Cooperative Edge Caching Based on Temporal Convolutional Networks

With the rapid growth of networked multimedia services in Internet, wireless network traffic has increased dramatically. However, current mainstream content caching schemes do not take into account cooperation different edge servers, resulting deteriorated system performance. In this paper, we propose a learning-based scheme to enable mutual among servers with limited resources, thus effectively reducing delivery latency. Specifically, formulate cooperative problem as an optimization problem, which is proven be NP-hard. To solve design new strategy (LECS) that encompasses three key components. Firstly, temporal convolutional driven popularity prediction model developed estimate high accuracy. Secondly, predicted popularity, concept value (CCV) introduced weigh cached on given server. Thirdly, novel dynamic programming algorithm maximize overall CCV. Extensive simulation results have demonstrated superiority our approach. Compared state-of-the-art schemes, LECS can improve cache hit rate by 8.3%-10.1%, and reduce average delay 9.1%-15.1%.