DOA estimation of coherent signals based on modified music algorithm

Abstract

The presence of Abstract-coherent signal sources causes the vectors in the signal subspace to emanate into the noise subspace, resulting in incorrect DOA estimation. In this paper, a MUSIC algorithm for multiple signal classification in feature space based on forward-backward spatial smoothing is used. This algorithm is further improved and its performance is simulated under matlab, while the principle of this improved algorithm is to preprocess the coherent signal and then apply the eigenspace MUSIC algorithm to it for accurate DOA estimation. The computer simulation results show that the algorithm is more effective in estimating and resolving the DOA of small signal-noise ratio sources close to each other with higher resolution, lower sidelobe, and sharper spectral peaks, without affecting the estimation of uncorrelated sources.