DOA estimation of non-coherent and coherent sources

Abstract

Array signal processing is currently widely used in many fields. It has been a hot research area for several decades. Direction of arrival (DOA) estimation has evolved from high-resolution to super-resolution. In view of the wide application of DOA estimation, we have considered this topic under various conditions. Traditional DOA estimations are based on signal subspace, such as the MUSIC algorithm and ESPRIT algorithm. They have excellent estimation performance when the signal sources are non-coherent. These algorithms are affected by some parameters, such as signal-to-noise ratio (SNR), the number of array elements and the number of snapshots being used. Then, with the progress of compressed sensing, many researchers have applied this theory to DOA estimation. This algorithm based on sparse representation still has good estimation performance even at low SNR and a small number of snapshots. However, in actual situations, due to the influence of multipath, the signal sources received by the array are usually coherent, which will greatly reduce the performance of DOA estimation algorithms based on subspace. Then, a serious of decoherence algorithms appeared. Among them, the commonly used method is the spatial smoothing algorithm, which recovers the rank of the data covariance matrix at the cost of reducing the degree of freedom. The DOA estimation algorithm based on sparse representation is not affected by coherent sources, and the excellent characteristics of this algorithm can be achieved.