Optimizing SDN Controller Placement for Enhanced Performance and Scalability in Large-Scale Networks
Vishwanath Hiremath
Abstract
Software-Defined Networking (SDN) has emerged as a transformative paradigm that decouples the control and data planes, enabling centralized network management. However, the optimal placement of SDN controllers remains a critical challenge, particularly in largescale networks, as it directly impacts network performance, scalability, and fault tolerance. Inefficient placement strategies can lead to increased latency, network congestion, and uneven load distribution, affecting overall network reliability. This paper presents an optimized controller placement strategy that employs a hybrid approach combining Genetic Algorithms (GA) and Particle Swarm Optimization (PSO) to enhance network efficiency. The proposed method minimizes the average propagation delay, balances controller workloads, and improves network resilience against failures. A mathematical model is formulated to optimize controller distribution while considering constraints such as latency, load balancing, and fault tolerance. The approach is implemented and tested using Mininet and POX controllers, with experimental evaluations demonstrating its effectiveness over traditional placement techniques. Results indicate that the proposed hybrid optimization method significantly reduces latency, enhances load distribution, and improves network reliability compared to conventional clustering and heuristic-based approaches. The findings contribute to advancing SDN controller placement strategies, providing a scalable and efficient solution for modern largescale networks. Future work will explore adaptive placement mechanisms for dynamic network environments to further enhance performance and flexibility.