摘要
为了进一步改善阵列信号处理中盲源分离算法的分离性能,提出了一种新的基于阵列结构的盲分离算法。该算法的基本思想是利用已有的盲源分离算法(EASI和FastICA算法)估计混合矩阵,根据估计出来的混合矩阵和均匀线阵的特点来重构混合矩阵,对分离矩阵进行较正,达到改善算法分离性能的目的。仿真结果表明,该文提出的EASI-1算法的平均干信比比EASI算法低7.5 dB,FastICA-1算法的平均干信比比FastICA算法低4.3 dB。
To improve the separation performance of the blind source separation algorithm in array signal processing, a novel separation algorithm based on array constructure is presented here. The basic idea of the algorithm is that the mixing matrix is estimated by using the available blind source separation algorithm （i. e. EASI or FastICA）. The mixing matrix is reconstructed according to the estimated mixing matrix and the characteristic of the unitary linear array. This reconstructed mixing matrix can be used to revise the separation matrix and improve the performance of the algorithm. Simulation results show that the average interference signal ratio of EASI-1 algorithm presented here is lower 7.5 dB than that of the EASI algorithm, and that of the FastICA-1 is lower 4.3 dB than that of the FastICA algorithm.
出处
《南京理工大学学报》
EI
CAS
CSCD
北大核心
2009年第2期168-171,共4页
Journal of Nanjing University of Science and Technology
基金
国家自然科学基金(60702060)
高等学校学科创新引智计划(B08038)
关键词
盲信号处理
盲源分离
均匀线阵
blind signal processing
blind source separation
unitary linear array
