A Markov Decision Process-based service migration procedure for follow me cloud

The Follow-Me Cloud (FMC) concept enables service mobility across federated data centers (DCs). Following the mobility of a mobile user, the service located in a given DC is migrated each time an optimal DC is detected. The detailed criterion for optimality is defined by operator policy, but it may be typically derived from geographical proximity or load. Service migration may be an expensive operation given the incurred cost in terms of signaling messages and data transferred between DCs. Decision on service migration defines therefore a tradeoff between cost and user perceived quality. In this paper, we address this tradeoff by modeling the service migration procedure using a Markov Decision Process (MDP). The aim is to formulate a decision policy that determines whether to migrate a service or not when the concerned User Equipment (UE) is at a certain distance from the source DC. We numerically formulate the decision policies and compare the proposed approach against the baseline counterpart.

I. Introduction

To cope with the ever-growing mobile traffic, mobile operators have been looking into the decentralization of their core network, along with devising traffic offload-based solutions [1] [2]. On the other hand, the fast growing business of clouding computing is pushing for the deployment of regional Data Centers (DCs) [3] [4]. Internet Service Providers (ISPs) are also pushing micro DCs closer to customers, enabling third parties to rent them out [11]. Connecting these geographically distributed micro DCs as well as macro DCs, together into a common resource pool, to deliver a variety of cloud services forms the so-called federated cloud. The distribution of cloud computing resources over different locations in the network is beneficial for different reasons such as increasing availability, reducing bandwidth cost, and reducing latency by locating resources nearby users.