A Novel Hybrid, BERT and Deep Learning Model Network Intrusion Detection System for Healthcare Electronics
Ali Alferaidi, Kusum Yadav, Yasser Alharbi, Eissa Jaber Alreshidi, Abdulrahman Alreshidi, Bassam W. Aboshosha, Rohit Sharma, Ahmed Alkhayyat, Daniel Gavilanes Aray
Abstract
IoMT has become an attractive playground for cybercriminals due to its market value and rapid growth. As the amount of sensitive data transmitted across IT infrastructures grows, healthcare organizations and companies that generate wearable data become targets for attackers. Recently, government agencies and healthcare organizations have prioritized collecting this data and using machine learning to protect users’ privacy. In all test cases, the proposed method performed better than any other method, including the IoMT intrusion dataset, with an accuracy increase of 2.9%. It can also monitor IoMT networks within the healthcare and medical environment to protect IoMT devices and networks from attackers. Based on the ECU-IoHT dataset, it achieves 99% performance improvement in accuracy, precision, recall, and F1-score compared with existing anomaly detection models. The proposed model shows higher detection accuracy than the existing latest state-of-arts.