Improvement of VPN Authentication Using Facial Recognition Technology
| Autor: | Cheng-Kang Liu, 劉政鋼 |
|---|---|
| Rok vydání: | 2014 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 102 SSL VPN (Secure Sockets Layer Virtual Private Network) is a form of VPN that can be used with a standard web browser via a secure and authenticated pathway by encrypting all network traffic. This protocol achieves a higher level of compatibility with client and server platforms and hence provides a more reliable connection. In general, SSL VPN access can be granted by a user certificate and password in a LDAP. In some particular applications, SSL VPN intends to access important resources that are restricted and not as a general access solution. The resources require much higher secure authentication. In this thesis, we propose a novel SSL VPN double authentication scheme. An advanced feature is proposed based on endpoint security to check a connecting computer to make sure it passes facial recognition rules before allowing a user to log in to SSL VPN. In the proposed scheme, if the authentication successfully passes the SSL handshake process and certification, it is then passed the face recognition as the second layer authentication. The server requests the facial recognition parameters from the client for authentication. If the authentication is successful, the protocols go to the next stage or the protocols fail. The client can login to access the server if only all the protocols are successful. In our experiments, the accurate rate is 88.7% on images of person himself, and 77% on images of person non-self. From the experimental results, it indicates that our proposed scheme is an advance feature of identification on login to SSL VPN to decrease identification spoofing on the internet. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|