Intrusion Detection Based on Stacked Autoencoder for Connected Healthcare Systems

Abstract

With the people-oriented medical concept gradually gaining popularity and the rapid development of sensor network technology, connected healthcare systems (CHSs) have been proposed to remotely monitor the physical condition of patients and the elderly. However, there are many security issues in these systems. Threats from inside and outside the systems, such as tampering with data, forging nodes, eavesdropping, and replay, seriously affect the reliability of the systems and the privacy of users. After an overview of CHSs and their security threats, this article analyzes the security vulnerabilities of the systems and proposes a novel intrusion detection method based on a stacked autoencoder. We have conducted extensive experiments, and the results demonstrate the effectiveness of our proposed method.