Parametric Analysis of Mobile Cloud Computing Frameworks Using Simulation Modeling

Mobile Cloud Computing (MCC) frameworks implement mechanisms for selecting tasks in an application and offloading those tasks for execution on a cloud server. Task partitioning and task offloading aim to optimize performance objectives, like lower energy usage on mobile devices, faster application execution, while operating even in unpredictable environments. Offloading decisions are influenced by several parameters, like varying degrees of application parallelism, variable network conditions, trade-off between energy saved and time to completion of an application, and even user-defined objectives. In order to investigate the impact of these variable parameters on offloading decision, we present a detailed model of the offloading problem incorporating these parameters. Implementations of offloading mechanisms in MCC frameworks often rely on only a few of the parameters to reduce system complexity. Using simulation, we analyze influence of the variable parameters on the offloading decision problem, and highlight the complex interactions among the parameters.