匹配追踪(modified adaptive matching pursuit,MAMP)算法。该算法在支撑集选择过程中对稀疏度进行了初步估计,并优化了迭代停止的条件。实验表明,该算法相比于SAMP有更快的收敛速度,并且实现更优的重建效果。To improve the operating efficiency of sparsity adaptive matching pursuit （SAMP）, a modified a- daptive matching pursuit algorithm （MAMP） is presented. The new algorithm estimates the value of sparsity preliminarily in the process of the choice of the support list, meanwhile, it corrects the condition when the circu-lation stops. The experiments demonstrate that the new algorithm saves much more operating time than SAMP, and improves the performance of the recovery.

关 键 词：压缩感知 自适应 匹配追踪 

分 类 号：TN911.72[电子电信—通信与信息系统]