Scholarlink

<p>Real-time water quality monitoring remains challenging due to the high latency of centralized laboratory analysis and the substantial computational payload of Machine Learning (ML) inference. While existing Internet of Things (IoT) frameworks deploy raw sensors, they frequently lack a standalone Edge calculation layer capable of operating completely independent of cloud infrastructure. This study proposes a mathematical-based Internet of Things Water Quality Index (IoT-WQI) architecture engineered specifically to act as an autonomous Fog/Edge Node for immediate first-line screening. The framework reduces continuous cloud dependency and network communication overhead by circumventing ML dependencies, replacing discrete look-up tables with mathematically continuous Gaussian and Polynomial Curve Fitting. This zero-training algorithmic optimization accelerates local computation by 83.55% over traditional linear interpolation and reduces network payload transmission bandwidth by 83.3%, while achieving an exceptional approximation fidelity (<InlineEquation ID="IEq1"><EquationSource Format="TEX">\(\:{R}^{2}&gt;0.996\)</EquationSource></InlineEquation>). Concurrently, a cross-sectoral Analytical Hierarchy Process (AHP) was structurally aligned with local river ecosystem profiles to extract robust contextual weights (CR = 0.02). Validated through a hardware-algorithmic stress test at the Troso River (Indonesia), the fault-tolerant topology yielded 6,779 validated records across a 6.49-hour deployment window via asynchronous queueing, achieving 98.1% of transmissions within 0–1&#xa0;s end-to-end latency. Ultimately, this framework seamlessly integrates edge computational efficiency with cloud distribution frameworks, providing a highly scalable architecture for real-time anomaly triage in resource-constrained IoT environments.</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

An edge-IoT water quality index (IoT-WQI) for first-line screening: accelerating computation via deterministic mathematical equations and grouped AHP

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月7日

查看全文
  • Wibowo Harry Sugiharto,
  • Agung Budi Prasetijo,
  • Heru Susanto

摘要

Real-time water quality monitoring remains challenging due to the high latency of centralized laboratory analysis and the substantial computational payload of Machine Learning (ML) inference. While existing Internet of Things (IoT) frameworks deploy raw sensors, they frequently lack a standalone Edge calculation layer capable of operating completely independent of cloud infrastructure. This study proposes a mathematical-based Internet of Things Water Quality Index (IoT-WQI) architecture engineered specifically to act as an autonomous Fog/Edge Node for immediate first-line screening. The framework reduces continuous cloud dependency and network communication overhead by circumventing ML dependencies, replacing discrete look-up tables with mathematically continuous Gaussian and Polynomial Curve Fitting. This zero-training algorithmic optimization accelerates local computation by 83.55% over traditional linear interpolation and reduces network payload transmission bandwidth by 83.3%, while achieving an exceptional approximation fidelity (\(\:{R}^{2}>0.996\)). Concurrently, a cross-sectoral Analytical Hierarchy Process (AHP) was structurally aligned with local river ecosystem profiles to extract robust contextual weights (CR = 0.02). Validated through a hardware-algorithmic stress test at the Troso River (Indonesia), the fault-tolerant topology yielded 6,779 validated records across a 6.49-hour deployment window via asynchronous queueing, achieving 98.1% of transmissions within 0–1 s end-to-end latency. Ultimately, this framework seamlessly integrates edge computational efficiency with cloud distribution frameworks, providing a highly scalable architecture for real-time anomaly triage in resource-constrained IoT environments.