BALANCE THE NETWORK TRAFFIC LOAD IN SOFTWARE-DEFINED NETWORKING

Abstract

Traditional network architecture is too rigid to be scalable and efficient in traffic distribution. Software-defined networking (SDN) addresses these issues by separating the control and data planes to improve network management. However, SDN faces challenges, such as controller bottlenecks and traffic congestion. To overcome these, load-balancing algorithms are implemented. This project examines current load-balancing approaches and proposes a solution that distributes load evenly between optimal paths. The employed OpenFlow protocol allows checking real-time network distribution and testing the solution’s efficiency. The study reviews existing research on various models for enhancing traffic distribution in SDN. Following a comprehensive analysis of traditional, heuristic, and AI-driven techniques, the findings reaffirm that load balancing is an efficient and scalable solution for SDN routing. Mininet is used for network emulation, Floodlight serves as the controller, and the REST API is queried to check real-time statistics. Initial results show that round-robin load balancing slightly improves network distribution performance. However, it lacks adaptability for complex topologies. Further work explores the integration of blockchain to enhance security and transparency.