Fault tolerance technique using replication and checkpoint in mobile grid

Mobile grid computing is a computing environment which incorporates mobile devices to an existing grid environment and supports user mobility. As this environment has inherent characteristic, that is, not stable, methodologies to cope with the reliability issue are required. We apply replication and checkpoint techniques for task fault tolerance to a SimGrid simulator in mobile environment and show the results in this paper. The results demonstrate that using a fault tolerance scheme can increase the completion rate of tasks and the best solution for fault tolerance in mobile grid computing depends on network operational situations including the number of hosts and tasks. The contribution of this paper is the use of real-life trace data to simulate fault tolerance and comparison of fault tolerance schemes in a mobile grid computing.