A centralized location-based job scheduling algorithm for inter-dependent jobs in mobile ad hoc computational grids

This study addresses the problem of job scheduling to inter-dependent jobs in mobile ad hoc computational grids. To maximize the utilization of shared computational resources and to improve application performance, an effective job scheduling algorithm plays a key role. Previously, numerous job scheduling algorithms have been proposed, but most of them are either targeted towards large scale infrastructure-based systems or they don't consider the characteristics of mobile devices and inter-job dependencies. As the performance of inter-dependent jobs is greatly affected by communication performance; therefore, to improve communication performance of inter-dependent jobs, we have proposed a centralized job scheduling algorithm that takes into account the location of nodes, inter-job dependencies and communication traffic among inter-dependent jobs to schedule them on closely located nodes. Simulation results demonstrated that our proposed algorithm performs well as compared to existing approaches and reduces the communication cost thus energy consumption.