相干信号的空间谱估计算法研究

摘要

Aiming at the MUSIC algorithm's limitation on spatial spectrum estimation of coherent signals and its low estimation accuracy, this paper presented an advanced algorithm for spatial spectrum estimation of coherent signals——fixed global smoothing algorithm, which breaks the limitation of the placement of arrays through replacing thernsmoothing subarrays and makes the usefull number of subarrays more than that of forward/backward spatial smoothing. It uses all the arrays as a slidable subarray and gets a new covariance matrix containing spatial spectrum information, then decomposes its eigenvalue and uses MUSIC algorithm to estimate the DOA. The simulation results prove that the fixed global smoothing algorithm can estimate the DOA of coherent signals very well, and it has much higher estimation accuracy than the forward/backward smoothing algorithm when the SNR is quite low.%研究相干信号谱估计精度问题,由于空间信号噪声影响,识别精度低.为此提出了一种改进的解相干谱估计算法——全局滑动修正算法.可通过对平滑子阵划分的修正,突破了阵列布局限制,有效地使可利用的子阵数目大于双向空间平滑中的子阵个数,采用全部的阵元当做一个滑动的予阵,得到一个新的包含空间谱信息的协方差矩阵,再对其进行特征值分解,然后按照MUSIC算法进行谱估计.仿真结果证明,全局滑动修正算法能很好地对相干信号进行解相干DOA估计,并且在信噪比较低时,比前后向空间平滑算法具有更高的估计精度.