Hybrid Cooperative Cache Based on Temporal Convolutional Networks in Vehicular Edge Network

With the continuous development of intelligent vehicles, people's demand for services has also rapidly increased, leading to a sharp increase in wireless network traffic. Edge caching, due to its location advantage, can provide more efficient transmission services and become an effective method to solve the above problems. However, the current mainstream caching solutions only consider content popularity to formulate caching strategies, which can easily lead to cache redundancy between edge nodes and lead to low caching efficiency. To solve these problems, we propose a hybrid content value collaborative caching strategy based on temporal convolutional network (called THCS), which achieves mutual collaboration between different edge nodes under limited cache resources, thereby optimizing cache content and reducing content delivery latency. Specifically, the strategy first obtains accurate content popularity through temporal convolutional network (TCN), then comprehensively considers various factors to measure the hybrid content value (HCV) of cached content, and finally uses a dynamic programming algorithm to maximize the overall HCV and make optimal cache decisions. We have obtained the following conclusion through simulation experiments: compared with the benchmark scheme, THCS has improved the cache hit rate by 12.3% and reduced the content transmission delay by 16.7%.