PRVC: A Novel Vehicular Ad-Hoc Network Caching Based on Pre-trained Reinforcement Learning

In recent years, network caching in Vehicular Ad-hoc Network (VANET) has gained significant interest, with particular interest around high mobility nodes. Traditional methods, based on Mobile Ad-hoc Network (MANET) caching, face limitations due to MANET’s computational constraints and struggle to effectively address VANET-specific caching needs, leading to subpar performance. Addressing this, this paper introduces the novel Vehicular Ad-hoc Network Caching Method based on Pre-trained Reinforcement Learning (PRVC). This new approach uses pre-trained reinforcement learning to enhance VANET caching Quality of Service (QoS) and adapt to changing cache interests. Our empirical experiments show that PRVC outperforms benchmarks in cache hit rate, latency, and link load.