A Cache Model And Consistency Maintenance Strategy In Mobile Cloud Computing

With the technological advancement of mobile internet and mobile terminal, the internet has become accessible anywhere any time. Specifically, as the 3rd generation mobile internet technology develops, the increased bandwidth enables people to connect to the Internet for various kinds of information via mobile terminals. While mobile cloud computing allows users to login to the cloud computing system from a mobile terminal to purchase computing power and services according to their own needs, the increasing storage capacity in mobile terminals makes it feasible and necessary to utilize caching technology to minimize expenses on internet data communication. Therefore, it has become an active research area to get better implementation of caching technology through reasonably allocating storage spaces of mobile terminals, in order to minimize the expenses and delay of data access. This thesis mainly studies the cache model and consistency maintenance strategy in mobile cloud computing. Noticing the increasing cache space in mobile terminals, the thesis proposes a cache model based on an active data-pushing mechanism. The model is constructed so that when a data request is received from the mobile terminal, the cloud computing system predicts and pushes the data that are likely to be used in the future, aiming at minimizing expenses and response delay. To construct this cache model, cache data are filtered according to time sensitivity and urgency level, thus formulating an improved cache consistency maintenance strategy based on the traditional probabilistic consistency maintenance mechanism. By filtering different types of data according to actual consistency requirement from the users, the model also seeks to strike a balance between satisfying these requirements and mitigating system overhead. Also, it is applied to the model consistency maintenance strategies based on dynamic push and dynamic pull, respectively, on the source date node and mobile terminal. Performance evaluation by simulation is conducted by the building the CacheSim system on the JiST simulation platform.