Towards Optimal Efficiencies in Software Defined Network SDN-Edge Cloud: Performance Evaluation of Load Balancing Algorithms

Software defined network (SDN) has brought advancements in the field of network management by isolating it from switch/routing operations. Hence, SDN has been integrated into other platforms such as edge and cloud computing. However, due to the limited computation-resources built into the controller, issues related to cognition and scalability may arise. This can lead to a degradation of Quality of Service (QoS), especially in data-sensitive applications. This paper explores a comparative analysis of four distinct load balancing algorithms which are greedy, round robin, weighted round robin, and resource based load balancing algorithm in Software-Defined Network (SDN)-Edge Cloud environment. The performance of each balancing algorithm is examined under different packet arrival rates. The objective of this study is to provide insights into the capabilities of an SDN controller to adapt to unpredictable data flows and efficiently perform load balancing in terms of QoS, specifically delay and packet overdue ratio in multi-tier controller system.