Adaptive Variable Length Sliding Window Network Coding for Low Latency Transmission in FANETs via Deep Reinforcement Learning

Future flying ad-hoc networks (FANETs) need to address issues related to delay and channel interference while ensuring high data transmission accuracy. In this letter, we propose a proximal policy optimization (PPO)-based adaptive adjustment algorithm for sliding window-based network coding to mitigate channel contention and signal degradation caused by increased traffic load. We model the adaptive adjustment of the sliding window size as a Markov decision process, considering effective rate, node congestion, and total delay in the reward function. To further improve the algorithm's performance, we enhance PPO with long short-term memory (LSTM) networks to process time-series data. Experimental results demonstrate that our method reduces delay, improves packet delivery rate and throughput compared with traditional sliding window network coding algorithms.