Energy-Aware Hybrid AI Framework for Intelligent Threat and Anomaly Detection in Resource-Constrained Wireless Sensor Networks

Abstract

This paper presents an energy aware hybrid Artificial Intelligence (AI) approach to intelligent threat detection and anomaly identification in resource restricted Wireless Sensor Networks (WSNs). However, securing the computer systems remains one of the challenging issues due to limited energy resources, computational capacity, and dynamic characteristics of WSN nodes. To overcome all of the above problems, this paper introduces a hybrid framework that combines Convolutional Neural Networks (CNN) and Random Forest (RF) algorithms to enable accurate and efficient detection of multiple cyber threats such as Distributed Denial of Service (DDoS) attacks, malware intrusions, data breaches, unauthorized access attempts and more. Furthermore, an adaptive anomaly detection method is designed to learn from network traffic and sensor data in order to detect abnormal patterns that may reflect security vulnerabilities. The proposed model adopts an energy-awareness approach, unlike traditional approaches; minimizing computational efforts while achieving high detection performance that is suitable for WSN environments with resource constraints. The results of our exploration show that the hybrid CNN+RF model provides high accuracy for threat classification and uniquely detects anomalies in real-time approaches. The proposed study shows a greener way of providing lightweight energy efficient intelligent detection for an anomalous event using modern techniques in WSN by means of deep learning and machine learning approaches. thus, this gives rise to better threat discovery and detection at earliest stages which in turn, helps in enhanced functionality from a security point of view all over WSN.