基于多级安全策略的二维标识模型

A Planar Attributes Model Based on Multi Level Security Policy

安全模型是用形式化的方法来描述如何满足系统的安全要求 .经典的安全模型都只能要么满足系统的保密性要求 (如BLP模型 ) ,要么满足完整性的要求 (如Biba模型 ) .该文提出了一个多级安全策略的二维标识模型 ,在对可信主体必须遵守最小特权原则的前提下 ,利用保密性标识和可信度标识共同构成主客体的访问标识 ,并利用两个约束条件 ,使得既能防止越权泄露信息 ,又能控制信息的非授权修改 ,从而同时保证了系统的保密性和完整性 .

Security model is the foundation for meeting the security needs of information systems. Classical security models can only meet either confidentiality or integrity. None of model could resolve both of security requirements simultaneously. This paper presents a planar attributes model based on multi level security policy. In the model, the trusted subjects must accord with the least privilege principle. The concepts of confidence level and trusted level as labels of subjects and objects for constraining subjects accessing to objects are briefly described. This paper details a formal description and verification on the model which provides both the confidentiality and integrity for the system. It also explain the model in informal language briefly to help understand. Lastly, some open issues about model implementation are discussed, and the future direction of the model is pointed out. This model will be beneficial to be successful applied to constructing security systems in practices.