基于相对梯度的鲁棒的盲源分离算法

Journal of Systems Engineering and Electronics ›› 2010, Vol. 32 ›› Issue (2): 226-228.

基于相对梯度的鲁棒的盲源分离算法

付卫红1, 刘乃安1, 杨小牛2, 曾兴雯1

（1. 西安电子科技大学综合业务网理论及关键技术国家重点实验室，陕西西安 710071；
2. 江南电子通信研究所通信系统信息控制技术国家重点实验室，浙江嘉兴 314033）

出版日期:2010-02-03 发布日期:2010-01-03

Robust blind source separation algorithm based on relative gradient

FU Weihong1, LIU Naian1, YANG Xiaoniu2, ZENG Xingwen1

(1. State Key Lab. of Integrated Services Networks, Xidian Univ., Xi’an 710071, China;
2. National Lab. of Information Control Technology for Communication Systems，
Jiangnan Inst. of Electronic Communications， Jiaxing 314033, China)

Online:2010-02-03 Published:2010-01-03

摘要/Abstract

摘要：

针对有噪环境中通信信号的盲分离问题，说明了在有噪环境中，对观察信号进行白化处理后得到的混合矩阵为正交矩阵，并在此基础上给出了正交分离矩阵的迭代公式。另外，为了消除噪声影响，改善算法的分离性能，根据噪声的概率密度模型，提出了一种改进方法。该方法对分离信号进行迭代，从而得出鲁棒的盲源分离算法。仿真结果表明，提出的分离算法相对著名的EASI算法，其性能有很大改善。在信噪比为20 dB时，本算法获得的干信比比EASI算法低10 dB。

Abstract:

Aiming at the problem of the communication signal blind separation in the noisy circumstance, the mixing matrix is transformed into an orthogonal matrix after prewhitening of the observed signals. Based on the result, an iterative equation of the orthogonal separating matrix is given. Finally, in order to eliminate the effect of the noise and improve the separating performance of the algorithm, a novel method is proposed according to the model of the probability density function of the noise. Thus the robust blind source separation algorithm is obtained. The simulation result shows that the performance of the proposed algorithm is greatly improved relative to the famous equivuiriant adaptive separation via independence (EASI) algorithm. When the signaltonoise ratio is 20 dB, the interferencetosignal ratio of the proposed algorithm is 10 dB lower than that of the EASI algorithm.

付卫红1, 刘乃安1, 杨小牛2, 曾兴雯1. 基于相对梯度的鲁棒的盲源分离算法[J]. Journal of Systems Engineering and Electronics, 2010, 32(2): 226-228.

FU Weihong1, LIU Naian1, YANG Xiaoniu2, ZENG Xingwen1. Robust blind source separation algorithm based on relative gradient[J]. Journal of Systems Engineering and Electronics, 2010, 32(2): 226-228.