Privacy protection in trust management

As pervasive computing leading to an increased collection of information in the computational world as well as the real world, privacy leaking becomes a serious issue. Although privacy protection has drawn great attention in recent years, the privacy-preserving trust management has not been brought forward. Credential chain discovery problem is the key issue in trust management, and traditionally credentials are full open in the whole system. This could expose users' access control policies as well as other private information, and leads to a great risk of being cheated by malicious users. This paper advocates a privacy-preserving credential chain discovery mechanism, in which credentials are no longer available to everyone. By merely learning credentials of one's logic neighbors, this mechanism still achieves good results. The simulations show that this mechanism is practical.