Improving Response Time using Lightning based Biased Random Sampling (LBRS) for Load Balancing

Abstract

Abstract Cloud computing has rapidly emerged as a successful paradigm for providing IT infrastructure, resources and services on a pay-per-use basis over the past few years. The wider adoption of Cloud has led to the establishment of large scale data centers that consume excessive energy. Biased Random Sampling is an existing load balancing algorithm used to choose the load allocation node from the neighboring nodes to satisfy the user requests. Proposed work applies the behavior of lightning to Biased Random Sampling algorithm which is used for load balancing technique in cloud to achieve minimum response time. The work also computes the capacity of the machine to achieve high accuracy in calculating response time while assigning cloudlet received to VM. The work also includes obtaining resources from public cloud when load increases in the private cloud. Response time of proposed work is compared with existing biased random sampling and minimum response time is obtained for proposed load balancing technique. The response time obtained in LBRS is 82% less than BRS algorithm and 74% less than FCFS algorithm. The experimental results showed that the load is evenly distributed for LBRS algorithm and unevenly distributed for BRS algorithm.