基于洛伦兹范数的MWC快速重构算法  

Fast recovery algorithm for MWC based on Lorentzian norm

作　　者：彭军伟[1,2] 韩志韧 李杨[2] 孙景芳[2] Peng Junwei;Han Zhiren;Li Yang;Sun Jingfang(College of Information and Communication Engineering, Harbin Engineering University, Harbin 150001, China;Wuhan Maritime Communication Research Institute, Wuhan 430205, China)

机构地区：[1]哈尔滨工程大学信息与通信工程学院,黑龙江哈尔滨150001 [2]武汉船舶通信研究所,湖北武汉430205

出　　处：《华中科技大学学报（自然科学版）》2018年第5期6-11,16,共7页Journal of Huazhong University of Science and Technology(Natural Science Edition)

基　　金：船舶工业国防科技预研基金资助项目(11J3.4.2)

摘　　要：针对现有调制宽带转换器（MWC）压缩采样信号重构算法对联合稀疏结构的限制及其受滤波器非理想因素影响的问题,提出了一种基于洛伦兹范数的MWC快速重构算法.该算法采用洛伦兹范数来拟合误差项,设计了冲击噪声背景下的稀疏优化目标函数,可有效削弱异值点对重构结果的影响;同时进一步将零范数优化问题转化为矩阵平滑零范数的函数极值求解问题,并利用固定步长共轭梯度算法进行并行重构,以提高算法收敛速度和运行效率.仿真结果表明：与现有重构算法相比,在受滤波器过渡带等非理想因素影响下,所提算法可以提高稀疏信号的重构成功率,同时在重构速度方面具有较好的性能.A fast recovery algorithm of modulated wideband converter（MWC） based on Lorentzian norm was proposed to surmount the restriction of the existing MWC signal recovery algorithm on joint sparse structure,and reduce the influence of some non ideal factors in the filter.An objective function for sparse optimization in impulsive noise environment was designed,weakening the influence of abnormal data points on the result of recovery.The designed objective function used a Lorentzian norm to fit the error term.To improve the convergence speed and computational efficiency of recovery algorithm,the recovery process was optimized via the solution of function extremum of matrix smoothed L0-norm.Finally,a parallel recovery was implemented in a unified parametric framework using conjugate gradient algorithm with the fixed-step.Simulation results show that,compared with traditional recovery algorithm,the proposed algorithm can not only improve the performance of sparse signal recovery,but also has obvious advantages in the recovery speed.

关 键 词：调制宽带转换器 洛伦兹范数 快速重构算法 任意稀疏结构 冲击噪声 压缩感知 

分 类 号：TN911[电子电信—通信与信息系统]