QoS-aware resource allocation for mobile media services in cloud environment

摘要

With the advancement of technology in smartphones and their popularity, users are expecting higher performance and functionality from mobile devices just as powerful non-mobile devices, especially in the context of multimedia applications. However, despite significant advances in handheld devices, there still exists a gap between device capabilities and the requirements of applications. Mobile Cloud Computing emergence has overcome the barriers of the mobile device restrictions. To improve the performance of media services processed in clouds, an efficient resource allocation mechanism is needed. In this paper, we propose a resource allocation scheme in Mobile Cloud Computing based on Learning Automata technique, taking into account three fundamental concerns: total response time of various types of mobile media services, if they are to be executed on cloud; level of uncertainty (i.e. prone to failure measurement); and computational capacity requirement of services. Simulation results demonstrate that the proposed resource allocation scheme can optimally allocate cloud resources to each media service while ensuring minimal QoS requirements of applications.