Methods and tools for intrusion-aware software systems

软件系统必须使用可靠的保护机制来设计，同时仍然在预算的成本和时间内交付软件的预期价值。实现这两个不同但相互依赖的目标的主要障碍是，当前的软件工程过程没有提供足够的方法和工具来实现安全目标。因此，安全漏洞往往使系统容易受到恶意攻击和滥用。在本研究中，将开发各种方法和工具来构建入侵感知软件系统。可用的软件规范语言并不完全适合指定软件的安全相关方面，而攻击语言对软件开发也没有用处。在软件规范和安全规范中使用两种不同的语言涉及到许多不必要但复杂的问题。本研究将著重于现有入侵场景描述机制及相应入侵侦测器的限制因素，并设计克服这些限制的方法。提出了一种规范方法，既可以规范目标的正常操作行为，也可以规范目标在受到已知攻击时的行为。将提供入侵场景规范所需的功能的正式语法和语义。构建入侵场景描述工具，并根据入侵场景自动生成签名。最后，将开发入侵检测系统，通过比较系统的运行时行为和生成的签名来检测对软件系统的任何入侵。这项研究将在弥合软件工程和安全工程实践之间的差距方面发挥至关重要的作用。这项研究计划的一个独特之处在于为研究生（2名博士和2名硕士）提供一个平衡的培训设施，包括构建安全软件产品的软件工程和安全工程原理。

Software systems must be engineered with reliable protection mechanisms, while still delivering the expected value of the software within the budgeted cost and time. The principal obstacle in achieving these two different but interdependent objectives is that current software engineering processes do not provide adequate methods and tools to achieve security goals. As a result, security loopholes often leave the system vulnerable to malicious attacks and abuses. In this research, various methods and tools will be developed with a view to building intrusion-aware software systems.    Available software specification languages are not completely suitable for specifying security related aspects of software, while attack languages are not useful for software development. The use of two different languages for software specification and security specification involves a number of unwanted but complicated issues. This research will focus on the factors related to the limitations of the existing intrusion scenario description mechanisms and the corresponding intrusion detectors, and will design ways to overcome them. A specification method will be proposed which can be used to specify both normal operational behavior and the behavior of the target when it is under known attack. Formal syntax and semantics of the features required for intrusion scenario specifications will be provided. Tools will be built for intrusion scenario description and automatic signature generation from the scenarios. Finally, intrusion detection systems will be developed to detect any intrusions into the software systems by comparing the run-time behavior of the systems with the generated signatures.    This research will play a vital role in bridging the gap between software engineering and security engineering practices. One of the unique aspects of this research proposal is to provide a balanced training facility for graduate students (2 PhD and 2 MSc) including both software engineering and security engineering principles for building secure software products.