Separation and tracking of multiple broadband sources with one electromagnetic vector sensor

Abstract

A structure for adaptively separating, enhancing and tracking uncorrelated sources with an electromagnetic vector sensor (EMVS) is presented. The structure consists of a set of parallel spatial processors, one for each individual source. Two stages of processing are involved in each spatial processor. The first preprocessing stage rejects all other sources except the one of interest, while the second stage is an adaptive one for maximizing the signal-to-noise ratio (SNR) and tracking the desired source. The preprocessings are designed using the latest source parameter estimates obtained from the source trackers, and a redesign is activated periodically or whenever any source has been detected by the source trackers to have made significant movement. Compared with conventional adaptive beamforming, the algorithm has the advantage that no a priori information on any desired signal location is needed, the sources are separated at maximum SNR, and their locations are available. The structure is also well suited for parallel implementation. Numerical examples are included to illustrate the capability and performance of the algorithm.