Self-Socio Adaptive Reliable Particle Swarm Optimization Load Balancing in Software-Defined Networking

The ever-increasing heterogeneous connections and the demands of the users pose many new challenges to the network service providers to sustain by providing improved quality of service (QoS). Software-defined networking (SDN) is a game changer in networking by allowing user customization to enhance performance. With the advent of 5G and the increasing user requests, a massive volume of heterogeneous traffic is generated in the network, increasing load. Currently, the existing load balancing techniques lack efficiency in handling the load under unicontroller deployment. In addition, the network paths selected must also be reliable and optimal. We proposed the self-socio adaptive, reliable particle swarm optimization (SSAR-PSO) load balancing technique to address the issue of load balancing in the unicontroller deployment of SDN. In the proposed technique, the performance of the node itself, known as direct information, and the performance of the neighbouring nodes, known as indirect information, were considered to identify the reliable node to form an optimal path. Simulation results showed that the proposed technique outperforms the existing state-of-the-art techniques under TCP and UDP load in the following network performance metrics: latency, packet loss ratio, throughput, average round trip time, and bandwidth utilization ratio.