Scholarlink

<p>Wireless Sensor Networks (WSNs) play a vital role in critical applications such as environmental monitoring, healthcare, and military surveillance. However, their limited resources and vulnerability to security threats, particularly Distributed Denial-of-Service (DDoS) attacks, significantly degrade performance and network lifetime. To address these challenges, this paper proposes a Lightweight Trusted Clustering Framework that integrates Bidirectional Long Short-Term Memory (Bi-LSTM) networks for malicious node detection with the Walrus Optimization Algorithm (WOA) for optimized Cluster Head (CH) selection. The Bi-LSTM model identifies malicious nodes based on behavioral metrics such as Packet Delivery Rate (PDR), energy consumption, and response time, ensuring that only trusted nodes are included in clustering. Subsequently, WOA selects CHs by optimizing residual energy, distance to the base station, node density, and trust value, thereby achieving energy efficiency while maintaining resilience against attacks among the 3-level energy heterogeneous nodes. The proposed framework is evaluated under DDoS scenarios using MATLAB simulations, with performance assessed through stability period, network lifetime, remaining energy, throughput, and detection rate. Results demonstrate that the proposed model extends the stability period by 35%, improves network lifetime by 18%, and achieves a 2.6% higher DDoS detection rate compared to recent secure routing approaches. These findings highlight the effectiveness of combining Bi-LSTM with WOA to deliver a robust, energy-efficient, and secure clustering solution for WSNs operating in adversarial 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

Trusted clustering framework for secure wireless sensor networks using Bi-LSTM and walrus optimization algorithm

  • 期刊论文

  • 发表日期：

    2026年1月3日

查看全文
  • Anupriya .,
  • Aruna Malik

摘要

Wireless Sensor Networks (WSNs) play a vital role in critical applications such as environmental monitoring, healthcare, and military surveillance. However, their limited resources and vulnerability to security threats, particularly Distributed Denial-of-Service (DDoS) attacks, significantly degrade performance and network lifetime. To address these challenges, this paper proposes a Lightweight Trusted Clustering Framework that integrates Bidirectional Long Short-Term Memory (Bi-LSTM) networks for malicious node detection with the Walrus Optimization Algorithm (WOA) for optimized Cluster Head (CH) selection. The Bi-LSTM model identifies malicious nodes based on behavioral metrics such as Packet Delivery Rate (PDR), energy consumption, and response time, ensuring that only trusted nodes are included in clustering. Subsequently, WOA selects CHs by optimizing residual energy, distance to the base station, node density, and trust value, thereby achieving energy efficiency while maintaining resilience against attacks among the 3-level energy heterogeneous nodes. The proposed framework is evaluated under DDoS scenarios using MATLAB simulations, with performance assessed through stability period, network lifetime, remaining energy, throughput, and detection rate. Results demonstrate that the proposed model extends the stability period by 35%, improves network lifetime by 18%, and achieves a 2.6% higher DDoS detection rate compared to recent secure routing approaches. These findings highlight the effectiveness of combining Bi-LSTM with WOA to deliver a robust, energy-efficient, and secure clustering solution for WSNs operating in adversarial environments.