Document Details
|
Document Type |
: |
Thesis |
|
Document Title |
: |
MULTI-BIOMETRIC CONTINUOUS AUTHENTICATION ON PCS USING ARTIFICIAL IMMUNE SYSTEM المصادقة المستمرة متعددة البيومترية على أجهزة الكمبيوتر باستخدام نظام المناعة الاصطناعي |
|
Subject |
: |
Faculty of Computing and Information Technology |
|
Document Language |
: |
Arabic |
|
Abstract |
: |
Authenticating a legitimate user is an important aspect of any system security; which indicates that the authentication process has to be secure and effective. In case that authentication method is compromised, the other security aspects like availability, authorization, integrity, confidentiality, could be compromised also. However, most of the existing systems used Knowledge-based authentication methods, for example, a password or PIN; due to its simplicity and flexibility. Meanwhile, this type of authentication is usually performed as one time identity proof over the initial log on process, where the user is expected to be the same one during the full session. Since this authentication occurs only once, an unauthorized user could have access to the genuine user resources and steal secret information, by stealing user password or by exploiting unattended open computer. On other hand, continuous authentication overcomes this limitation by continuously authenticates and monitors the genuine user. The user authenticity is verified continuously based on user biometric behavioral to detect anomalous behavior, to limit the information disclosure.
This research proposes several enhancements to better secure the authentication system by exploring the mouse and keystroke dynamics as a biometrics behavior. To achieve this goal; artificial immune systems (AIS), Negative Selection (NS) algorithm is proposed to authenticate the users continuously.
Our proposed system was conducted with a unique biometric dataset. The used dataset was collected over a completely uncontrolled setting from 30 users. In this research, a combination of two biometric, mouse and keystroke user behavior were used. Our proposed system shows that AIS was correctly able to continuously authenticate the users with high accuracy. The best achieved accuracies for 20 NS runs, where the number of generated detectors is 100, 200, 300, and 600, are 98.10%, 99.24%, 99.5%, and 99.8% respectively. |
|
Supervisor |
: |
Dr. Fahad Alsolami |
|
Thesis Type |
: |
Master Thesis |
|
Publishing Year |
: |
1439 AH
2018 AD |
|
Added Date |
: |
Thursday, May 10, 2018 |
|
Researchers
| عهود عواد الجهني | Aljohani, Ohood Awad | Researcher | Master | |
|
Back To Researches Page
|