Fast detection algorithm for the number of signal sources based on modified hung-turner projection

Detection of the number of spatial signals is one of the key issues in array signal processing community. Most of the existing methods for detection of the number of signal sources require the eigenvalues of the sample covariance matrix. In this paper, a Fast Detection Method via Modified Hung-Turner Projection (FDM-MHTP) is proposed to dectect the number of multiple emitters, where a threshold is derived exploiting the asymptotic distribution properties of the estimation errors of the sample covariance matrix. The threshold is utilized to detect the number of signal sources in conjunction with performing Gram-Schmidt(GS) orthogonalization. Without eigenvalue decomposition, the algorithm has good detection performance and low computation complexity. Computer simulation shows the correctness and efficiency of the algorithm.