Grid is a heterogeneous and very dynamic environment where a scalable authorization system is needed. Existing grid authorization systems adopt mechanisms for storing resources’ security policies that suffer from huge repetitions while checking security rules. To overcome this problem, the Hierarchical Clustering Mechanism (HCM) was introduced. When HCM is used to find the SET of resources which the user is authorized to access, it tremendously reduces the redundancy in checking security rules compared to the classical methods (the Brute Force Approach and the Primitive Clustering Mechanism (PCM)). However, HCM demands a centralized authorization database where all the resources’ security policies have been kept in order to build its decision tree which is difficult to achieve in a heterogeneous environment like the grid. Moreover, HCM adds extra complexity when it is used in case of SINGLE resource authorization scenarios, which may happen very frequently in the grid. This paper tries to address these issues of HCM and comes up with an efficient solution.
分享
引用
