A dynamic execution offloading model for efficient mobile cloud computing

摘要

With the emerging of diverse applications, power-sensitive mobile devices are facing the severe challenge of computational stress. Many researchers propose the schemes of migrating the mobile devices' computational burden to their adjacent powerful cloud servers. In these schemes, the execution context as well as the program itself must be sent to the cloud. The transfer time is the key element affecting the performance. In this paper, we propose the scheme of shrinking and reusing the migrated data to accelerate the transmission process. Besides, on the cloud side, we design a caching algorithm of IB-LRU (Importance-based LRU) to manage the migrated data. We evaluate our scheme by comprehensive experiments. Our IB-LRU algorithm surpasses traditional caching algorithms such as LRU by discriminating frequently and infrequently referenced cached data. The experiment results show that our scheme significantly improves the overall performance of mobile application migration by reducing the transmission time.