Scholarlink

<p>In the evolution of contemporary computing paradigms, edge computing has emerged as a pivotal technology to handle delay-sensitive tasks and manage computational cost. This paper addresses the challenges of ensuring efficient task offloading and system stability in heterogeneous computing resource environments and aims to minimize system cost while maximizing the number of successfully completed tasks. The task offloading problem is transformed into a many-to-one matching game with externalities. To address this, we propose a distributed Multi-stage Adaptive Deferred Acceptance (MA-DA) algorithm that enables a stable and Pareto-optimal assignment of tasks to edge computing nodes (ECNs). The algorithm integrates task laxity time into the task offloading strategy, thus determining a reasonable task execution sequence and ensuring the prioritized completion of delay-sensitive tasks. The experimental results demonstrate that, compared to the Least Connection algorithm, the proposed MA-DA algorithm is able to reduce the system energy consumption by about 20.9% under low system load conditions and increase the task completion rates by about 29.1% under high system load conditions.</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

Stable matching based efficient task offloading in heterogeneous edge environment

  • 期刊论文

  • 发表日期：

    2026年7月29日

查看全文
  • Shoulu Hou,
  • Zhenquan Liu,
  • Yaru Zhao,
  • Qiang Tong,
  • Xiulei Liu

摘要

In the evolution of contemporary computing paradigms, edge computing has emerged as a pivotal technology to handle delay-sensitive tasks and manage computational cost. This paper addresses the challenges of ensuring efficient task offloading and system stability in heterogeneous computing resource environments and aims to minimize system cost while maximizing the number of successfully completed tasks. The task offloading problem is transformed into a many-to-one matching game with externalities. To address this, we propose a distributed Multi-stage Adaptive Deferred Acceptance (MA-DA) algorithm that enables a stable and Pareto-optimal assignment of tasks to edge computing nodes (ECNs). The algorithm integrates task laxity time into the task offloading strategy, thus determining a reasonable task execution sequence and ensuring the prioritized completion of delay-sensitive tasks. The experimental results demonstrate that, compared to the Least Connection algorithm, the proposed MA-DA algorithm is able to reduce the system energy consumption by about 20.9% under low system load conditions and increase the task completion rates by about 29.1% under high system load conditions.