Abstract:

A Khatri-Rao (KR) sparse decomposition algorithm based on KR product with sparse signal representation is proposed in the light of direction of arrival (DOA) estimation of quasi-stationary signals. The received data are processed as the jointsparse multiplemeasurement vectors through KR product, and DOA is estimated by solving the jointsparse inverse problem. The existence and uniqueness condition of the jointsparse inverse problem is also provided, which addresses the issue that the present DOA estimation with sparse decomposition can not handle the underdetermined situation. The simulation results demonstrate the effectiveness of the proposed approach and the higher resolution in a lower signal-to-noise ratio environment.