Efficient management of large-scale trajectory data for mobile objects is a critical and challenging task. Existing systems struggle to handle the high volume of real-time trajectory messages in mobile edge scenarios, where high data transmission latency and overhead impede efficiency and real-time data management. Additionally, support for fine-grained representation and efficient query processing of trajectory data is often inadequate. This paper introduces MCTM, a multi-chord distributed system that efficiently stores and manages trajectory data on edge nodes, reducing network latency and overhead. MCTM presents a fine-grained temporal index (TI), an Index Cache, and a spatio-temporal index (TSI), significantly improving query efficiency and accuracy. In various query scenarios, MCTM reduces the number of retrievals by 56.5% at most, and its query speed exceeds the baselines by up to 50.5% to 6.32 times.

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

MCTM: Multi-chord Distributed System for Efficient Trajectory Data Management in Mobile Edge Computing

  • Yucheng Tao,
  • Haopeng Chen,
  • Zihong Lin,
  • Jiahao Xu,
  • Xiaojian Gao,
  • Jinghao Wang,
  • Yan Jiao,
  • Yongming Xu

摘要

Efficient management of large-scale trajectory data for mobile objects is a critical and challenging task. Existing systems struggle to handle the high volume of real-time trajectory messages in mobile edge scenarios, where high data transmission latency and overhead impede efficiency and real-time data management. Additionally, support for fine-grained representation and efficient query processing of trajectory data is often inadequate. This paper introduces MCTM, a multi-chord distributed system that efficiently stores and manages trajectory data on edge nodes, reducing network latency and overhead. MCTM presents a fine-grained temporal index (TI), an Index Cache, and a spatio-temporal index (TSI), significantly improving query efficiency and accuracy. In various query scenarios, MCTM reduces the number of retrievals by 56.5% at most, and its query speed exceeds the baselines by up to 50.5% to 6.32 times.