一种色噪声背景下的信源方位估计快速算法

针对实际环境中相干信源普遍存在的情况,提出一种基于对称均匀线阵的波达方向(DOA)分步估计方法。该算法在未知噪声协方差矩阵为复对称Toeplitz(色噪声)结构的情况下,利用空间差分方法和相干信源Toeplitz矩阵重构方法相结合,来处理同时存在相干(或相关)和独立信源的情况。首先利用常规谱估计算法估计独立信源;然后用差分的方法将其排除掉,同时可以排除色噪声信息;然后用Toeplitz重构的方法将剩下的相干信源恢复为满秩,进而可以利用传播算子的方法进行DOA估计。与传统的去噪、解相干算法相比,该算法在提高阵列信源过载能力的同时,可明显减小算法的运算量。计算机仿真结果证明了新算法的有效性和正确性。

A Fast Algorithm of Azimuth Estimation for Sources under Stationary Correlated Noise

Aiming at the ubiquity of coherent sources in real circumstances,an ordinal algorithm to direction of arrival(DOA) estimation for symmetric uniform linear array is presented.under the condition of colored noise field.The algorithm combined with spatial differencing and Toeplitz reconstructing is used to cope with the situation where both uncorrelated sources and coherent sources exist,The uncorrelated sources is first estimated by using MUSIC,simultaneously the noise covariance matrix and uncorrelated sources are eliminated,Then a Toeplitz matrix is constructed for the remained coherent sources to renew full rank.After preprocessing,propagator method(PM) is employed to find the DOAs without any eigen decomposition.By using this approach,the number of sources resolved exceeds that of array elements and at the same time the computational complexity is obviously reduced.Simulation results show that the proposed method is effective and correct.