Adaptive Intelligent Routing in Ultra-High-Speed FANETs Using Deep Reinforcement Learning

With the rapid advancement of drone technology, Flying Ad Hoc Networks (FANETs) have gained significant importance in applications such as disaster response, security patrols, and environmental monitoring. The emergence of ultra-high-speed UAVs presents unique challenges to the stability and efficiency of network communications within FANETs due to their high mobility. This paper proposes a novel DRL-based adaptive intelligent routing strategy specifically designed for ultra-high-speed FANETs, leveraging deep reinforcement learning to make autonomous routing decisions in highly dynamic network environments. The method includes a hierarchical priority experience replay pool mechanism to guide the DRL model in balancing immediate and overall optimal routing decisions, overcoming the limitations of traditional normalized reward functions. Experimental results show that the proposed DRL-based adaptive intelligent routing strategy significantly improves packet delivery rate, average end-to-end delay, and per-hop delay compared to traditional routing protocol. However, high energy consumption remains a limitation, requiring further research and optimization to improve energy efficiency in the future.