Data Prefetching Driven by User Preference and Global Coordination for Mobile Environments

Most of the prefetch technologies used in the current cache service schemes did not consider systematically different issues such as the mobile clients´ user preferences, Client/Server cooperation and global coordination. This paper presents an approach named Multiple Goals-Oriented Data Prefetching (MGODP) addressing the above problem. MGODP takes the user preferences into account to best satisfy user requirements. Since MGODP focuses on Client/Server data access model, it can adopt global coordination to achieve optimal allocation of network bandwidth from the server perspective and balance the workload between the mobile client and the backend server through autonomic cooperation. Simulation results demonstrate the positive effect of achieving users´ specific goal and show that this mechanism significantly outperforms several prefetching algorithms from the whole system´s perspective.