Scholarlink

As train communication networks evolve toward greater intelligence and informatization, traditional real-time Ethernet is increasingly unable to meet operational requirements due to issues such as latency jitter and packet loss. To address this challenge, Time-Sensitive Networking (TSN) technology has been introduced into train communication nerworks. In this study, a multi-queue heuristic scheduling algorithm is proposed based on the modeling of high-speed train communication networks. The algorithm aims to enhance schedulability by utilizing multiple data flow queues while significantly reducing scheduling time. Finally, the algorithm is validated using the OMNeT++ simulation software. Experimental results demonstrate a substantial improvement in train communication network performance, with total scheduling latency reduced by more than half on average compared to the baseline algorithms.

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

A Heuristic Multi-Queue Scheduling Algorithm for Time-Sensitive Information Acquisition in Train Sensing Systems

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Guodong Ma,
  • Jianan Cui,
  • Xuechao Wu,
  • Yuexian Wang,
  • Jiaqi Yang,
  • Mingyu Sun

摘要

As train communication networks evolve toward greater intelligence and informatization, traditional real-time Ethernet is increasingly unable to meet operational requirements due to issues such as latency jitter and packet loss. To address this challenge, Time-Sensitive Networking (TSN) technology has been introduced into train communication nerworks. In this study, a multi-queue heuristic scheduling algorithm is proposed based on the modeling of high-speed train communication networks. The algorithm aims to enhance schedulability by utilizing multiple data flow queues while significantly reducing scheduling time. Finally, the algorithm is validated using the OMNeT++ simulation software. Experimental results demonstrate a substantial improvement in train communication network performance, with total scheduling latency reduced by more than half on average compared to the baseline algorithms.