Software Defined Network Based Load Balancing in Cloud Computing

Abstract

Allocation and management of resources is one of the major challenges in the current era where the number of users are exponentially increasing. Hence, the main focus is to manage the dynamic workload in an equal manner so that no node will be overloaded. Load balancing is the technique that balances the load by equally distributing it over all nodes, thereby reducing the over-burdened or idle node as per its capability. The conventional system can be improved by taking benefits of Software Defined Network (SDN). SDN helps network administrators to have a centralized control over the network so that they can meet the business needs of the users easily. Current techniques for load balancing in the network requires many hardware resources for processing the network traffic. SDN based load balancing an emphasis on effective resource utilization to minimize request processing time. The proposed algorithm selects the least loaded server from a number of servers. It minimizes request processing time and maximizes the throughput as compared to Round Robin and Weighted Round Robin polices. Proposed algorithm is simulated on mininet simulator using Floodlight SDN based OpenFlow enabled open source controller.