Joint Subarray Selection and Power Allocation for Cognitive Target Tracking in Large-Scale MIMO Radar Networks

This article develops a joint subarray selection and power allocation (JSSPA) strategy for tracking multiple targets in clutter environments using large-scale distributed MIMO radar networks. The mechanism of our strategy is to implement the best resource allocation on the basis of the information in the tracking recursive manner, aiming at improving the overall tracking accuracy. By integrating with the information reduction factor, we derive the predicted conditional Cramér-Rao lower bound (PC-CRLB) in clutter, which offers a more accurate measure of target state estimate than the standard posterior Cramér-Rao lower bound. Then, the sum of weighted PC-CRLBs is utilized as the optimization criterion to guide our JSSPA strategy. It is shown that the optimization model is a nonconvex problem that involves three variables, and a two-stage local search-based algorithm is proposed to solve it. Numerical simulations verify the tracking performance improvement by the proposed method, compared with four other resource allocation strategies.