A Secure ECG based Smart Authentication Scheme for IoT Devices

Abstract

In this super-evolving world, there exist numerous hidden vulnerabilities and any kind of security breach in recent times would have implications in ways which are even difficult to imagine. To counter such probable scenarios the demand evolved to design security protocols which are sound enough to counter most of such situations in near future. In this scheme, an ECG based user authentication protocol for smart IoT devices is proposed. ECG is one of the unique characteristic traits of any individual which is experimented to remain similar throughout, provided any biological changes have not occured. The scheme uses the captured ECG signal from a smart IoT device taken as raw input by the computation model. The signal is then used to extract the user characteristics such as the QRS complex, P peak, T peak, etc. A feature vector is derived after pre-processing of the signal, which is then used by the computation model to check for the authenticity of the individual. The final authentication of the user depends upon the similarity index of the feature vector generated by the existing user’s record captured during registration. The data used here is from the ECG ID database and the obtained accuracy is measured close to 94.2%. A comprehensive study of the performance of the protocol is presented taking communication cost, computation cost, storage cost, and even the security features as the parameters for comparison.