Cloudlet deployment in local wireless networks: Motivation, architectures, applications, and open challenges

摘要

In past few years, advancement in mobile applications and their integration with Cloud computing services has introduced a new computing paradigm known as Mobile Cloud Computing. Although Cloud services support a wide range of mobile applications, access to these services suffers from several performance issues such as WAN latency, jitter, and packet losses. Cloudlet frameworks are proposed to overcome these performance issues. More specifically, Cloudlets aim to bring the Cloud or a specific part of the Cloud closer to the mobile device by utilizing proximate computing resources to perform compute-intensive tasks. This paper presents a comprehensive survey on the state-of-the-art mobile Cloudlet architectures. We also classify the state-of-the-art Cloudlet solutions by presenting a hierarchical taxonomy. Moreover, the areas of Cloudlet applications are also identified and presented. Cloudlet discovery, resource management, data security, mobility, application offloading, and most importantly incentives to deploy a Cloudlet are areas that still needs to be investigated by the research community. The critical aspects of the current Cloudlet frameworks in Mobile Cloud Computing are analyzed to determine the strengths and weaknesses of the frameworks. The similarities and differences of the frameworks based on the important parameters, such as scalability, mobility support, Internet dependency, dynamic configuration, energy savings, and execution cost, are also investigated. The requirements for deploying the Cloudlet in a Local Wireless Network are also highlighted and presented. We also discuss open research challenges that Cloudlet deployments face in Local Wireless Networks. (C) 2015 Elsevier Ltd. All rights reserved.