Robust Power Allocation for Resource-Aware Multi-Target Tracking With Colocated MIMO Radars

Abstract

Power allocation is an essential part of the design of colocated multiple-input and multiple-output (C-MIMO) radar systems for multi-target tracking (MTT). A widely adopted power allocation approach is to directly minimize the amount of allocated power while satisfying the desired performance for each tracked object so as to conserve the power resources. However, this approach may not lead to a satisfactory solution when the available power is not sufficient to achieve the desired performance for all the targets. To overcome the limitations of this approach, a robust power allocation (RPA) methodology is proposed in this paper based on the quality of service framework. At its core, the proposed RPA employs a task utility function that quantifies the tracking performance for different power allocations in a flexible manner. The Bayesian Cramér-Rao lower bound (BCRLB) is utilized to formulate the task utility function since it provides a lower bound on the accuracy of target state estimates. By using a set of weights to quantify the importance of different tracked objects, the objective function of RPA is modeled as the weighted sum of task utility functions. This formulation of the RPA problem is demonstrated to be a non-convex optimization problem in general. We show that the task utility function is unimodal. Based on the specific structure of the task utility function, we propose an iterative parallel search algorithm to find the solution. Numerical experiments involving scenarios with both sufficient and insufficient power resources demonstrate the robustness and efficiency of the proposed strategy.