Cooperative Caching Strategy Based on Dynamic Cache Value Perception in Edge Networks
Yue Guo, Guangxu Zhou, Junfeng Hou, Hongjun Zhang, Yaofeng Li · 2022
Caching the content that users will request at the edge of the network can effectively alleviate the pressure of the core network, but the existing research results pay less attention to the cache value of the content, which makes the cache strategy performance low. A dynamic cache value aware collaborative cache strategy is proposed, which is different from the traditional single mode. The user layer caches the content according to the user’s request intention, and the MBS and SBS layers consider the global information for collaborative cache placement. Combined with the user’s movement trajectory, residence time and user’s request intention, a value perception method of dynamic content cache is proposed to determine the content to be cached. A content cache decision algorithm based on deep reinforcement learning is also designed, which uses the cooperative cache placement decision with MBS and SBS layers. Simulation results show that the cache hit rate of the proposed cache strategy is increased by 5.0% 10.8% and 18.9% on average compared with the three baseline schemes, indicating that the proposed strategy can significantly offload core network traffic and improve edge cache performance.