Scholarlink

<p>Self-healing at the network edge is vital for sustaining IoT deployments in settings where safety is measured by seconds and access by weeks. This paper presents an Edge AI-powered Self-Healing framework for IoT networks in harsh environments such as offshore rigs, mines, Arctic bases, and disaster zones. Harsh environments are defined as contexts with extreme temperatures (− 40&#xa0;°C to + 85&#xa0;°C), high vibration (&gt; 5&#xa0;g RMS), electromagnetic interference (&gt; 10&#xa0;V/m), and physical inaccessibility. The framework integrates a lightweight Distributed Fault Detection Network (DFDN) with Self-Healing Protocol (SHP) mechanisms, achieving 95.8% recovery rate with 12.4-min mean time to recovery. Evaluation on Arctic-IoT and Industrial-Harsh datasets demonstrates superior performance while maintaining computational efficiency for resource-constrained edge devices.</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

Edge AI for Self-Healing IoT Networks in Harsh Environments

  • 期刊论文

  • 发表日期：

    2026年5月24日

查看全文
  • Shweta Gode

摘要

Self-healing at the network edge is vital for sustaining IoT deployments in settings where safety is measured by seconds and access by weeks. This paper presents an Edge AI-powered Self-Healing framework for IoT networks in harsh environments such as offshore rigs, mines, Arctic bases, and disaster zones. Harsh environments are defined as contexts with extreme temperatures (− 40 °C to + 85 °C), high vibration (> 5 g RMS), electromagnetic interference (> 10 V/m), and physical inaccessibility. The framework integrates a lightweight Distributed Fault Detection Network (DFDN) with Self-Healing Protocol (SHP) mechanisms, achieving 95.8% recovery rate with 12.4-min mean time to recovery. Evaluation on Arctic-IoT and Industrial-Harsh datasets demonstrates superior performance while maintaining computational efficiency for resource-constrained edge devices.