MFQ: Network Bandwidth Utilization among Data Centers on Software Defi ned Network Using Multilevel-Feedback Queue

Abstract

The growing complexity of network traffic patterns and congestion within data centers necessitate innovation in traditional network infrastructure. While advancements in storage, computing, and applications have progressed rapidly, networking has remained relatively stagnant, limiting the overall performance of data centers. Software-Defined Networking (SDN) emerges as a powerful solution by decoupling the control plane from the data plane, offering enhanced flexibility and efficiency in network management. This decoupling allows for centralized control and dynamic configuration, making SDN well-suited for modern, high-demand environments. As network architectures evolve, SDN is expected to enable more adaptive and context-sensitive routing protocols that can adjust to real-time conditions. However, challenges in optimizing resource utilization across data centers persist. This paper proposes a novel approach utilizing a multilevel-feedback queue mechanism to enable real-time traffic processing and dynamic scheduling, improving bandwidth utilization. Simulation results demonstrate the effectiveness of this method in optimizing traffic scheduling, highlighting its potential for addressing congestion issues and enhancing network performance in data centers.