Scholarlink

<p>With the rapid growth of the Internet of Things, vast amounts of data are generated, creating challenges for effective management and processing. Limited processing resources, network congestion, and fluctuating workloads-particularly during peak times-can significantly increase response times or lead to data loss. To address these challenges, efficient task offloading strategies are essential for ensuring timely responses and optimal resource utilization in fog-based environments. This paper proposes a network traffic-aware decentralized task offloading method that uses Decision Tree intelligence to enable fog nodes to make adaptive offloading decisions. By considering both processing workload and network traffic conditions, the proposed approach determines whether to execute tasks locally or offload them, thereby dynamically distributing workloads across the network. Unlike traditional methods that directly rely on latency or communication delay, our method incorporates network traffic as a holistic parameter, inherently reflecting system congestion and queue delays. Experimental evaluations using the iFogSim simulator demonstrate that the proposed method outperforms existing approaches, and other AI-based models, under varying workload conditions, network bandwidth constraints, and system scalability. Results indicate a significant reduction in response time and improved resource efficiency, particularly under heavy load and network bottleneck scenarios.</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

Dynamic task offloading for IoT-Fog-Cloud systems: a network traffic-aware decision tree approach

  • 期刊论文

  • 发表日期：

    2025年3月13日

查看全文
  • Mohammad Zolghadri,
  • Parvaneh Asghari,
  • Seyed Ebrahim Dashti,
  • Alireza Hedayati

摘要

With the rapid growth of the Internet of Things, vast amounts of data are generated, creating challenges for effective management and processing. Limited processing resources, network congestion, and fluctuating workloads-particularly during peak times-can significantly increase response times or lead to data loss. To address these challenges, efficient task offloading strategies are essential for ensuring timely responses and optimal resource utilization in fog-based environments. This paper proposes a network traffic-aware decentralized task offloading method that uses Decision Tree intelligence to enable fog nodes to make adaptive offloading decisions. By considering both processing workload and network traffic conditions, the proposed approach determines whether to execute tasks locally or offload them, thereby dynamically distributing workloads across the network. Unlike traditional methods that directly rely on latency or communication delay, our method incorporates network traffic as a holistic parameter, inherently reflecting system congestion and queue delays. Experimental evaluations using the iFogSim simulator demonstrate that the proposed method outperforms existing approaches, and other AI-based models, under varying workload conditions, network bandwidth constraints, and system scalability. Results indicate a significant reduction in response time and improved resource efficiency, particularly under heavy load and network bottleneck scenarios.