Scholarlink

<p>Large-scale Internet of Things (IoT) service composition is a computationally intensive, NP-complete problem characterized by a massive combinatorial search space, rendering it intractable for conventional systems and a prime candidate for High-Performance Computing (HPC). Achieving an effective IoT service composition remains a challenge due to the need to jointly address several conflicting objectives. First, compositions must satisfy diverse Quality of Service (QoS) requirements, including response time, availability, throughput, and cost. Second, it is essential to minimize energy consumption to prolong device battery life and support the principles of Green IoT. Third, the reliability and stability of composed services must be ensured by reducing QoS fluctuations, which often result from the highly dynamic and unpredictable nature of IoT environments. These conflicting objectives, coupled with the real-time demands of applications in intelligent transportation and industry, necessitate scalable and high-throughput solutions. These challenges are rarely simultaneously addressed in existing approaches, which motivates the need for a unified and efficient solution. In light of this shortcoming, we propose MOMRFO-SC, a Multi-Objective Manta Ray Foraging Optimization approach for large-scale IoT Service Composition leveraging swarm intelligence principles. It combines a preselection phase based on the Compromise Ratio Method (CRM) which filters low-quality candidate services to reduce the search space with a Manta Ray Foraging Optimization (MRFO) algorithm that generates high-quality compositions. MOMRFO-SC is designed to efficiently handle large-scale scenarios in simulated environments, and its inherent scalability suggests that future extensions could benefit from parallel or distributed infrastructures to further enhance the real-time capabilities in dynamic IoT deployments. Experimental results show that MOMRFO-SC effectively balances the trade-offs between QoS satisfaction, energy efficiency, and QoS fluctuation minimization.</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

MOMRFO-SC: a multi-objective framework for large-scale IoT service composition

  • 期刊论文

  • 发表日期：

    2025年12月12日

查看全文
  • Macilia Boukhama,
  • Zoubeyr Farah,
  • Lynda Alkama,
  • Mohamed-Lamine Messai

摘要

Large-scale Internet of Things (IoT) service composition is a computationally intensive, NP-complete problem characterized by a massive combinatorial search space, rendering it intractable for conventional systems and a prime candidate for High-Performance Computing (HPC). Achieving an effective IoT service composition remains a challenge due to the need to jointly address several conflicting objectives. First, compositions must satisfy diverse Quality of Service (QoS) requirements, including response time, availability, throughput, and cost. Second, it is essential to minimize energy consumption to prolong device battery life and support the principles of Green IoT. Third, the reliability and stability of composed services must be ensured by reducing QoS fluctuations, which often result from the highly dynamic and unpredictable nature of IoT environments. These conflicting objectives, coupled with the real-time demands of applications in intelligent transportation and industry, necessitate scalable and high-throughput solutions. These challenges are rarely simultaneously addressed in existing approaches, which motivates the need for a unified and efficient solution. In light of this shortcoming, we propose MOMRFO-SC, a Multi-Objective Manta Ray Foraging Optimization approach for large-scale IoT Service Composition leveraging swarm intelligence principles. It combines a preselection phase based on the Compromise Ratio Method (CRM) which filters low-quality candidate services to reduce the search space with a Manta Ray Foraging Optimization (MRFO) algorithm that generates high-quality compositions. MOMRFO-SC is designed to efficiently handle large-scale scenarios in simulated environments, and its inherent scalability suggests that future extensions could benefit from parallel or distributed infrastructures to further enhance the real-time capabilities in dynamic IoT deployments. Experimental results show that MOMRFO-SC effectively balances the trade-offs between QoS satisfaction, energy efficiency, and QoS fluctuation minimization.