Low angle estimation with colored noise in bi-static MIMO radar

This paper proposes a method to estimate the unknown angular location of a bi-static multiple input multiple output (MIMO) radar target in the low angle scenario. The multipath model is formulated, in which the specular reflection is treated as an image target and the diffuse reflection is taken as spatial correlated stationary Gaussian noise. Based on this model, the colored noise is cancelled by covariance differencing approach first. Then the differencing array covariance matrix is reconstructed into a block Hankel matrix to differentiate coherent signals of the actual targets and image targets. A unitary ESPRIT algorithm is utilized to give the accurate low angle estimations with low computational burden at last. Simulation results are presented to verify the effectiveness of the method.