A Comparative Study of the Application of Glowworm Swarm Optimization Algorithm with other Nature-Inspired Algorithms in the Network Load Balancing Problem

Abstract

Vast amounts of data are transferred through communication networks resulting in node congestion, which varies according to peak usage times. The Glowworm Swarm Optimization (GSO) algorithm is inspired by the rummaging and courtship behavior of glowworms. The glow intensity of glowworms is a measure of fitness that attracts other glowworms in its neighborhood. This work applies the GSO algorithm to the computer network congestion problem in order to lessen the network burden by shifting loads to the fittest neighborhood nodes, thereby enhancing network performance during peak traffic times, when the response of systems on the network would go down. The proposed solution aims to alleviate the burdened nodes, thereby improving the flow of traffic throughout the network, improving the users’ experience and productivity, and efficiency. In this paper, three swarm algorithms, namely Particle Swarm Optimization (PSO), Cuckoo Search (CK), and GSO have been employed to solve the network load balancing problem. The results produced by GSO show improvement of 71.17%, 74.14%, and 84.15% in networks consisting of 50, 100, and 200 nodes in peak hour load, while PSO shows 13.87%, 11.75%, and 23.72%, and CK 10.61%, 3.19%, and 6%. The results prove the superior performance of GSO.

Keywords

• network congestion
• load balancing
• gso
• throughput
• swarm intelligence