吝啬语义信任协商

Parsimonious Semantic Trust Negotiation

自动信任协商是通过数字证书的交互披露在陌生实体间建立信任的过程,现有的自动信任协商框架要求主体出示完整属性证书来证明自己满足对方资源披露策略中的身份断言约束条件,导致了属性证书中身份信息过度披露问题.该文中作者提出一种通过交换DL-TNL语义身份断言来建立信任关系的吝啬语义信任协商框架,避免了完整属性证书的直接出示,减少信任建立过程中身份信息被披露的程度,并提出一种正确、完备且有效的吝啬语义信任协商策略.在该策略下,交易双方的协商引擎可以快速有效地从由身份断言权威签发的包含多个DL-TNL语义身份断言的身份断言证书中,自动计算出批露最少信息且符合对方策略的身份断言集进行出示,以最大限度地减少信任建立过程中被披露的身份信息,并保证理论上存在成功可能性时,使用该策略必然可以有效地帮助主体最终获得网络资源的访问权限.有关该策略完备性和正确性的证明以及策略实施中所使用的符合性检测算法的实现和分析均在文中给出.

Automated Trust Negotiation （ATN） is a process in which two unfamiliar entities ex- change their digital attribute certificates in turn to set up mutual trust relationship with each oth- er. Since existing ATN frameworks require entities to release entire attribute certificates to prove that they satisfy the identity constraint condition stated by the other entities＇ access control poli- cies, private identity information of entities is often over-revealed in many circumstances. In this paper, the authors propose a novel parsimonious semantic trust negotiation framework in which entities can build trust relationship by exchanging DL-TNL semantic identity assertions instead of entire attribute certificates. This framework can greatly reduce the degree of disclosed private identity information. Under the framework, the authors propose a correct, complete and efficient parsimonious semantic trust negotiation strategy to allow the negotiation agencies of participants to compute and disclose a satisfying set of DL-TNL assertions which contains least private identi- ty information at every exchange step. In a word, the parsimonious semantic trust negotiation strategy can minimize the disclosed private identity information in the trust negotiation processes and guarantees the resource requesters participating in the processes to get the access rights finally if there does exist a theoretical successful trust negotiation sequence. All of the proofs about the properties of the strategy are given in this paper, so are the relevant compliance checking al gorithms and their analyses.