Underdetermined Direction-of-Arrival Estimation From Second Order Statistics

Over the decades, two major methods have been developed to deal with the underdetermined Direction-of-Arrival (DOA) estimation problem where there are more sources than sensors. One is based on the higher-order statistics of the received signals, which needs a plenty number of snapshots to guarantee the estimation accuracy and thus is computational complex. The other relies on the sparsity of sources, which assumes that there exists at most one active signal in any time-frequency point. To overcome the aforementioned shortcomings, An extension of the MUSIC algorithm is presented to appropriate the underdetermined case by a set of covariance matrix with different lags, giving rise to the second-order underdetermined MUSIC (SOU-MUSIC) algorithm. The proposed algorithm has advantages over the existing underdetermined DOA estimation methods including robustness to the number of snapshots, computation efficacy as well as not requiring the source sparseness. Simulation results confirm the validity and high performance of the proposed algorithm.