The Research On The Visulized Anonymity Metric For Network Data

With the disclosure of privacy of individuals and corporations, many network security researchers have proposed a number of protective approaches, where anonymity is one of the most effective techniques for privacy. To prove the efficiency of anonymity and encourage more people to share their network data, a lot of metrics for evulating anonymity of network and microdata have been proposed, and these anonymity metrics effectively quantify the performance of anonymous communication system to the extent. However, the metric for both network and data is a currently hot topic.The thesis attempts to present a visualize metric for measuring anonymity of data network in order to further improve the theory of anonymity measure and better express the anonymity metrics, and then apply this measure to analyze and evaluate the degree of the anonymity that P2P anonymous communication system provided.A new visualized anonymity metric is firstly proposed for fulfiling the need of anonymity measure for both network and data, it takes the auxiliary information attackes obtained into consideration, based on the profile function of the conditional probability distributions, the external and internal degree of anonymity of anonymous communication system is fully quantified. Where the external degree of anonymity is composed by the degree of participients and data, the internal degree of anonymity of the system itself is made up of the relationship between the participients and the data. Then the difference between some previous anonymity metrics and the metric proposed by this paper is compared by the authentication mechanisms, and the comparison results show that the new method has its obvious advantages.Finally, the novel measurement is applied to evaluate the degree of the anonymity provided the P2P anonymous systems inoder to show the good feasibility and university of the new method. The degree of anonymity provided P2P system is calcualated by building the models of P2P anonymous system and the attacker, expectively. The results show that the degree of anonymity relates with the system size, the number of malicious nodes and the forwarding probability.