Energy-Efficiency for Smartphones Using Interaction Link Prediction in Mobile Cloud Computing

Jiuyun Xu, Chao Guan, Xiangrui Xu

Abstract: Along With vast deployment of mobile cloud computing systems, users accessing any information on the Internet by smart phones are often based on continuous data communication. However, when the communication status is unstable, the mobile client needs to establish multiple connections with the cloud. This leads to great energy consumption, which poses a huge challenge to the usability of mobile cloud computing systems. In this article, considering the similarity of the accessibility data of strong interactive users and the predictability of user behaviour data, we proposes a link prediction method based on the maximization of user interaction behaviour (Maximize Interaction Link Prediction) in a specific environment for the mobile cloud computing: First, based on the data prediction model, we use the interaction degree method to improve the access data prediction for known users; Secondly, combining with the social network method we analyze and filter the prediction data; At last, we pre-store the above prediction data by the pre-storage mechanism. The Evaluations show that it can reduce mobile energy consumption significantly by around 20%.