Joint Task Selection and Resource Allocation for Multi-Target Tracking Under Suppression Jamming in Networked Radar Systems

Improving target tracking performance under suppression jamming remains a challenge for radar systems. A practical method to address this issue is coordinating multiple radars through a networked radar system (NRS). This paper introduces a joint target tracking framework that includes radar active-passive task selection and dwell time allocation. A constraint objective function that aims to increase tracking accuracy on multiple targets is constructed by building both active and passive measurement models. An iterative approach based on reinforcement learning is proposed. Specifically, we train an agent to optimize radar task selection and employ convex optimization tools to allocate dwell time. The agent is tested in simulation scenarios involving different target settings and varying levels of jamming intensity. The results demonstrate the method's capability to effectively improve overall tracking performance of the NRS compared to traditional rule-based approaches. The relationship between jamming intensity and radar task selection is further analyzed.