一种适合任意阵列流型的相干信号波达方向估计算法

A Method for Estimating the Direction of Arrival of Coherent Signals on Arbitrary Antenna Array in Multi-path Scenarios

针对多径环境中相干信号波达方向（DOA）估计问题，提出了一种适合任意阵列流型的相干信号波达方向估计算法（DOA—CSAA）．先通过联合使用时域和空域相关函数估计出每个用户组的广义导向矢量，再估计出每个用户组内多条路径上的波达方向．由于DOA—CSAA算法所利用的信号的空域相关信息和时域相关信息均为信号的二阶统计量，因而运算量远小于基于四阶统计量的算法．此外，DOA—CSAA算法由于未使用子空间平滑算法，所以不受阵列流型限制，且阵列孔径没有降低，因此当使用M个阵元的任意阵列流型时，可估计出的DOA个数最多可达M（M—1）个．仿真实验表明，与经典子空间平滑算法相比，DOA—CSAA算法在均方根误差为10^-4～10^-3时可获得20dB左右的信噪比增益．

In order to estimate direction of arrival（DOA） of coherent signals in multi-path propagation, a new estimation algorithm DOA-CSAA is proposed. The generalized steering vector corresponding to each user is estimated by jointly adopting the time-domain and space-domain correlations. Then the direction-of-arrival of every multi-path is estimated within each user＇s group. The proposed method, which is capable of identifying more DOAs than the methods that are based on fourth-order statistics, uses second-order statistics only, hence its computational complexity is much lower. Since the forward/backward spatial smoothing technique is not adopted in the proposed method, arbitrary geometry arrays are applicable. The method can estimate at most M（M-1） impinging DOAs using an arbitrary geometry array with M antennae. Simulation results and comparison with the forward/backward spatial smoothing techniques under the same simulation conditions show that the DOA-CSAA can obtain nearly 20 dB signal-to-noise ratio gain when the root mean square error of approximation is between 10^-4 and 10^-3.