授权服务中策略决策点的高效评估方法  

Study of high efficiency of evaluation of a PDP in authorization service

作　　者：邓凡[1,2] 陈平[3] 张立勇[3] 王献青[3] 厉孙德 王彬[1] 

机构地区：[1]西安电子科技大学计算机学院,陕西西安710071 [2]西安电子科技大学软件学院,陕西西安710071 [3]西安电子科技大学软件工程研究所,陕西西安710071

出　　处：《西安电子科技大学学报》2013年第6期105-110,共6页Journal of Xidian University

基　　金：高等学校基本科研业务费资助项目(K50511230002)

摘　　要：授权服务中策略决策点的评估效率是影响授权服务效率的关键点,现有诸多策略决策点的评估效率往往受到规则中条件因素的影响.分析了根据用户历史访问记录来对策略中的规则进行调整以提高评估速度的方法在实时性方面存在的不足,提出了根据用户访问记录实时调整策略内规则顺序及采用为策略决策点添加索引与自适应的方法以提高评估速度.实验结果表明,当策略中包含的规则数为9 795时,采用索引与自适应方式的策略决策点的评估时间及完成一次授权操作调用的时间分别比未采用索引与自适应方式的策略决策点的缩短了93.33%和76.32%;当同一用户重复访问次数为7时,采用索引加自适应方式的策略决策点完成一次授权服务的调用时间比仅采用索引的策略决策点的缩短了24.29%.In authorization service, the efficiency of evaluation of a PDP（policy decision point） is a key issue that affects the efficiency of authorization service. However, the efficieneies of evaluation of many PDPs available are frequently influenced by the factors of conditions in rules. Said Marouf et al put forward a method in which the evaluation speed can be increased by adjusting the rules in policies according to user history access records. This paper analyzes the shortcomings of this method in real-time applications, and proposes that the rules order in policies can be instantaneously adjusted by user history access records and indexes as well as self-adaptation are added to a PDP so that the evaluation speed can be increased substantially. Experimental results indicate that when there are 9 795 rules in a PDP, the evaluation time and the time consumption taken for a PDP applied with indexes and self-adaptation to complete an authorized operation call are shortened by 93.33% and 76.32% respectively, compared with those for a PDP which is not applied with indexes and self-adaptation. When the number of repeat visits from the same user is 7, the time consumption taken for a PDP applied with indexes and adaptation to complete an authorized operation call is shortened by 24.29% compared with that for a PDP which is applied with indexes only.

关 键 词：授权服务 策略决策点 评估效率 索引 自适应 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]