Model-based adaptive detection and DOA estimation using separated sub-arrays

The potential performance of adaptive detection and direction-of-arrival (DOA) estimation using a certain class of sparse linear arrays, characterized by two widely separated sub-arrays, in combination with model-based techniques, is investigated. With this array structure, a large baseline is obtained with a limited amount of hardware. It is found that the very narrow main-lobe obtained by separating the sub-arrays can be utilized to obtain accurate DOA estimates, high angular resolution, and good detection performance even in close vicinity of a localized interference source. However, the presence of grating lobes deteriorates the performance at low SNR and also when the angular separation between two targets, or between a target and an interfering source, corresponds to an integer times the distance between the grating lobes.