Classification of Deep Learning methods in Intrusion Detection for IoT Devices
Kalra Ashish, Kumar Manoi · 2024
Recent advancements of internet of things (IoT) has resulted in sizable deployment of IoT devices. Most homes have IoT devices in one or another form present now days. Deployment of IoT devices at home results in large quantity of data being processed while providing intelligent services to user. This results in several security concerns related to security of IoT systems. IoT devices has limitation related to their computation capability and this need to be considered while checking intrusion detection techniques. Researcher has implemented different techniques for identification and prevention of such attacks. Intrusion detection can be done using traditional methods or deep learning-based methods or using hybrid combination among both of these. New type of attacks which can happened to IoT systems are called zero-day attacks. These are not known to anyone first. Traditional methods are efficient but they are weak in detection of zero-day attacks. Our survey present detailed comparison and review of latest deep learning-based methods for intrusion detection and also check future research direction in this domain. We also provide a comparative analysis with focus on compatibility, challenges, feasibility and real time issues. This survey will be helpful to both academia and industry to check future direction in intrusion detection for IoT devices.