小波盲源分离算法的仿真及FPGA实现  

Simulation and FPGA Implementation Scheme for the Wavelet-based BSS Algorithm

作　　者：杨军[1] 杜兴民[2] 茹乐[2] 

机构地区：[1]空军工程大学工程学院航空电子工程系,陕西西安710038 [2]空军工程大学

出　　处：《微计算机信息》2008年第22期262-264,共3页Control & Automation

基　　金：航空航天部国家空间微波技术重点实验室(编号不公开)

摘　　要：提出了一种基于小波变换的盲源分离方法,在理论分析和仿真结果的基础上,给出了FPGA的实现方案。针对传统盲分离算法对源信号统计特征敏感的弱点,该方法在小波变换域实施盲分离算法。同时,在分离过程中引入了空域相关滤波,使本文方法具有抗噪特征。将该方案应用于多接收天线系统中的多跳频信号的分离应用中,仿真结果证明了该方案的可行性和有效性。This paper proposed a method for wavelet-transform-based blind source seperation, and FPGA implementation scheme was also designed based on the theorial analysis and simulation results. Aiming at the weakness of the sensitivity of source signals statistical distribution in traditional BSS algorithm, this method applied BSS in wavelet tranform domain. Meanwhile, because of introducing spatial-correlation filtering in source seperation process, this method featured anti-noise characteristic. The application in multiple frequency-hopping signals blind seperation in multi-antenas system was also examined in this work as well, and simulation results indicated the availability and efficiency of this method.

关 键 词：小波变换 盲源分离 空域相关滤波 FPGA 

分 类 号：TN76[电子电信—电路与系统]