الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 136 |  |  |
| | | from | 136 | | |
Abstract
This thesis is mainly concerned with a vital topic encountered in the medical community. Nowadays technology has invaded the medical community and this is represented in the utilization of Wireless Body Area Network (WBAN) and Internet of things (IoT) for the interchanging the confidential patient data. These types of networks need efficient security tools represented in good authentication system. Biometrics find an important role in all modern networks.
Moreover, the trend of cancellable biometrics is rising now. Its basic idea is to use distorted or encrypted version of the biometrics instead of the original ones. The objective is to allow the authentication process, while protecting the original biometrics from any hacking attempts.
Electrocardiogram (ECG) is a tool used to measure the heart electrical activity when it contracts over some time, also it is an appropriate signal to be used as a biometric in such types of networks. Hence we are interested in this thesis with cancellable ECG-based authentication. Four different algorithms are presented in this thesis to generate Cancellable ECG version.
The first algorithm depends on the Chirp Z-transform (CZT), absolute value estimation, and finally Discrete Cosine-transform (DCT). Cancellable templates are generated through the cascaded implementation of these stages.
The second algorithm depends on block based encryption of ECG signal segments. Chaiotic Baker map is used in the encryption process.
The third algorithm adopts the same methodology of the second one, but with a different encryption algorithm using matrix manipulation processes.
The fourth algorithm depends on the generation of the cepstra of the ECG signal segments and the convolution of these cepstra with masks extracted from the ECG signals to generate a high level of user privacy. Circular shifts is used in this algorithm to enhance the security levels.
The introduced cancellable ECG authentication systems are evaluated statistically based on probability density function of genuine and imposter distributions. In addition, the Area Under ROC Curve (AUC) is used as a metric for quality of authentication. Simulation tests prove the high quality and security of the proposed cancellable ECG recognition algorithms.