针对载频未知目标方位估计方法的二维搜索运算量大的问题,提出一种基于分离变量方法的目标方位估计新方法,该方法通过泰勒展开将阵列的导向矢量近似为仅含有方位因子的矩阵和仅含有频率因子的向量的积,并通过矩阵变换将二维MUSIC方位估计的空间谱函数中的频率因子约去,在此基础上进行目标的方位估计。整个计算过程只需进行一维搜索即可获得多个目标的方位,有效地减少了原二维搜索算法的计算量。仿真结果表明:该方法高信噪比条件下均方根误差约为0.3°,可以满足工程需要。
A new method for estimating directions-of-arrival(DoA) of multiple spatial narrowband signals whose carrier frequency unknown is presented to lighten computational complexity of classical 2 dimension searching method.The steering vector of the array can be modeled as the product of a matrix containing the azimuth factor and a vector containing the frequency factor.Factor of frequency is removed from the spatial function of MUSIC method through matrix transfer.DoA of targets is got through one dimension searching using transferred MUSIC spatial function.Since the method provides multiple targets' azimuth estimation via one dimension searching,the complexity computation of method is less compare with that of classical MUSIC method.Simulation results show that RMSE of RMUSIC is about 0.3° in high SNR.This method is feasible in engineering practice.