摘要
给出一种采用多层次优化技术的XACML(extensible access control markup language)策略评估引擎实现方案MLOBEE(multi-level optimization based evaluation engine).策略判定评估前,对原始策略库实施规则精化,缩减策略规模并调整规则顺序;判定评估过程中,在引擎内部采用多种缓存机制,分别建立判定结果缓存、属性缓存和策略缓存,有效降低判定引擎和其他功能部件的通信损耗.通过两阶段索引实现的策略缓存,可显著降低匹配运算量并提高策略匹配准确率.仿真实验验证了MLOBEE所采用的多层次优化技术的有效性,其整体评估性能明显优于大多数同类系统.
This paper proposes an implementation scheme of XACML （extensible access control markup language） policy evaluation engine based on multi-level optimization technology, MLOBEE （multi-level optimization based evaluation engine）. Before evaluating these policies, the scenario implements rule refinement to lessen scale policies and adjust the sequence at the rule. During evaluation, the engine adopts a multi-cache mechanism that includes result cache, attribute cache, and policy cache to reduce the communication cost between engine and other components. To decrease matching magnitudes and enhance matching exactitudes, policy cache practices two stage index techniques. Finally, emulation tests validate that the overall evaluation performance of MLOBEE, using multi-level optimization technology, is better than most other similar systems.
出处
《软件学报》
EI
CSCD
北大核心
2011年第2期323-338,共16页
Journal of Software
基金
国家自然科学基金(61003228
60803129)
中国科学院知识创新工程项目(YYYJ-1013)
下一代互联网业务试商用及设备产业化专项(CNGI-09-03-03)
关键词
XACML
访问控制
策略评估
规则精化
缓存机制
策略索引
XACML （extensible access control markup language）
access control
policy evaluation
rule refining
cache mechanism
policy index
