用改进的MUSIC算法实现相干多径信号分离

基于特征值分解的MUSIC算法是建立在非相干信号模型基础之上的,对于相干多径信号,MUSIC算法将会失效。与传统的拟补空间协方差矩阵秩亏损的空间平滑去相关法不同,从另一个角度出发,通过特殊的天线阵列模型,重构一个Toeplitz矩阵,使其秩只与信号的波达方向有关,而不受信号相关性的影响,从而达到去相关的目的,并对信号子空间和噪声子空间作出正确的估计。仿真结果验证了该方法的有效性,且较传统的空间平滑方法具有更低的信噪比门限和更小的运算量。

Separation of coherent multi-path signals with improved MUSIC algorithm

The MUSIC algorithm, which is based on the eigenvalue decomposition and has excellent performance and high efficiency, can provide very high resolution and asymptotically unbiased DOA estimation. However, the MUSIC algorithm is constructed on the model of uncorrelated signals and will be ineffect on coherent multi-path signals. Unlike the conventional "spatial smoothing techniques" which attempt to eliminate the rank loss of the spatial covariance matrix, the presented method, from another point of view, reconstructs a special antenna-array-model-based Toeplitz matrix whose rank is only related to the DOA of signals and will not be affected by the coherency between the signals, thus the signal subspace and the noise subspace can be estimated properly. The computer simulation shows that it is effective and has lower SNR threshold and less computational burden than conventional "spatial smoothing techniques".