Dual Threshold Load Balancing in SDN Environment Using Process Migration

Nowadays, Software Defined Networking (SDN) is a potential evolving technology and considered as an innovative network programming approach. In SDN, energy management and load balancing are considered to be current major research challenges. Previously, these issues are addressed in cloud environments. In this research, we propose a dual threshold load balancing (DTLB) method for SDN environment. The DTLB model works in four steps: process selection, destination physical machine selection, process-physical machine assignment and finally, process migration through docker container. Processes and physical machines are selected based on energy consumption, CPU utilization, and available memory and response time. The effectiveness of the DTLB model is achieved through using the matching technique which assigns a process to an appropriate physical machine and docker container. The docker container is lightweight and easy to develop while providing load balancing, resource sharing, and fault tolerance. The simulation shows that the DTLB model provides energy efficient load balancing with better throughput and faster response time than that of the traditional model without load balancing.