A Survey on Deep Learning Enabled Intrusion Detection System for Internet of Things

Abstract

Internet of Things (IoT) has led the world to grow at a revolutionary rate that is bringing profound impact in various domains without human intervention. Although IoT networks are enhancing different application domains, they are vulnerable and possess different cyber threats to its adoption and deployment. Thus, the continuously growing threats demand security measures that can lead IoT adoption without risk around the globe. Also, the security methods that are deployed for traditional networks lack security provisioning for IoT networks as these are resource constraint, deploy usage of diverse protocols, therefore existing security schemes demand to be evolved to be adopted for the security of IoT systems. As the IoT ecosystem grows rapidly with time and devices are connected via wireless sensor-based connections, deploying intelligent attack detection methods are one of the promising techniques which can detect different cyber-attacks and monitor IoT networks intelligently. Thus, the intrusion detection systems (IDS) play an important role in the security of IoT networks and evolving IDS with the application of deep learning networks leads to the generation of more powerful security systems that can be capable of detecting even zero birth attacks and making IoT networks more adaptable. The aim of this paper is to provide a comprehensive survey of IDS in IoT networks which has evolved in past years with the usage of machine learning and resulted in improving the performance of IDS. The main objective of this survey is to thoroughly review different deep learning methods for IDS.