Scholarlink

<p>Wireless sensor networks (WSN) are used extensively in applications like healthcare monitoring, industry monitoring, environmental monitoring, etc. But the security of the wireless sensor network still remains a major concern because most wireless sensor networks are deployed openly, which makes them more susceptible to potential cyber-attacks. This study proposed an advanced version of the intrusion detection framework that can achieve better efficacy in detecting the intruded nodes through the proposed approach. A weighted intrusion risk score (WIRS) model is proposed in this study that utilizes the characteristics of the nodes in the wireless sensor network, such as remaining energy, received signal strength indicator (RSSI) value sent to the cluster head, distance between the nodes and the cluster-head, advertisement packets sent out, and the anomalies associated with the nodes. Cross-datasets were mapped with the proposed model, which can be used on diverse datasets like the NSL-KDD dataset, CICIDS 2017, and the proposed approach worked efficiently on all three datasets. An overall accuracy level of 97% was achieved with the proposed approach on the NSL-KDD dataset, followed by 96% on the CICIDS 2017 dataset, while on the WSN-DS dataset, the overall efficacy proved better with an accuracy level of 98%. Moreover, there is considerable improvement in detecting WSN specific attacks like Black holeattack Grey hole attack. These results establish WIRS as an effective risk-aware feature that enhances intrusion detection across diverse network scenarios, supporting scalable and interpretable cybersecurity solutions for WSNs.</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

Advancing wireless sensor network security through enhanced intrusion detection techniques

  • 期刊论文

  • 发表日期：

    2026年2月23日

查看全文
  • Mohit Angurala,
  • Sivaneasan Bala Krishnan,
  • Pravin R. Kshirsagar,
  • Balajee Maram

摘要

Wireless sensor networks (WSN) are used extensively in applications like healthcare monitoring, industry monitoring, environmental monitoring, etc. But the security of the wireless sensor network still remains a major concern because most wireless sensor networks are deployed openly, which makes them more susceptible to potential cyber-attacks. This study proposed an advanced version of the intrusion detection framework that can achieve better efficacy in detecting the intruded nodes through the proposed approach. A weighted intrusion risk score (WIRS) model is proposed in this study that utilizes the characteristics of the nodes in the wireless sensor network, such as remaining energy, received signal strength indicator (RSSI) value sent to the cluster head, distance between the nodes and the cluster-head, advertisement packets sent out, and the anomalies associated with the nodes. Cross-datasets were mapped with the proposed model, which can be used on diverse datasets like the NSL-KDD dataset, CICIDS 2017, and the proposed approach worked efficiently on all three datasets. An overall accuracy level of 97% was achieved with the proposed approach on the NSL-KDD dataset, followed by 96% on the CICIDS 2017 dataset, while on the WSN-DS dataset, the overall efficacy proved better with an accuracy level of 98%. Moreover, there is considerable improvement in detecting WSN specific attacks like Black holeattack Grey hole attack. These results establish WIRS as an effective risk-aware feature that enhances intrusion detection across diverse network scenarios, supporting scalable and interpretable cybersecurity solutions for WSNs.