A Comparative Analysis of Machine Learning Models for Enhancing Wormhole Attack Detection in Wireless Sensor Networks

protocol [30]. Given their dynamic nature and expanding use across diverse domains, the demand for robust security mechanisms in WSNs is becoming increasingly critical. These networks frequently operate in hostile environments where node-to-node communication can be unstable, complicating the deployment of effective security solutions. Safeguarding nodes against potential security threats presents a significant challenge, as WSNs are prone to various forms of attacks [3], including jamming, collision, wormhole, flooding, sinkhole, selective packet drop, Sybil, cloning, denial-of-service, and tampering. Among these, the wormhole attack stands out as a particularly severe threat, targeting the routing protocols within WSNs [6]. Objectives: The objective of this study is to evaluate the performance of various machine learning algorithms in detecting wormhole attacks within wireless sensor networks. We conduct a comparative analysis of different models using key performance metrics such as accuracy, precision, recall, F1-score, and computational efficiency. Furthermore, the study aims to identify the most suitable model for real-time detection in resource-limited WSN environments. Methods: This study conducts a comparative evaluation of machine learning models for detecting wormhole attacks in WSNs, addressing the limitations of traditional methods like distance estimation and hop count. By leveraging anomaly-based analysis, the research involves selecting suitable models, preprocessing a simulated dataset, and training the models for evaluation using metrics such as accuracy, precision, recall, F1-score, and computational efficiency. The goal is to identify the most effective and resource-efficient models for real-time deployment in WSN environments. Results: The results of our comparative analysis of machine learning models for wormhole attack detection in wireless sensor networks (WSNs) reveal critical insights into the effectiveness and limitations of each algorithm. Using real-world network simulation data, we evaluated the models based on accuracy, precision, recall, F1-score, and computational time. This section highlights the comparative strengths and weaknesses of the models in identifying wormhole attacks, offering practical considerations for selecting suitable models in resource-constrained WSN environments and guiding future improvements in detection systems. Conclusions: In conclusion, this study presents a comparative analysis of seven machine learning models for wormhole attack detection in WSNs, highlighting their strengths, limitations, and practical applicability. The findings offer valuable insights into the trade-offs between accuracy and computational efficiency, contributing to the development of effective, resource-aware security solutions in wireless sensor networks.