Evolutionary Adaptation Mechanism for Edge Caching Under Propagation Dynamics

In mobile-edge network, adaptive cache placement and resource allocation are vital for meeting the dynamic requirements of users. The current network cache performance has not been fully implemented, due to the lack of comprehensive knowledge of content popularity dynamics and the impact of network resource deployment on the dynamics. In this article, an epidemic model is applied to characterize the content propagation among users. Based on this model, a joint optimization problem of cache placement and base station resource allocation is formulated, maximizing the satisfaction rate of users. The joint optimization problem is decomposed into two subproblems: 1) caching placement and 2) resource allocation. First, the cache placement strategy is determined based on the dynamic propagation model. Second, considering the impact of network decisions on content propagation, a belief propagation-based iterative algorithm is proposed for dynamic resource allocation. Finally, simulation results reveal that the proposed joint optimization algorithm outperforms the benchmark algorithms.