Dynamic Clone Sharing Scheme in Mobile Cloud Computing: A Delaunay Triangulation Approach

Abstract

Owing to the nature of elastic resource pooling, cloud computing offers a great opportunity to extend the potentials of Mobile Devices (MDs) for supporting the resources-hungry applications. In the mobile cloud platforms, each MD is associated with a system-level clone in the cloud infrastructure. It provides the potentials in supporting the clone sharing. The clone sharing can reduce the communication delay between an MD and its clone when the MD moves far away from its clone. It is because that clone sharing can migrate the service from the remote source clone to a nearby target clone. Considering the mobility of MDs, how to select the appropriate target clone is still a big challenge. In this paper, we first pose the clone sharing question. The clone sharing can reduce the service delay between a MD and its clone when the MD move far away from its clone. It is because that it migrate the service from the remote source clone to a local target clone. However, how to select the appropriate target clone is a big challenge. We propose a clone sharing scheme, called as Triangulation-based Clone Sharing Scheme (TCS), to solve the problem. In the TCS scheme, we firstly disclose the relationship between clone serving area and the MD. Secondly, the TCS scheme splits the clone network into Triangle topology by our Clone Triangulation Topology Construction and Reconstruction (CTTCR) algorithm. It can accurately obtain the target clone for clone migration. We also prove the correctness of CTTCR algorithm. Simulation results show that our scheme works very well.