基于奇异值分解的方向估计改进方法

摘要

The modified singular value decomposition method based on signal phase matching (MSVDSPM) is presented to make the root mean square error of the direction of arrival (DOA) estimation of singular value decomposition based on signal phase matching (SVDSPM) close to the Cramer-Rao bound at high signal-to-noise ratio. Firstly, the sensor outputs are transformed to the frequency domain. Then the reciprocal of the square summation of the dis tance between the sensor output spectra and their mean value at the center frequency bin is tak en as the DOA estimator. The simulation results show that the MSVDSPM has a better perfor mance in DOA estimation than that of SVDSPM. MSVDSPM is a beamforming method pre serving the sharp peak of the SVDSPM spectrum in the case of single source. The beam width of the MSVDSPM spectrum is independent of the analysis frequency.%基于相位匹配原理的奇异值分解法(Singular value decomposition based on signal phase matching,SVD-SPM)的波达方向估计的均方根误差在高信噪比下无法逼近克拉美罗界,针对该问题提出了基于相位匹配原理的修正奇异值分解法(Modified singular value decomposition based on signal phase matching,MSVDSPM).该方法将阵列接收信号转换到频域,取相位匹配后各阵元中心频点频谱与其均值差值的距离平方和的倒数作为方向估计算子.仿真表明MSVDSPM方向估计的均方根误差可以在高信噪比下逼近克拉美罗界.MSVDSPM保持了SVD-SPM在单源入射时的尖锐谱峰,它等价于常规波束形成方法,并且其主瓣宽度与分析频率无关.