A multilevel relational database system handles data at different security classifications and provides access to users with different security clearances. Many methods for enforcing security in this environment have been investigated. This paper presents a unified approach to multilevel database security based on two ideas: a trusted filter and an inference engine. These two approaches are introduced separately and then the motivation for the unified system and the system model itself are presented.
All Science Journal Classification (ASJC) codes
- Food Science
- Hardware and Architecture