A Novel Algorithm for Load Balancing in Mobile Cloud Networks: Multi-objective Optimization Approach

Cloud computing is a type of computing, because they share computing resources for better throughput. In cloud, virtualization technique is used to maximize the power of cloud computing. Mobile cloud computing architecture is based on geographical distributed cloud services. The resource allocation for the request is an important criterion for better performance of the computations. We need to efficiently handle these multiple cloud domains for better performance, quality of service and cost efficient job completion. In this paper, a novel method for efficiently allocating requirements to the available resources is proposed. Based on the prediction method a single cloud domain is selected from the multi cloud domain system. Then prioritize the nodes, based on the results obtained from the prediction approach. The priority ranks are arranged in descending order. The best resource allocation in the selected node (cloud domain) is determined by using genetic algorithm. Multi objective optimization method gives the constrained values for repetition of genetic approach. Using this priority based selection of nodes; unimportant node selection time is reduced, irrelevant node usage eliminated and the network congestion is reduced. Simulation results reveals how efficiently place the request to the appropriate resources on the nearest cloud domain and find out the best VM (virtual machine) for the individual request. Usage of rejected requests handling queues effectively handles the rejected requests. So the request rejection problem is eliminated.