Scholarlink

<p>The evolution of Internet of Things (IoT) technologies plays a crucial role in advancing smart cities and industrial applications. One of the emerging technologies facilitating sensing and data transfer processes in IoT applications is Wireless Sensor Networks (WSNs). WSN nodes typically operate with constrained battery capacities, making energy efficiency a critical concern for clustering and routing. Furthermore, WSNs are vulnerable because they are deployed in unmonitored and harsh environments. Timely identification of faults in such conditions is challenging and can significantly affect network stability, leading to inaccurate sensed data within the sensing field. Numerous optimization techniques have been proposed to enhance resilience and fault tolerance against malicious attacks, node failures, and communication disruptions. However, many existing algorithms lack adequate mechanisms to protect against malicious threats, leading to unreliable and insecure network communications. To address these challenges, this work introduces a low-power cluster-based routing protocol featuring a robust fault detection system for WSNs. The protocol utilizes the fuzzy logic-enhanced Improved Whale Optimization Algorithm (IWOA) for Cluster Head (CH) selection. Furthermore, the Adaptive Elephant Herding Optimization (AEHO) technique determines optimal routes for efficient inter-cluster data transmission, thereby promoting energy efficiency within the WSN. Finally, the CH deploys a robust fault detection system using a Deep Feed-Forward Neural Network (DFFNN) to identify erroneous data within the network and ensure efficient data transmission in cluster-based WSNs. Experimental results show that it performs better across a variety of performance metrics. The simulation results of the suggested approach achieved better Quality of Service (QoS) parameters, such as the dispersion value (0.3401), cost function (601.09), network lifetime (4100 rounds), and prediction accuracy (99.5%). Compared to the Quadrature-LEACH (Q-LEACH), Enhanced Clustering Hierarchy (ECH), Adaptive Energy Efficient Clustering (AEEC), and Mobile Sink based Fault Diagnosis Scheme (MSFDS) protocols, the proposed protocol outperforms them.</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

A hybrid optimization and machine learning based energy-efficient clustering algorithm with self-diagnosis data fault detection and prediction for WSN-IoT application

  • 期刊论文

  • 发表日期：

    2025年1月6日

查看全文
  • N. Nathiya,
  • C. Rajan,
  • K. Geetha

摘要

The evolution of Internet of Things (IoT) technologies plays a crucial role in advancing smart cities and industrial applications. One of the emerging technologies facilitating sensing and data transfer processes in IoT applications is Wireless Sensor Networks (WSNs). WSN nodes typically operate with constrained battery capacities, making energy efficiency a critical concern for clustering and routing. Furthermore, WSNs are vulnerable because they are deployed in unmonitored and harsh environments. Timely identification of faults in such conditions is challenging and can significantly affect network stability, leading to inaccurate sensed data within the sensing field. Numerous optimization techniques have been proposed to enhance resilience and fault tolerance against malicious attacks, node failures, and communication disruptions. However, many existing algorithms lack adequate mechanisms to protect against malicious threats, leading to unreliable and insecure network communications. To address these challenges, this work introduces a low-power cluster-based routing protocol featuring a robust fault detection system for WSNs. The protocol utilizes the fuzzy logic-enhanced Improved Whale Optimization Algorithm (IWOA) for Cluster Head (CH) selection. Furthermore, the Adaptive Elephant Herding Optimization (AEHO) technique determines optimal routes for efficient inter-cluster data transmission, thereby promoting energy efficiency within the WSN. Finally, the CH deploys a robust fault detection system using a Deep Feed-Forward Neural Network (DFFNN) to identify erroneous data within the network and ensure efficient data transmission in cluster-based WSNs. Experimental results show that it performs better across a variety of performance metrics. The simulation results of the suggested approach achieved better Quality of Service (QoS) parameters, such as the dispersion value (0.3401), cost function (601.09), network lifetime (4100 rounds), and prediction accuracy (99.5%). Compared to the Quadrature-LEACH (Q-LEACH), Enhanced Clustering Hierarchy (ECH), Adaptive Energy Efficient Clustering (AEEC), and Mobile Sink based Fault Diagnosis Scheme (MSFDS) protocols, the proposed protocol outperforms them.