An Intrusion Detection System Using Machine Learning for Internet of Medical Things

Abstract

The COVID-19 pandemic has impacted everyday life, the global economy, travel, and commerce. In many cases, the tight measures put in place to stop COVID-19 have caused depression and other diseases. As many medical systems over the world are unable to hospitalize all the patients, some of them may get home healthcare assistance, while the government and healthcare organizations have access to substantial sickness management data. It allows patients to routinely update their health status and have it sent to distant hospitals. In certain cases, the medical authorities may designate quarantine stations and provide supervision equipment and platforms (such as Internet of Medical Things (IoMT) devices) for performing an infection-free treatment, whereas IoMT devices often lack enough protection, making them vulnerable to many threats. In this paper, we present an intrusion detection system (IDS) for IoMTs based on the following gradient boosting machines approaches: XGBoost, LightGBM, and CatBoost. With more than 99% in many evaluation measures, these approaches had a high detection rate and could be an effective solution in preventing attacks on IoMT devices.