Priority-Aware DDPG Empowered Multi-Active-RIS Beamforming Framework for Energy-Efficient ISAC

This paper investigates energy-efficient integrated sensing and communication (ISAC) aided by multiple active reconfigurable intelligent surfaces (active-RISs). We develop a multi-active-RIS-enabled joint beamforming framework and formulate a system-level power optimization problem that explicitly accounts for the active-RIS amplification power in ISAC. To support real-time beamforming decisions, we propose a priority-aware deep deterministic policy gradient (DDPG) algorithm with a hierarchical reward design that provides clear exploration guidance for high-dimensional joint beamforming. Simulation results show that the proposed framework improves overall power efficiency compared with passive-RIS and conventional DDPG baselines under different ISAC requirements and network scales.