Multi-target localization for non-uniform linear array of MIMO radar

A multi-target localization method for non-uniform linear array of multiple input multiple output (MIMO) radar is presented. For this approach, the received data of the non-uniform linear array of MIMO radar can be dealt with array interpolation technique, which satisfies the characteristics of a virtual uniform linear array of MIMO radar. Then the signal subspace of the virtual uniform linear array of MIMO radar is obtained and the target's azimuth angle can be estimated by estimating signal parameter via rotational invariance techniques (ESPRIT). The strongpoint is that the ESPRIT can be extended to other array geometries and the virtual sensors are increased by the proposed method. The simulation results show that more targets can be estimated and the performance of angle estimation is close to the root of Cramer-Rao bound by the proposed method. Furthermore, the estimated parameters are automatically paired without additional paring computation.