Mobile Cloud Computing (MCC) helps increasing performance of intensive mobile applications by offloading heavy tasks to cloud computing infrastructures. The first step in this procedure is partitioning the application into small tasks and identifying those that are better suited for offloading. The method call partitioning strategy splits the code into a set of method calls that are offloaded to remote servers. Quite often, many applications need to make use of multiple servers for parallel processing of intensive computational operations. Predicting the behavior of such parallelizable applications is not an easy task. Deciding the number of remote servers determines the performance of the applications and the costs of the cloud usage. On one hand, users are interested in improving the performance of their applications, so they would like to use as many servers as possible, but on the other hand, they would also like to reduce their costs by using fewer cloud resources. In this paper, we propose a Stochastic Petri Net (SPN) modeling strategy to represent method call executions of mobile cloud systems. This approach enables a designer to plan and optimize MCC environments in which SPNs represent the system behavior and estimate the execution time of parallelizable applications.

Mobile Cloud Computing (MCC) combines mobile computing and cloud computing aiming to aid performance of mobile devices. The idea is simple: thin devices offload heavy methods to resource-rich servers in the clouds. We believe that in the near future MCC will adopt more advanced offloading techniques. In particular, in this paper we envision a scenario where offloading frameworks will have to deal with GPU code offloading. Amazon already offers instances with Graphics Processing Units (GPU), which can be used for this purpose. We propose and implement MCC-Adviser, a simulation tool that can predict the performance of GPUs with different number of cores using Stochastic Petri Nets. We tested MCC-Adviser in a case study with one of the expensive Amazon GPU instances. The simulations showed that it is possible to minimize costs, while satisfying user's quality of service requirements, by utilizing less powerful instances.