Adaptive Resource Management in Cognitive Radar via Deep Deterministic Policy Gradient

In this paper, scanning for target detection, and multi-target tracking in a cognitive radar system are considered, and adaptive radar resource management is investigated. In particular, time management for radar scanning and tracking of multiple maneuvering targets subject to budget constraints is studied with the goal to jointly maximize the tracking and scanning performances of a cognitive radar. We tackle the constrained optimization problem of allocating the dwell time to track individual targets by employing a deep deterministic policy gradient (DDPG) based reinforcement learning approach. We propose a constrained deep reinforcement learning (CDRL) algorithm that updates the DDPG neural networks and dual variables simultaneously. Numerical results show that the radar can autonomously allocate time appropriately so as to maximize the reward function without exceeding the time constraint.