Scholarlink

<p>Fog computing extends cloud services to the network edge, providing a more efficient and responsive infrastructure for time-critical IoT applications. However, improper service placement on fog nodes can lead to increased power consumption, higher latency, and diminished system reliability. To address these challenges, this paper presents a multi-objective, fault-tolerant service placement framework tailored for fog computing environments. A novel <i>mega node</i> strategy is introduced to minimize power consumption and improve fault tolerance by restricting the number of fog nodes utilized per application within a logically clustered mega node. Furthermore, an enhanced optimization algorithm—named the Improved Multi-Objective Artificial Hummingbird Algorithm (I-MOAHA)—is proposed to efficiently solve the service placement problem. Building upon the Multi-Objective Artificial Hummingbird Algorithm (MOAHA), the I-MOAHA integrates opposition-based learning, advanced territorial foraging strategies, and enhanced differential evolution mechanisms to mitigate premature convergence and enhance exploration capabilities. The proposed algorithm is evaluated using standard benchmark functions and the iFogSim2 simulator. Experimental results demonstrate that I-MOAHA significantly improves upon the baseline MOAHA algorithm, achieving a 14.16% reduction in power consumption and a 10% decrease in service latency.</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

Toward efficient and reliable fog computing: a mega node-based service placement strategy with improved metaheuristic optimization

  • 期刊论文

  • 发表日期：

    2026年7月24日

查看全文
  • Hamidreza Khaksar,
  • Reihaneh Khorsand

摘要

Fog computing extends cloud services to the network edge, providing a more efficient and responsive infrastructure for time-critical IoT applications. However, improper service placement on fog nodes can lead to increased power consumption, higher latency, and diminished system reliability. To address these challenges, this paper presents a multi-objective, fault-tolerant service placement framework tailored for fog computing environments. A novel mega node strategy is introduced to minimize power consumption and improve fault tolerance by restricting the number of fog nodes utilized per application within a logically clustered mega node. Furthermore, an enhanced optimization algorithm—named the Improved Multi-Objective Artificial Hummingbird Algorithm (I-MOAHA)—is proposed to efficiently solve the service placement problem. Building upon the Multi-Objective Artificial Hummingbird Algorithm (MOAHA), the I-MOAHA integrates opposition-based learning, advanced territorial foraging strategies, and enhanced differential evolution mechanisms to mitigate premature convergence and enhance exploration capabilities. The proposed algorithm is evaluated using standard benchmark functions and the iFogSim2 simulator. Experimental results demonstrate that I-MOAHA significantly improves upon the baseline MOAHA algorithm, achieving a 14.16% reduction in power consumption and a 10% decrease in service latency.