Freshness-Aware Target Tracking in UAV-Assisted Integrated Sensing and Communication Systems

Unmanned aerial vehicles (UAVs) play a crucial role in time-critical missions, such as surveillance and dynamic target tracking, requiring precise sensing and rapid data delivery. This paper investigates an Integrated Sensing and Communication (ISAC) system enabled by UAV for detecting and tracking moving targets while updating the base station (BS) with fresh information about the target’s position. We propose a scheduling framework for sensing, communication, and flight that leverages the Age of Incorrect Information (AoII) metric to ensure information timeliness and the Cramer-Rao Bound (CRB) to maintain correctness by addressing uncertainty from imperfect sensing, while considering energy consumption. We formulate the problem as a Markov Decision Process (MDP) to capture the sequential decision-making for sensing schedule, communication decisions, and flight trajectory. The MDP introduces challenges of a large state space and high computational complexity. To address these challenges, we employ a Deep Q-Network (DQN)-based reinforcement learning approach. Simulation results demonstrate that the proposed method effectively ensures the freshness of accurate information and reduces energy consumption.