A Cognitive SDN-Driven Adaptive Load Balancing Architecture for Next-Generation Wireless Edge Networks

Abstract

Software-Defined Networking (SDN) has gained significant attention as a flexible solution for managing modern wireless edge networks. Despite its advantages, issues such as uneven traffic distribution and node congestion continue to affect overall network performance. In this work, an intelligent SDN-based adaptive load balancing framework is proposed to address these challenges. The approach continuously observes network conditions, including traffic load and available bandwidth, and makes real-time decisions to redirect data flows toward less congested paths. To evaluate its effectiveness, the proposed framework is implemented using Mininet-WiFi and compared with conventional techniques such as Round Robin and static load balancing. Performance is assessed using key metrics including throughput, latency, packet loss, and bandwidth utilization. The results indicate noticeable improvements, with higher throughput, reduced delay, and lower packet loss under varying traffic conditions. In particular, the system demonstrates better adaptability in handling dynamic network scenarios. Overall, the proposed framework provides a practical and efficient solution for improving resource utilization and Quality of Service in wireless edge environments, making it suitable for applications such as IoT systems, smart cities, and next-generation wireless networks.