A Context-Aware Framework for Adaptive Software Security

自适应软件安全的上下文感知框架

• 批准号: RGPIN-2017-04488
• 负责人: Tahvildari, Ladan
• 金额: $ 1.89万
• 依托单位: University of Waterloo
• 依托单位国家: 加拿大
• 项目类别: Discovery Grants Program - Individual
• 财政年份: 2019
• 资助国家: 加拿大
• 起止时间: 2019-01-01 至 2020-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=681337
• 关键词: Context; Aware; Framework; Adaptive; Software

Dynamically changing environments and threat landscapes require adaptive software that enables the change and modification of security mechanisms at runtime. Research has been conducted with two different emphases in this area: context-awareness and self-protection. Research focused on context-awareness is more concerned with how to model, process, and manage contextual information, but is limited in its understanding of how a system adapts itself in response to unanticipated security changes in the context information. On the other hand, research into self-protection focuses more on how to adapt the system's structure and/or behavior in response to requirements and/or unanticipated security changes, but pays less attention to how the context is modeled, processed, and managed. ******The long-term goal of the proposed research program is to gain full-awareness of cyber attacks at application-layers. In the next five years, the goal is to investigate, design, and develop a novel framework that supports consideration of context-awareness and adaptive security in an integrated manner. The specific objectives are: (1) to design an adaptive model for context recognition, (2) to design a set of requirements and runtime models for context comprehension, and (3) a set of algorithms to handle context projection and resolution. ******Existing approaches in literature lack the reasoning and learning capabilities required to gain context-awareness for cyber defense. However, the recent PI's preliminary work on modeling the interactions between the attacker and the adaptation manager as a Markov game suggests incorporating context in adaptive security to handle application-layer attacks. The PI will propose a set of integrated models and methods for monitoring, detecting attack-type uncertainty, and applying adaptation decisions in cyber space that are currently unachievable with existing techniques.******This research program will create high-quality research capabilities. It will open a new research direction that significantly reduces the cost impact of cyber security threats for more than 1,300 Canadian software companies, and improves Canada's competitiveness in this vital sector. Industry will benefit from the findings of this program in two areas. The provision of a superior context-aware framework will allow for the development and implementation of sophisticated machine learning algorithms and methods for adaptive security at runtime, and the adaptive software security modeling techniques will help companies to employ effective methods to keep self-protecting systems up-to-date during their evolution and maintenance. This research program will train High Qualified Personnel (HQP) to work in the field of software engineering and adaptive security. The HQP will acquire a unique blend of interdisciplinary knowledge and apply that knowledge to improve adaptive software security.

动态变化的环境和威胁场景需要自适应软件，以便在运行时更改和修改安全机制。在这方面的研究有两个不同的重点：上下文意识和自我保护。关注上下文感知的研究更多地关注如何建模、处理和管理上下文信息，但对系统如何适应上下文信息中未预料到的安全变化的理解有限。另一方面，对自我保护的研究更多地关注于如何适应系统的结构和/或行为以响应需求和/或未预料到的安全更改，但较少关注如何对上下文进行建模、处理和管理。******提出的研究计划的长期目标是全面了解应用层的网络攻击。在接下来的五年里，目标是研究、设计和开发一个新的框架，以集成的方式支持考虑上下文感知和自适应安全性。具体目标是：(1)设计一种用于上下文识别的自适应模型；(2)设计一套用于上下文理解的需求和运行时模型；(3)一套用于处理上下文投影和解析的算法。******文献中现有的方法缺乏获得网络防御上下文感知所需的推理和学习能力。然而，最近PI的初步工作是将攻击者和适应管理器之间的交互建模为马尔可夫博弈，建议在自适应安全中结合上下文来处理应用层攻击。该项目将提出一套集成模型和方法，用于监测、检测攻击类型的不确定性，并在网络空间中应用适应性决策，这些是目前现有技术无法实现的。******该研究计划将创造高质量的研究能力。它将开辟一个新的研究方向，为1300多家加拿大软件公司显著降低网络安全威胁的成本影响，并提高加拿大在这一重要领域的竞争力。工业将受益于该计划在两个方面的发现。提供优越的上下文感知框架将允许在运行时开发和实现复杂的机器学习算法和自适应安全方法，自适应软件安全建模技术将帮助公司采用有效的方法，使自我保护系统在其发展和维护期间保持最新状态。该研究计划将培养在软件工程和自适应安全领域工作的高素质人才（HQP）。HQP将获得跨学科知识的独特混合，并应用这些知识来提高自适应软件安全性。