Title: A novel wavelet packet-based anti-spoofing technique to secure ECG data

Authors: Fahim Sufi, Seedahmed Mahmoud, Ibrahim Khalil

Addresses: Science, Engineering and Technology Portfolio, RMIT University, 330 Swanston St., Melbourne, VIC-3000, Australia. ' Science, Engineering and Technology Portfolio, RMIT University, 330 Swanston St., Melbourne, VIC-3000, Australia. ' Science, Engineering and Technology Portfolio, RMIT University, 330 Swanston St., Melbourne, VIC-3000, Australia

Abstract: Research related to ECG based biometric authentication is recently gaining popularity. However, there is substantial lack of research in anti-spoofing measurement to protect ECG recognition data from being captured in the wrong hands, where it might be subject to replay attack. Anonymisation of ECG data not only protects this valuable biometric data from being utilised for unauthorised access to restricted facility, but also hides major cardiovascular details of a particular person upholding HIPAA regulations. This paper proposes a novel ECG anonymisation technique based on wavelet packets. It was proven to provide 100% anonymisation, showing robustness against replay attack by the spoofer. Even with the most recent available technology, the anonymised ECG remained totally unidentified. A key, which is only 5.8% of the original ECG, is securely distributed to the authorised personnel for reconstruction of the original ECG.

Keywords: ECG-based biometrics; ECG anonymisation; wavelet packets; secure ECG transmission; anti-spoofing; spoofing; security; anonymity; privacy; electrocardiograms.

DOI: 10.1504/IJBM.2008.020144

International Journal of Biometrics, 2008 Vol.1 No.2, pp.191 - 208

Published online: 30 Aug 2008 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article