Collaborative Research: CT-T: Flexible, Decentralized Information-flow Control for Dynamic Environments

合作研究：CT-T：动态环境下灵活、分散的信息流控制

• 批准号: 0524035
• 负责人: Stephan Zdancewic
• 金额: $ 30万
• 依托单位: University of Pennsylvania
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2005
• 资助国家: 美国
• 起止时间: 2005-09-01 至 2009-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0524035&HistoricalAwards=false
• 关键词: Collaborative; Research; CT; Flexible; Decentralized

ABSTRACT0524035Zdancewic, StephanUniversity of PennsylvaniaNon-Lead Collaborative with 0524010, 0524132, 0524036Protecting confidential information in dynamic, distributed environments whose participants have heterogeneous trust relationships is an important and difficult challenge. An important problem is how to consistently enforce security constraints in a practical manner as policies and the systems themselves evolve over time. Without care, inconsistently-viewed updates to policy could allow a principal toperform actions granted by an old policy, or worse, could allow a principal to release information or perform actions authorized by neither an old nor a new policy, but rather an illegal combination of the two.The objective of the proposed research is to develop, implement, and evaluate security infrastructure for providing strong end-to-end security guarantees in dynamic environments where policy can change. The researchers will explore a novel synthesis of the state-of-the-art in security-typed programming languages for governing information flow; analysis for uncovering dependencies between principals, programs, and policies; and distributed protocols for coordinating policy updates. To motivate and validate the design decisions, this infrastructure will be incorporated into the programming language Cyclone, a type-safe variant of C, and used to build and evaluate distributed file system.

摘要 0524035Zdancewic，Stephan 宾夕法尼亚大学非主导合作者 0524010、0524132、0524036 在参与者具有异构信任关系的动态分布式环境中保护机密信息是一项重要且艰巨的挑战。 一个重要的问题是，随着策略和系统本身随着时间的推移而发展，如何以实际的方式一致地实施安全约束。 如果不小心，不一致的策略更新可能会允许委托人执行旧策略授予的操作，或更糟糕的是，可能会允许委托人发布信息或执行既不是旧策略也不是新策略授权的操作，而是非法组合两者。拟议研究的目标是开发、实施和评估安全基础设施，以便在策略可能发生变化的动态环境中提供强大的端到端安全保证。 研究人员将探索一种新颖的综合方式，将最先进的安全类型编程语言用于管理信息流；分析以揭示主体、计划和政策之间的依赖关系；以及用于协调策略更新的分布式协议。 为了激励和验证设计决策，该基础设施将被纳入编程语言 Cyclone（C 的类型安全变体）中，并用于构建和评估分布式文件系统。