Scholarlink

<p>The objective of this research is to find the various flows of optimal path in the multi-path transmission control protocol (MPTCP), which is a key component scheduler that determines the path taken by a data packet. However, the performance of the previous methods was lower because they did not take into account all the properties. In this paper, a multi-path scheduler based on Deep Q-Reinforcement Learning (DQRL) named MPTCP-DQRL is proposed to find a better path for a given application, depending on the present network environment. Here, the proposed model manages path utilization between a number of connections by combining both the MPTCP transmission control model and QRL. Thus, the proposed model increases the throughput, reduces energy consumption and latency. The proposed technique’s effectiveness is assessed by measuring and comparing various performance metrics with existing techniques. The comparison results show that the proposed technique outperforms existing methods in selecting the optimal path for data packet transmission, achieving a throughput of 30 Mbps, an energy consumption ratio of 4.71%, and a delay of 8.74% for 500 nodes.</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

MPTCP-DQRL Based Learning Protocol for Finding Various Flows of Optimal Path in MPTCP Scheduling System

  • 期刊论文

  • 发表日期：

    2025年2月1日

查看全文
  • Sanjit Kumar Barik,
  • Bharat J. R. Sahu

摘要

The objective of this research is to find the various flows of optimal path in the multi-path transmission control protocol (MPTCP), which is a key component scheduler that determines the path taken by a data packet. However, the performance of the previous methods was lower because they did not take into account all the properties. In this paper, a multi-path scheduler based on Deep Q-Reinforcement Learning (DQRL) named MPTCP-DQRL is proposed to find a better path for a given application, depending on the present network environment. Here, the proposed model manages path utilization between a number of connections by combining both the MPTCP transmission control model and QRL. Thus, the proposed model increases the throughput, reduces energy consumption and latency. The proposed technique’s effectiveness is assessed by measuring and comparing various performance metrics with existing techniques. The comparison results show that the proposed technique outperforms existing methods in selecting the optimal path for data packet transmission, achieving a throughput of 30 Mbps, an energy consumption ratio of 4.71%, and a delay of 8.74% for 500 nodes.