Scholarlink

<p>The Internet of Things (IoT) networks generate substantial traffic due to the large number of connected devices and their constant data communication. To efficiently monitor IoT networks, Software-Defined Networking (SDN) is used for its programmability, flexibility, and scalability. However, managing the traffic flow from IoT devices to controllers via switches presents a major challenge, particularly because of the heterogeneity in the switch load distribution. One way to deal with this issue is by optimizing the controller placement to ensure efficient communication between IoT devices and SDN controllers, known as the <i>Controller Placement Problem</i>. In this paper, we propose a new Particle Swarm Optimization (PSO)-based approach that mainly considers the heterogeneity in switch loads to strategically determine controller placement. It efficiently explores the solution space and converges to optimal controller placement by balancing the trade-off between switch to controller latency and reliability, aiming to optimize both. The extensive simulation results show that our <i>PSO-HSL</i> approach as an average latency gap of 6.15%, a flow rate improvement of 5.38%, and a reliability enhancement of 1.95% compared to the baseline method (<i>Opt-HSL</i>; the exact method that finds the optimal solution through an exhaustive exploration of all possible combinations). Furthermore, it notably outperforms both <i>K-means-BC</i> and <i>Louvain-BC</i> approaches, reducing latency by up to 60%, improving flow rate by 42.84%, and enhancing reliability by up to 15.55%, while maintaining near-optimal performance with significantly lower execution time.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

PSO-HSL: A controller placement strategy using PSO with heterogeneous switch load awareness in SD-IoT

  • 期刊论文

  • 发表日期：

    2025年10月9日

查看全文
  • Raid Boudi,
  • Mohammed Lalou,
  • Nardjes Bouchemal

摘要

The Internet of Things (IoT) networks generate substantial traffic due to the large number of connected devices and their constant data communication. To efficiently monitor IoT networks, Software-Defined Networking (SDN) is used for its programmability, flexibility, and scalability. However, managing the traffic flow from IoT devices to controllers via switches presents a major challenge, particularly because of the heterogeneity in the switch load distribution. One way to deal with this issue is by optimizing the controller placement to ensure efficient communication between IoT devices and SDN controllers, known as the Controller Placement Problem. In this paper, we propose a new Particle Swarm Optimization (PSO)-based approach that mainly considers the heterogeneity in switch loads to strategically determine controller placement. It efficiently explores the solution space and converges to optimal controller placement by balancing the trade-off between switch to controller latency and reliability, aiming to optimize both. The extensive simulation results show that our PSO-HSL approach as an average latency gap of 6.15%, a flow rate improvement of 5.38%, and a reliability enhancement of 1.95% compared to the baseline method (Opt-HSL; the exact method that finds the optimal solution through an exhaustive exploration of all possible combinations). Furthermore, it notably outperforms both K-means-BC and Louvain-BC approaches, reducing latency by up to 60%, improving flow rate by 42.84%, and enhancing reliability by up to 15.55%, while maintaining near-optimal performance with significantly lower execution time.