Cramer-Rao lower bound for multitarget localization with noncoherent statistical MIMO radar

In this paper, we focus on the theoretical localization accuracy of two localization algorithms in noncoherent MIMO radar systems with widely separated antennas. The first one is the optimal method for multitarget localization which is simply to expand the dimension of the parameter vector and thus perform a global maximum of the joint likelihood function of all the targets. The second one is a suboptimal called successive-interference-cancellation (SIC) algorithm proposed in our previous work [1] which localizes targets one-by-one and clears the interference of previous declared targets. The Cramer-Rao lower bound (CRLB) for these two algorithms has been derived and compared with emphasis on special cases where some targets share no common range bins with any other targets. Numerical results demonstrate that the suboptimal SIC algorithm has little theoretical performance loss compared with the optimal method even when targets share some common range bins and the loss may be reduced by increasing the number of radar elements.