摘要
针对无线通信网络近些年出现的大数据量信号的情况,在对信号采样的同时进行适当压缩,利用合适的重构算法实现了用少量的采样值或观测值对信号进行重构。在原有基于OMP(正交匹配追踪)的压缩感知重构算法的基础上引入AS(交替步长),提出一种新的改进算法——GP-OMP(梯度正交匹配追踪)算法。实验结果表明,该改进算法利用前一感知时刻获得的频谱信息,不仅能有效地降低重构算法的计算量,还能有效地减少重构耗时,并得到与原有方法基本一致的重构效果。
In view of the emergence of mega data signals in wireless communication networks in recent years,this paper proper-ly compresses these signals while performing signal sampling and realizes their reconstruction with small amount sampling val-ue or observed value by using an appropriate reconstruction algorithm.On the basis of the original Orthogonal Matching Pur-suit (OMP)-based compressed sensing reconstruction algorithm,the Alternating Step-size (AS)is introduced and an improved algorithm,i.e.GP-OMP algorithm proposed.Experimental results show that the spectrum information this algorithm ob-tained by using the previous sensing moment can not only effectively reduce the amount of computations by the reconstruction algorithm,but also the reconstruction time,with almostthe same results as those by using the original method.
出处
《光通信研究》
北大核心
2016年第1期74-78,共5页
Study on Optical Communications
关键词
重构算法
压缩感知
交替步长
频谱感知
reconstruction algorithm
compressed sensing
alternately step
spectrum sensing
