Distributed mobile cloud computing: A multi-user clustering solution

Abstract

Edge computing through local mobile cloud computing platforms is a key enabler for coping with the ever increasing data traffic requirements. A key enabler for this technology is the awaited ultra-dense deployment of radio access points for future 5G networks. Local cloud platforms allow maintaining a scalable network design by jointly managing local radio and computational resources. The Fog, a platform with rich services, introduces distributed intelligence at the edge of the network where entities such as radio access points form a local computing resources pool. In this paper, we address the problem of radio access points clustering for fog computing applications. We focus on the multi-user case where the local cloud resources are to be shared by several devices. We propose a novel clustering algorithm in which management functionalities are split into two layers: centralized and decentralized. The proposed strategy compromises centralized optimality with decentralized distribution intelligence for faster and less complex decision making. We compare, through simulations, the performance of the proposed algorithm to centralized and decentralized strategies, and show how it can achieve good quality of experience.