Caching hit ratio maximization in mobile edge computing with node cooperation

In this study, we investigate the cooperation problem of edge nodes in mobile edge scenarios. We study two performance metrics: caching hit ratio and cooperation cost. We analyze the relationship of the two metrics for cooperative edge networks in which multiple edge nodes form a forwarding group to serve users. We construct a spanning tree to organize the forwarding group, and maximize the hit ratio subject to a budget constraint on cooperation cost. We first address the optimization problem using the method of Lagrangian multipliers and derive the optimal group parameter. Furthermore, a distributed optimization algorithm with alternating direction method of multipliers (ADMM) via constraint projection and variable splitting is proposed to achieve the desired goal. Finally, we build an edge computing simulation platform consisting of hundreds of edge nodes and users as a case study to verify the effectiveness and efficiency of our proposed method. Numerical results show that it significantly improves the caching hit ratio up to 60%, compared to the classic scheme.