Loading...

Vulnerability Analysis of Anonymity Protocols

Emamdoost, Navid | 2010

495 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 41144 (19)
  4. University: Sharif University of Technology
  5. Department: Computer Engineering
  6. Advisor(s): Jalili, Rasool
  7. Abstract:
  8. Anonymity helps in preserving user’s privacy in communication networks. In this respect many protocols with different characteristics have been proposed, as yet. Beside these protocols there have been introduced a lot of attacks, in which the ultimate goal is reducing the degree of provided anonymity. There is a class of attacks, where attacker tries to relate users of anonymity system to each other, just by observing execution of protocol. There is no need to compromise system’s nodes or even manipulating sent or received messages. Here, the attacker omits internal mechanism of protocol and assumes it has no vulnerability. The Statistical Disclosure Attack (SDA) is one of these attacks. Using SDA, the attacker is able to probabilistically relate users to each other. The mere bottleneck of these attacks is the number of observations required to success. In this thesis, we improve and extend SDA, and also propose a solution to postpone the attack by requiring more observations to success. To improve SDA, all the observations available to the attacker are incorporated, to estimate other users’ behavior. Also, we propose a way of using background knowledge of user profiles to degrade anonymity. All of these contributions decrease the number of observations needed for attack success. On the other hand, we extend SDA to a low latency protocol. Already SDA was applicable just to protocols which use batch processing, but most of popular protocols are low latency and therefore do not use batch processing. At last we propose a way of postponing the attack. In this solution, a user based on her needs decides to cooperate with n other users (colleagues). In each round that she wants to send, her colleagues will send a message, too. We show the effectiveness of our improvement and resistance through simulations
  9. Keywords:
  10. Anonymity ; Privacy ; Unlinkability ; Mix Net ; Statistical Disclosure Attack (SDA)

 Digital Object List

 Bookmark

No TOC