Anomaly Detection in Wireless Sensor Networks: A Comparative Analysis of Performance Metrics and Model Accuracy

This research paper explores hybrid anomaly detection frameworks combining Particle Swarm Optimization (PSO), Bee Algorithm (ABC), and Genetic Algorithm (GA) using the CICIDS2017 dataset for Wireless Sensor Networks (WSNs). It evaluates their effectiveness across diverse network conditions, focusing on anomaly detection accuracy, false positives, and computational efficiency. The study aims to advance anomaly detection methodologies in WSNs by contributing insights into bio-inspired algorithms' impact on anomaly detection. It targets the research community, industry professionals, and scholars, emphasizing the documentation of methodologies and suggesting future research on hybridization techniques, dynamic adaptation, edge computing integration, and machine learning fusion for enhanced anomaly detection in WSNs, fostering improved network security and adaptability.