基于观测矩阵优化的自适应压缩感知算法

发布时间：2018-01-20 07:53

本文关键词： 自适应压缩感知观测矩阵优化观测信噪比特征分解 Gram矩阵　出处：《计算机应用》2017年12期 　论文类型：期刊论文

【摘要】：为提高传统压缩感知(CS)恢复算法的抗噪性能,结合观测矩阵优化和自适应观测的思想,提出一种自适应压缩感知(ACS)算法。该算法将观测能量全部分配在由传统CS恢复算法估计的支撑位置,由于估计支撑集中包含支撑位置,这样可有效提高观测信噪比(SNR);再从优化观测矩阵的角度推导出最优的新观测向量,即其非零部分设计为Gram矩阵的特征向量。仿真结果表明,随着观测数增大,Gram矩阵非对角元素的能量增速小于传统CS算法,并且分别在观测次数、稀疏度和SNR相同的条件下,所提算法的重构归一化均方误差低于传统CS恢复算法10 d B以上,低于典型的贝叶斯方法 5 d B以上。分析表明,所提自适应观测机制可有效提高传统CS恢复算法的能量利用效率和抗噪性能。
[Abstract]:In order to improve the anti-noise performance of the traditional compression sensing (CSM) restoration algorithm, the idea of optimization of observation matrix and adaptive observation is combined. An adaptive compression sensing (ACS) algorithm is proposed, in which the observed energy is allocated to the support position estimated by the traditional CS recovery algorithm, because the estimated support set contains the support position. In this way, the SNR can be improved effectively. From the point of view of optimizing observation matrix, the optimal new observation vector is derived, that is, the non-zero part of the vector is designed as the eigenvector of Gram matrix. The simulation results show that with the increase of the number of observations. The energy growth rate of non-diagonal elements of Gram matrix is smaller than that of traditional CS algorithm, and the observed times, sparsity and SNR are the same. The normalized mean square error of the proposed algorithm is lower than that of the traditional CS recovery algorithm above 10 dB and the typical Bayesian method above 5 dB. The proposed adaptive observation mechanism can effectively improve the energy utilization efficiency and anti-noise performance of the traditional CS recovery algorithm.
【作者单位】：重庆邮电大学通信与信息工程学院;移动通信技术重庆市重点实验室(重庆邮电大学);
【分类号】：TN911.7
【正文快照】： 0引言压缩感知(Compressed Sensing,CS)[1-3]技术因其高效的信息采样机制,已被广泛应用到信道估计、神经网络、雷达信号处理、核磁共振成像等领域。传统压缩感知技术采用非自适应的观测过程,即通过预先设计观测矩阵并对信号进行一次性观测,观测向量的能量均匀分配,这样做使得

【相似文献】