A Multi-agent Brokering Approach and Jumper Firefly Algorithm for Job Scheduling in Cloud Computing

With the advancements in Science and Technology, capabilities for handling complex problems increase through optimal usage and management of computing power. Resources being distributed are reality, identifying and accumulating required resources for solving problem is a difficult task. In order to overcome this burden, the recent trend is to effectively use cloud computing thereby supporting resource sharing. The goal is to aggregate idle network, processing resources such as CPU cycles, and storage spaces to facilitate optimal utilization and computing. For the above, we need an effective method for identifying service providers based on all needs of any job for successful execution. The assignments and scheduling of the jobs should result in efficient way to solve the problem and promote optimal use of resource. This project proposes to address the above problem using Multi Agent Brokering Approach for the identification of service providers in a cloud environment. The Multi Agent Approach can ensure that the agents can specialize in identification of service provider for scheduling jobs. Jumper firefly algorithm used for reducing the make span time by its status table, records the behaviour of each firefly in detail as well. Also the proposed algorithm makes weaker ones to jump to new position to attain high probability. This leads to increase the performance better in finding the optimal solution. The experimental result shows that the jumper firefly mechanism is more efficient in terms of execution time (make span) than standard firefly algorithm and other heuristic methods.