GLRT Based Subspace Detector for Joint Target Detection and Imaging in Monostatic MIMO-OFDM Radar Systems

This work presents a generalized likelihood ratio test (GLRT) based subspace detector for joint multiple target detection and 2D imaging in MIMO-OFDM Radar Systems. In particular, this paper considers a challenging scenario with weak stationary targets together with unknown ranges and angles of the target locations. Hence, the proposed GLRT framework aims to detect the targets followed by the estimation of the associated radar cross sections, ranges and angular bins. Further, this framework is also extended to a scenario with unknown noise variance. It is then demonstrated that the finite record performance of the derived GLRT based subspace detector, in terms of the probabilities of false alarm and detection, achieves the asymptotic bound. 2D imaging of the target in the angular and range dimensions is also developed through estimation of the scattering scene along with derivation of the associated Cramer-Rao lower bound (CRLB) for mean square error (MSE) of estimation. Simulation results demonstrate a significantly improved target detection, target parameter estimation performance in comparison to the conventional energy detector with/without noise uncertainty and existing MUSIC based imaging scheme for MIMO-OFDM radar systems.