Optimal Threat-Based Radar Resource Management for Multitarget Joint Tracking and Classification

Abstract

The Radar Resource Management problem in a multitarget joint tracking and classification scenario is considered. The problem is solved using a previously introduced dynamic budget-balancing algorithm that models the sensor tasks as Partially Observable Markov Decision Processes. It is shown that tracking and classification tasks can be considered as a single task type. Furthermore, it is shown how the task resource allocations can be jointly optimized using a carefully formulated cost function based on the task threat variance. Multiple two-dimensional radar scenarios demonstrate how sensor resources are allocated depending on the current knowledge of the target position and class. In contrast to previous approaches, the novelty of this paper lies in combining tracking and classification performance into asingle cost function, preventing heuristic trade-offs.