Deep Learning based Intrusion Detection for IoT Networks

Abstract

The Internet of Things (IoT) holds vast potential across a diverse range of applications, spanning from industrial automation to healthcare and defense networks. The security of an IoT network is crucial, as it directly impacts the overall security of the underlying computing and communication infrastructure. However, owing to resource constraints and limited computational capabilities, IoT networks usually using Message Queuing Telemetry Transport (MQTT) protocol are susceptible to various types of attacks and security threats. In this paper, we propose an Intrusion Detection System (IDS) for IoT networks that is based on Deep Learning concepts. In particular, we propose a Long Short Term Memory (LSTM) model for IoT intrusion detection and attack mitigation focusing on MQTT protocol. We have trained and tested our model on a typical IoT tailored dataset using optimal feature set. Extensive experiments have shown that, with less features, the proposed LSTM remains its ability with lower complexity.