Distributed Resource Allocation for Mobile Users in Cache-Enabled Software Defined Cellular Networks

Abstract

In this paper, we study the problem of resource allocation in a cache-enabled software defined cellular network (SDCN) with mobile users, where the SDCN controller has global information of the network and the popular contents that the users request are stored at the content server and cache-enabled small base stations (SBSs). We propose a Markov chain based model to predict the users' mobility patterns and then use the predicted mobility patterns to determine optimal resource allocation. The mobility prediction and resource allocation problem are jointly formulated as an optimization problem whose goal is to maximize the network throughput. Based on the predicted users' mobility patterns, a distributed alternating direction method of multipliers (ADMM) is proposed to solve the resource allocation problem. The proposed ADMM algorithm enables the multiple SBSs implement their resource allocation simultaneously and, hence decreases the control overhead of the SDCN controller. Simulation results show that the proposed algorithm achieves up to 9.35% and 33.17% gains in terms of the average throughput compared to the random algorithm and the nearest association with equal allocated resource algorithm.