Scholarlink

<p>Industrial Internet of Things (IIoT) gateways must balance energy consumption, queueing delay, packet loss, and effective goodput while operating under shared communication and hardware constraints. Manual configuration of gateway parameters becomes increasingly difficult because continuous and integer design variables interact through nonlinear system behaviour. This work formulates gateway configuration as a constrained four-objective mixed-integer optimisation problem using analytical surrogate models for queueing behaviour and per-epoch energy consumption subject to delay, packet-loss, goodput, airtime, and operating constraints. A hybrid multi-objective Harris Hawks optimiser (HybMOHHO) is proposed by combining Harris Hawks Optimization (HHO) exploration with NSGA-II environmental selection and type-aware mutation to accommodate mixed continuous-integer decision variables. The proposed framework is evaluated against NSGA-II, MOPSO, and MOHHO using 30 independent runs and assessed with Hypervolume (HV), Inverted Generational Distance (IGD), and Spread quality indicators. Across the benchmark experiments, HybMOHHO obtained a mean HV of 0.9694 and a mean IGD of 0.0542 while maintaining consistent feasible archive retention across independent runs. Friedman omnibus testing indicated statistically significant differences among the evaluated optimisation methods (<InlineEquation ID="IEq1"><EquationSource Format="TEX">\(p &lt; 10^{-4}\)</EquationSource></InlineEquation>). These results suggest that the proposed hybrid framework provides a competitive approach for constrained gateway-level IIoT multi-objective optimisation.</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

Hybrid multi-objective framework for constrained mixed-integer optimisation of industrial IoT gateway configuration

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年8月1日

查看全文
  • Nishant Vats,
  • Rajesh P. Barnwal

摘要

Industrial Internet of Things (IIoT) gateways must balance energy consumption, queueing delay, packet loss, and effective goodput while operating under shared communication and hardware constraints. Manual configuration of gateway parameters becomes increasingly difficult because continuous and integer design variables interact through nonlinear system behaviour. This work formulates gateway configuration as a constrained four-objective mixed-integer optimisation problem using analytical surrogate models for queueing behaviour and per-epoch energy consumption subject to delay, packet-loss, goodput, airtime, and operating constraints. A hybrid multi-objective Harris Hawks optimiser (HybMOHHO) is proposed by combining Harris Hawks Optimization (HHO) exploration with NSGA-II environmental selection and type-aware mutation to accommodate mixed continuous-integer decision variables. The proposed framework is evaluated against NSGA-II, MOPSO, and MOHHO using 30 independent runs and assessed with Hypervolume (HV), Inverted Generational Distance (IGD), and Spread quality indicators. Across the benchmark experiments, HybMOHHO obtained a mean HV of 0.9694 and a mean IGD of 0.0542 while maintaining consistent feasible archive retention across independent runs. Friedman omnibus testing indicated statistically significant differences among the evaluated optimisation methods (\(p < 10^{-4}\)). These results suggest that the proposed hybrid framework provides a competitive approach for constrained gateway-level IIoT multi-objective optimisation.