Comprehensive QoS Analysis for Enhancing Performance Evaluation in Software Defined Networking

Abstract

Software Defined Networking (SDN) introduces a paradigm shift in network architecture by cen tralizing control and configuration through software-based applications, separating the control plane from the data plane. This separation enhances network flexibility, enables automation, and improves programma bility, offering superior traffic management, resource allocation, and Quality of Service (QoS) compared to conventional network designs. This study focuses on evaluating SDN performance through four critical QoS parameters: Response Time, Throughput, Bandwidth Isolation, and Queue Impact. These metrics are vital for assessing network responsiveness, data handling capacity, resource isolation, and the effects of queue configurations. Notably, this research addresses bandwidth isolation and queue impact areas that have re ceived limited attention in prior studies making it a significant contribution to the field. To conduct this analysis, we implemented a single Linear Topology using the Mininet network emulator on a Linux platform. The network was configured using Python, with performance tests involving ICMP (ping) and TCP traffic under varying bandwidth and queue conditions. Experiments spanned multiple operational scenarios to capture detailed performance trends. The results highlight how adjustments to bandwidth and queue size influence the selected QoS parameters. These findings offer practical insights for network engineers and SDN practitioners, enabling them to optimize network performance and design more effective QoS-aware SDN deployments.