NSF Award
1623253
One of the main obstacles in providing extensive hands-on experience in cybersecurity classes is the substantial amount of manual work involved in creating and grading the exercise. Combined with the frequent need to update the exercises, this obstacle effectively limits that amount of hands-on work that gets incorporated into cybersecurity education. This project seeks to eliminate such barriers, and to greatly improve the efficiency of the educational process by automating the most time-consuming tasks. <br/><br/>This project makes two main contributions to cybersecurity education: the development of a specification-driven, dynamic environment for implementing realistic cyber defense and forensic analysis exercises; and the advanced support for class management and automated evaluation. The platform, AutoCUE, provides a high-level specification language, and an execution runtime that enable instructors to easily and efficiently run realistic scenarios that result in customized environments; based on the same methods, the system also be used to automatically create of realistic experimental data sets. The infrastructure provides an automated class management component, which consists of: a) deployment automation module, which guarantees consistent student lab environment, and central control by the instructor; b) scenario personalization module, which can generate customized exercises for each student (for evaluation purposes); and c) automated grading module, which combines ideas from capture-the-flag competitions and environment sensors to track student progress and automate the grading process. The project also provides ready-to-use seed content for two classes: digital forensics and network penetration testing.
