Scholarlink

<p>One of the key challenges in Software-Defined Networking (SDN) with multi-controller architectures is optimizing the controllers placement, which significantly impacts in network scalability. Improving network scalability is usually achieved by reducing network latency. To minimize network latency, a modified genetic algorithm called Wisdom of Artificial Crowds (WOAC) is used. The WOAC approach gives the best solution by merging individual solutions from the population. This optimized solution is used to place the controllers in precise locations to calculate the system reliability in SDN. The proposed method, Wisdom of Artificial Crowds based Controller Placement for minimum latency with higher reliability (ReLate) is introduced. ReLate is evaluated using standard network topologies namely Internet2 OS3E and Iris network. The comparison results show that ReLate outperforms existing state-of-the-art controller placement algorithms, namely K-Means, MDPC, and RCP-GWO. Simulation results also indicate that a 20.5%, 17.2%, and 17.2% improvement in average latency when compared to K-Means, MDPC, and RCP-GWO with four controllers respectively. In addition, 6.4%, 5.2%, and 1.3% increase in network reliability when compared to K-Means, MDPC, and RCP-GWO with four controllers respectively.</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

Relate: latency-based reliable controller placement using wisdom of artificial crowds in SDN-WAN

  • 期刊论文

  • 发表日期：

    2025年6月28日

查看全文
  • Vidya Sagar Thalapala,
  • Anandhu Mohan,
  • Guravaiah Koppala

摘要

One of the key challenges in Software-Defined Networking (SDN) with multi-controller architectures is optimizing the controllers placement, which significantly impacts in network scalability. Improving network scalability is usually achieved by reducing network latency. To minimize network latency, a modified genetic algorithm called Wisdom of Artificial Crowds (WOAC) is used. The WOAC approach gives the best solution by merging individual solutions from the population. This optimized solution is used to place the controllers in precise locations to calculate the system reliability in SDN. The proposed method, Wisdom of Artificial Crowds based Controller Placement for minimum latency with higher reliability (ReLate) is introduced. ReLate is evaluated using standard network topologies namely Internet2 OS3E and Iris network. The comparison results show that ReLate outperforms existing state-of-the-art controller placement algorithms, namely K-Means, MDPC, and RCP-GWO. Simulation results also indicate that a 20.5%, 17.2%, and 17.2% improvement in average latency when compared to K-Means, MDPC, and RCP-GWO with four controllers respectively. In addition, 6.4%, 5.2%, and 1.3% increase in network reliability when compared to K-Means, MDPC, and RCP-GWO with four controllers respectively.