Context-aware Distributed Storage in Mobile Cloud Computing

In this paper, we propose a context-aware distributed data storage mechanism to optimize the performance of reading/writing in the emerging mobile cloud computing environment. In particular, the exploited context information including user mobility pattern, network condition, and data access preference (intensity to read and write/update). The context information is utilized to improve the efficiency of read and write from/to mobile devices. In the context-aware distributed storage system, we consider the 3 types of user context in determining the optimal assignment of chunks based on Reed-Solomon erasure code. Then we propose to solve the optimization problem by a mixed integer linear programming (MILP). The efficiency and effectiveness of this context-aware mechanism is evaluated in a variety of scenarios by simulations.