Improved PSO based job scheduling algorithm for resource management in grid computing

Scheduling jobs to resources in grid computing is a complicated task due to the dynamic nature of resources. An efficient job scheduling algorithm is required to reduce the total time and cost of job execution and improve load balancing among resources in the network. The main problems in managing resources are hardware and software failures, jobs management downtime, etc. To solve this, the PSO is introduced. The PSO algorithm is based on a simplified model of social behaviour exhibited by the buzzing behaviour of insects, birds and fish. In this research, we propose a new algorithm for job scheduling, called improved particle swarm optimization (IPSO). The proposed algorithm generates a velocity vector that is used to point out that the direction of swarm movement and particle position are updated. Therefore, it refines and improves the efficiency of the execution, the research capacity of the global research, the accuracy of the solution and guarantees the load balancing of the programming of the grid activities. Consequently, the proposed work has been simulated with the help of the OptorSim simulator and it has been shown that our proposed algorithm provides an effective solution for planning the resources over grid scheduling network.