CAJ | 학술논문

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，本文提出了一种基于DOA估计的阵列信号分离方法。该方法的基本思想是利用MUSIC算法实现对阵列信号DOA的预先估计，构建一预估计方向加载矩阵W，使加载后的混合信号在该预估计DOA上的信号能量最大化，最后利用传统的盲源分离算法分离接收到的混合信号，得到原始信号波形。仿真结果表明，该方法分离效果优于直接的盲分离算法，原始信号与分离信号相关系数接近于1，但同时牺牲较大运算量。
위료진일보개선진렬신호처리중맹원분리산법적분리성능，본문제출료일충기우DOA고계적진렬신호분리방법。해방법적기본사상시이용MUSIC산법실현대진렬신호DOA적예선고계，구건일예고계방향가재구진W，사가재후적혼합신호재해예고계DOA상적신호능량최대화，최후이용전통적맹원분리산법분리접수도적혼합신호，득도원시신호파형。방진결과표명，해방법분리효과우우직접적맹분리산법，원시신호여분리신호상관계수접근우1，단동시희생교대운산량。
To improve the separation performance of the blind separation algorithm in array signal processing,a new re-search on the signals separation method of the uniform linear array is proposed in this paper.The basic idea of the algo-rithm is the original signal DOA being estimated by the algorithm of MUSIC and then a loading matrix W based on the estimated DOA being created to get the large energy of the mixing signals in the estimated DOA. At last, the mixing signals are separated by the traditional blind source separation algorithms and the original signals are got. Simula-tions show that the performance of proposed method is better than these of the traditional blind source separation algo-rithms, the related coefficient between the natural signals and the separated signals approaches to 1, but the computa-tional complexity is increased greatly.