Scholarlink

<p>In real-world scenarios, sensor node failures in underwater wireless sensor networks (UWSNs) may arise due to internal or external factors, often influenced by cluster behavior. Identifying faults in sensor nodes is challenging, particularly when nodes are not explicitly classified as faulty or fault-free, making it difficult to communicate fault status across the network. Intermittent disturbances occur at irregular or regular intervals and can be categorized as random, stuck-at-zero, transient, or spike-like. Traditional fault detection methods relying on statistical measures such as mean, mode, median, or repeated testing with K-means clustering often fall short in sparse UWSNs due to the variability in erroneous data transmission caused by defective nodes.To address these limitations, this paper proposes a K-means-based Distributed Self-Intermittent Fault Diagnosis (KDSIFD) algorithm to detect sensor nodes exhibiting soft (intermittent) faults. Each sensor node determines its fault status based on self-assessment and comparison with other cluster nodes. The approach relies on clustering and the distribution of test results among neighboring sensors to improve fault identification. Additionally, the K-means algorithm is utilized to diagnose intermittent failures within the sensing and communication aspects of the cluster network.The proposed KDSIFD system is implemented and simulated in Python, and its performance is evaluated against existing intermittent fault detection techniques. The test results show that KDSIFD greatly increases detection accuracy and lowers the rates of false alarms and false positives, especially in sparse UWSN environments where traditional methods have a hard time finding faults.</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

Improving Fault Detection in Sparse Underwater Sensor Networks with K-MeansBased Distributed Algorithms

  • 期刊论文

  • 发表日期：

    2025年7月11日

查看全文
  • Ruchika Padhi,
  • Debendra Muduli,
  • Bhabani Sankar Gouda

摘要

In real-world scenarios, sensor node failures in underwater wireless sensor networks (UWSNs) may arise due to internal or external factors, often influenced by cluster behavior. Identifying faults in sensor nodes is challenging, particularly when nodes are not explicitly classified as faulty or fault-free, making it difficult to communicate fault status across the network. Intermittent disturbances occur at irregular or regular intervals and can be categorized as random, stuck-at-zero, transient, or spike-like. Traditional fault detection methods relying on statistical measures such as mean, mode, median, or repeated testing with K-means clustering often fall short in sparse UWSNs due to the variability in erroneous data transmission caused by defective nodes.To address these limitations, this paper proposes a K-means-based Distributed Self-Intermittent Fault Diagnosis (KDSIFD) algorithm to detect sensor nodes exhibiting soft (intermittent) faults. Each sensor node determines its fault status based on self-assessment and comparison with other cluster nodes. The approach relies on clustering and the distribution of test results among neighboring sensors to improve fault identification. Additionally, the K-means algorithm is utilized to diagnose intermittent failures within the sensing and communication aspects of the cluster network.The proposed KDSIFD system is implemented and simulated in Python, and its performance is evaluated against existing intermittent fault detection techniques. The test results show that KDSIFD greatly increases detection accuracy and lowers the rates of false alarms and false positives, especially in sparse UWSN environments where traditional methods have a hard time finding faults.