Work

Using fuzzy control routing for dynamic load balancing over Software-Defined Networks

Year: 2025

Type: conference-paper

Abstract: The load-balancing and forwarding mechanisms of traditional software-defined networks (SDNs) often rely on static path selection.Conventional mechanisms lead to uneven resource utilization and bottlenecks under high-traffic conditions.Therefore, the existing methods require more flexibility in decision-making.This study introduces a loadbalancing algorithm based on fuzzy logic to address the difficulty of setting multivariable input thresholds.The algorithm uses fuzzy logic to combine the network parameters (queue length, link utilization, link delay, and packet loss rate) as fuzzy inputs.Converting the fuzzy input into a score is the key to achieving optimal traffic allocation in the mechanism.The proposed mechanism improves resource utilization and reduces bottlenecks and reliability in dynamic network-load situations.The results indicated that the proposed method achieved a higher throughput under high-load conditions.Moreover, the results maintained low packet loss and delay rates.Load balancing based on fuzzy logic provides an effective solution for SDN environments.

Source: Proceedings of International Conference on Artificial Life and Robotics

DOI: 10.5954/icarob.2025.os5-5

Authors: Wen-Shyang Hwang, Ming-Syuan Wu, Cheng-Han Lin, Yan-Jing Wu, Ming-Hua Cheng, Sian-Fong Huang

Institutions: National Kaohsiung University of Science and Technology, Fooyin University, Shih Chien University, Tzu Hui Institute of Technology

Language: English

Topic: Software-Defined Networks and 5G

Subfield: Computer Networks and Communications

Field: Computer Science

Domain: Physical Sciences

SDG: Industry, innovation and infrastructure

Open Access status: bronze

Funder: National Science and Technology Council