Optimal Power Allocation Under Different Power Availability Scenarios for Multitarget Tracking With C-MIMO Radar Systems

Power allocation has emerged to be a critical problem when exploiting colocated multiple-input and multiple-output (C-MIMO) radar for multi-target tracking. Several prior approaches employing the quality of service-based framework aim to minimize the weighted sum of the target task utility functions. In this article, to utilize power-resource efficiently and further improve the tracking performance of the C-MIMO radar system, an optimal power allocation (OPA) method is proposed. First, the quality of service based power allocation model is generalized to a more general and flexible model, where the task utility functions can be selected from a set of monotonically increasing convex functions, and the construction of the objective function is not limited to a particular filter to approximate the Bayesian Cramer-Rao lower bound (BCRLB). Thus, more efficient non-linear Bayesian filters can be used. Second, quasi-convexity of the non-convex OPA problem under the quality of service-based framework is explored, whose objective function is the weighted sum of a set of separable quasi-convex functions. Then the strong duality between the original non-convex problem and its dual problem is derived. Finally, under any given approximated BCRLB, a dual projection subgradient power allocation (DPSPA) algorithm is proposed to deal with the dual problem and obtain the optimal solution. Illustrative numerical results demonstrate the efficiency and generality of the proposed strategy under different power availability scenarios.