Adaptive Path Selection Algorithm with Flow Classification for Software-Defined Networks

Software-Defined Networking (SDN) is a trending architecture that separates controller and forwarding planes. This improves network agility efficiency. The proliferation of the Internet Things devices has increased traffic flow volume its heterogeneity in contemporary networks. Since SDN flow-driven network, it requires corresponding rule for each flowtable. However, complicates rules update operation due to varied quality service requirements en-route behavior. Some flows are delay-sensitive while others long-lived with propensity consume buffers, thereby inflicting congestion delays on network. must be routed through path minimal delay, congestion-susceptible guided along route adequate capacity. Although several efforts were introduced over years efficiently based different QoS parameters, current selection techniques consider either link or switch during decisions. Incorporating composite metrics classification decisions not been adequately considered. paper proposes technique differentiate congestion-prone reroute them appropriate paths avoid loss. integrated into guide suitable class. Compared other works, proposed approach improved load ratio by 25%, throughput 35.6%, packet delivery 31.7%.