Beamspace MIMO Radar Tensor Modeling and 2-D DOA Estimation

Multiple-input multiple-output （MIMO） radar suffers from the transmit energy loss due to its omni-directional transmit beampattern， and the DOA estimation methods mostly exploit only the signal covariance matrix in a single pulse， whose performance is poor at low signal-to-noise （SNR）. To improve the DOA estimation performance， a higher-order tensor model and a fast tensor decomposition method are proposed for beamspace MIMO radar in application to two-dimensional （2-D） direction of arrival （DOA） estimation. The designed tensor exploits the multi-linear structure of the received data in MIMO radar with multiple pulses. As compared to conventional tensor decomposition method， the proposed fast tensor decomposition method takes advantage of the Vandermonde factor matrix and requires only basic linear algebra. The computational complexity is reduced significantly. Simulation results show that the proposed method surpasses other DOA estimation methods with a better accuracy and a higher resolution.