IEEE Sensor Array and Multichannel Signal Processing Workshop, Darmstadt, Germany, 21 - 23 July 2008, pp.420-421
In this paper, we present a new application of the likelihood ratio test to solve the problem of enumeration and localization of multiple emitters (sources) with a uniform linear antenna array in a multi-path environment. First, spatial smoothing is applied to the antenna array data to decorrelate signals corresponding to each path in the estimated correlation matrix and the number of paths is then counted using an information theoretic criterion. Then, MUSIC is applied with this correlation matrix to estimate the angle of arrival of all possible incident paths. In our signal model, we assume that the correlation between signals from two paths is one if they originate from the same source but zero if they are from different sources. The likelihood ratio test is then applied to each path pair using this correlation model and the angles of arrival estimated by MUSIC in order to find out if the path pair originates from the same emitter, and hence the number of emitters and their angles of arrival are estimated. The proposed algorithm on the likelihood ratio test is compared with the correlation coefficient algorithm based on minimum variance beam-forming with a performance analysis.