A Deep Learning Model For Efficient Intrusion Detection In Wireless Sensor Networks

Abstract

The application of WSN has been rapidly growing in the field of agriculture, security, and manufacturing industry and healthcare. Sensors deployed in remote areas are difficult to reach and the resources such as power, storage, signal strength and communication range are constrained which makes the network to be vulnerable to different attacks and exploitations. WSN face security threats as the network is resource constrained due to establishment in hostile environment. Intrusion detection system plays an important role in mitigating WSN exploitations and attacks. Denial of Service (DoS) is the most common type of attack in WSN which disrupts the network functionality and affect sensor data. To efficiently detect intrusions and anomalies in WSN, a light weight deep learning model is proposed to detect different types of attacks. The performance of the proposed intrusion detection model is compared with state-of-art machine learning models. The performance of the proposed model demonstrated higher detection accuracy of different types of attacks in WSN.