HIDS-IoMT: A Deep Learning-Based Intelligent Intrusion Detection System for the Internet of Medical Things

Abstract

The expansion of the Internet of Medical Things (IoMT) has enhanced the accuracy, real-time functionality, connectivity, and intelligence of medical examination practices. However, increased interconnectivity of medical equipment has rendered IoMT networks susceptible to several cyberattacks, particularly the Distributed Denial of Service (DDoS) attacks. Current intrusion detection systems (IDS) are insufficiently focused to address and counteract these advanced threats. Furthermore, due to the dynamic nature of IoMT traffic, IDS has considerable difficulty preserving its current threat detection capabilities. This study presents a hybrid deep learning-based intrusion detection system for IoMT networks (HIDS-IoMT). The proposed model hybridizes the Convolutional Neural Network (CNN) for feature extraction and the Long Short Term Memory neural network (LSTM) for sequence data prediction. We implement the designed IDS on a Raspberry Pi device using a fog computing architecture, enabling decentralized processing closer to IoMT devices, thereby enhancing responsiveness and reducing latency. We evaluate the suggested approach for intrusion detection using the IoTID20 and the Edge-IIoTset datasets. These datasets comprise a substantial and varied assortment of traffic flows from actual DDoS attacks, including SYN floods, UDP floods, HTTP floods, and others. We evaluated our novel methodology against distinct DoS attacks such as DDoS, SQL injection, Vulnerability scanner, Scan Host Port, Mirai, etc. Our proposed model attains an accuracy of 99.92%, a precision of 99.91%, a recall rate of 99.99%, and an F1-score of 99.95%, outperforming the latest in advanced methodologies.