abstract
- In this paper, we propose DRHRP, a novel deep reinforcement learning (DRL)¿based hybrid routing protocol for unmanned aerial vehicle (UAV)¿enabled wireless networks. DRHRP integrates concepts from geocast routing, energy-efficient and connectivity maintenance strategies, and multi-agent DRL to dynamically adapt to the challenges of highly mobile and resource-constrained networks. By modeling the routing decision as a Markov Decision Process (MDP) and employing a Multi-Agent Deep Deterministic Policy Gradient (MADDPG) framework, our protocol learns optimal forwarding policies using only local observations. Extensive simulations demonstrate improvements in packet delivery ratio, delay reduction, and energy consumption compared to traditional schemes. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.