Antenna selection in MIMO radar with collocated antennas

An antenna adjustment strategy is developed for the target tracking problem in the collocated multiple-input multipleoutput (MIMO) radar. The basic technique of this strategy is to optimally allocate antennas by the prior information in the tracking recursive period, with the objective of enhancing the worst-case estimate precision of multiple targets. On account of the posterior Cramer-Rao lower bound (PCRLB) offering a quantitative measure for target tracking accuracy, the PCRLB of joint direction-of-arrival (DOA) and Doppler is derived and utilized as the optimization criterion. It is shown that the dynamic antenna selection problem is NP-hard, and an efficient technique which combines convex relaxation with local search is put forward as the solution. Simulation results demonstrate the outperformance of the proposed strategy to the fixed antenna configuration and heuristic search algorithm. Moreover, it is able to offer close-to performance of the exhaustive search method.