Scholarlink

<p>In the Industrial Internet of Things (IIoT), Time-Sensitive Software-Defined Networking (TSSDN) enables deterministic Quality-of-Service (QoS) through traffic shaping and scheduling. However, most existing scheduling approaches assume fully known traffic profiles and fail to handle unpredictable burst traffic, which competes for bandwidth and degrades the QoS of registered flows. To address this challenge, this paper proposes a preserve-and-invoke method that supports ultra-low-latency burst transmission while maintaining QoS guarantees for the registered traffic. Specifically, we proactively reserve bandwidth during registered traffic scheduling by optimizing traffic distribution, thereby preparing spare resources for potential burst arrivals without violating registered QoS. When burst traffic occurs, we schedule it using an as-soon-as-possible strategy to minimize its end-to-end latency while limiting disruption to existing schedules. We further invoke the preserved bandwidth through guided rescheduling to compensate affected registered flows, thereby restoring their QoS guarantees. Experimental results demonstrate that the proposed method improves burst schedulability by up to 25% and increases QoS maintenance ratio for registered traffic by 17% compared with state-of-the-art approaches.</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

Burst-Tolerance Design in Time-Sensitive Software-Defined Networking: A Preserve-and-Invoke Method

  • 期刊论文

  • 发表日期：

    2026年3月26日

查看全文
  • Liuqing Yang,
  • Heng Ding

摘要

In the Industrial Internet of Things (IIoT), Time-Sensitive Software-Defined Networking (TSSDN) enables deterministic Quality-of-Service (QoS) through traffic shaping and scheduling. However, most existing scheduling approaches assume fully known traffic profiles and fail to handle unpredictable burst traffic, which competes for bandwidth and degrades the QoS of registered flows. To address this challenge, this paper proposes a preserve-and-invoke method that supports ultra-low-latency burst transmission while maintaining QoS guarantees for the registered traffic. Specifically, we proactively reserve bandwidth during registered traffic scheduling by optimizing traffic distribution, thereby preparing spare resources for potential burst arrivals without violating registered QoS. When burst traffic occurs, we schedule it using an as-soon-as-possible strategy to minimize its end-to-end latency while limiting disruption to existing schedules. We further invoke the preserved bandwidth through guided rescheduling to compensate affected registered flows, thereby restoring their QoS guarantees. Experimental results demonstrate that the proposed method improves burst schedulability by up to 25% and increases QoS maintenance ratio for registered traffic by 17% compared with state-of-the-art approaches.