Efficient Data Distribution and Sharing in Mobile Cloud Computing

Abstract

The combination of Mobile computing and Cloud computing gave rise to a new computing paradigm called Mobile Cloud Computing (MCC) that gives the  exibility to access information and computing resources anywhere anytime. In many applications, mobile nodes capture images/video clips and exchange with other mobile peers (on demand) and use local access points for efficient smooth distribution of information in the wireless network. In such a situation, a three-way data management and dissemination technique is helpful because it provides both data management and distribution at different levels of granularity. The main motivation of this paper is that we seek a balance between accessing information from a remote cloud server, at the cost of increasing latency, and accessing data from other mobile hosts, at a cost of power, bandwidth and increased traffic. To manage this balance, we propose a layered architecture supported by mobile hosts, access points and cloud together for efficiency. The MCC architecture described in this paper also provides fault tolerance in case one of the mobile nodes fails or gets disconnected. In this paper, we discuss a layered architecture for MCC and present a data pre-distribution scheme for efficient sharing of data among potential users in MCC environment. We also propose update propagation and cache replacement policies in MCC. In addition, we report complexity analysis for cache accesses by mobile nodes using the proposed architecture in terms of communication messages. We implemented the proposed schemes using EC2 and mobile devices and evaluated the performance of accessing the data from different levels of architecture.