Research on SDN Load Balancing based on Ant Colony Optimization Algorithm

摘要

Network load balancing is one of the important methods to manage resources. Through network load balancing technology, both system resource utilization and system performance can be improved. This paper presents a Joint Load Balancing Algorithm based on Software Defined Network. The algorithm proposed in this paper not only considers the load balancing of data links, but also considers the server load balancing. When data flows reaches the data center, the Joint Load Balancing Algorithm first uses a Dynamic Server Load Balancing Algorithm to select the optimal server and then uses Hybrid Routing Algorithm to select the best path to the selected server. The Hybrid Routing Algorithm combines the Shortest Path Algorithm with the Ant Colony System. Finally, the Joint Load Balancing Algorithm proposed in this paper is implemented as a load balancing module in the OpenDayLight controller and simulated by Mininet. The experimental results prove the effectiveness of the algorithm. Compared with the shortest path polling algorithm and the shortest path random algorithm, the algorithm proposed in this paper makes the network performance significantly improved, and achieves higher network throughput and lower delay.