Collaborative Research: EAGER: SaTC-EDU: Dynamic Adaptive Machine Learning for Teaching Hardware Security (DYNAMITES)

合作研究：EAGER：SaTC-EDU：用于教学硬件安全的动态自适应机器学习 (DYNAMITES)

• 批准号: 2039610
• 负责人: Jeyavijayan Rajendran
• 金额: $ 15万
• 依托单位: Texas A&M Engineering Experiment Station
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2020
• 资助国家: 美国
• 起止时间: 2020-09-01 至 2023-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=2039610&HistoricalAwards=false
• 关键词: Collaborative; Research; EAGER; SaTC; EDU

Cybersecurity is key to safeguarding societal wellbeing in the present digital era. As threats at the hardware level become more prevalent, skills and knowledge for hardware security become more crucial for cybersecurity professionals. In addition, the rise of artificial intelligence (AI) promises to rapidly evolve the threat landscape. To prepare the next-generation cybersecurity workforce, students need opportunities to hone their skills on a variety of different hardware security problems. Current curriculum on hardware security focuses on theory and a small number of hand-crafted exercises, thus providing limited opportunity to apply learning to evolving scenarios. To address these drawbacks, this project intertwines AI and hardware security to develop new tools for preparing students to be creative and flexible, and ultimately, better prepared for dealing with newly emerging hardware security threats.To improve the state-of-the-art in hardware security and cybersecurity education, this project is seeking new insights at uncharted intersections of hardware security and AI-based decision making. The project will investigate how to enable students to attack and defend hardware by sparring against DYNAMITES, which is a dynamic adaptive machine learning tool for teaching hardware security. The project will also examine hardware security pedagogy to understand the impact of the tool in shaping students’ cognitive processes. The major goal is to develop and evaluate DYNAMITES through research in three directions: (1) investigating and adapting techniques to allow AI to understand hardware, (2) exploring how AI can be used to produce new problems intelligently, and (3) exploring how AI in the learning environment affects the "security mindset" in students. These findings will allow hardware security education to scale, reducing the barrier to entry and arming future professionals with the skills needed to protect critical systems, as well as jump-starting innovations in automated, scalable scanning and patching of hardware vulnerabilities. The hardware attack/defense artifacts emerging from this project will be released for use in teaching and research, and the project team will disseminate tools/techniques that emerge from this project.This project is supported by a special initiative of the Secure and Trustworthy Cyberspace (SaTC) program to foster new, previously unexplored, collaborations between the fields of cybersecurity, artificial intelligence, and education. The SaTC program aligns with the Federal Cybersecurity Research and Development Strategic Plan and the National Privacy Research Strategy to protect and preserve the growing social and economic benefits of cyber systems while ensuring security and privacy.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

网络安全是当今数字时代维护社会福祉的关键。随着硬件层面的威胁变得越来越普遍，硬件安全的技能和知识对于网络安全专业人员来说变得更加重要。此外，人工智能 (AI) 的兴起有望迅速演变威胁格局。为了培养下一代网络安全劳动力，学生需要有机会磨练他们在各种不同硬件安全问题上的技能。当前的硬件安全课程侧重于理论和少量手工练习，因此将学习应用于不断变化的场景的机会有限。为了解决这些缺点，该项目将人工智能和硬件安全结合起来，开发新工具，帮助学生培养创造力和灵活性，并最终为应对新出现的硬件安全威胁做好更好的准备。为了提高硬件安全和网络安全教育的最新水平，该项目正在硬件安全和基于人工智能的决策的未知交叉点寻求新的见解。该项目将研究如何让学生通过对抗 DYNAMITES 来攻击和防御硬件，DYNAMITES 是一种用于教授硬件安全的动态自适应机器学习工具。 该项目还将研究硬件安全教学法，以了解该工具对塑造学生认知过程的影响。主要目标是通过三个方向的研究来开发和评估 DYNAMITES：（1）研究和调整技术，让人工智能能够理解硬件，（2）探索人工智能如何智能地产生新问题，以及（3）探索学习环境中的人工智能如何影响学生的“安全心态”。这些发现将使硬件安全教育得以规模化，降低进入门槛，为未来的专业人员提供保护关键系统所需的技能，并在自动化、可扩展的扫描和硬件漏洞修补方面推动创新。该项目产生的硬件攻击/防御工件将被发布用于教学和研究，项目团队将传播该项目产生的工具/技术。该项目得到安全可信网络空间（SaTC）计划特别倡议的支持，旨在促进网络安全、人工智能和教育领域之间新的、先前未探索的合作。 SaTC 计划与联邦网络安全研究与发展战略计划和国家隐私研究战略相一致，旨在保护和维护网络系统不断增长的社会和经济效益，同时确保安全和隐私。该奖项反映了 NSF 的法定使命，并通过使用基金会的智力价值和更广泛的影响审查标准进行评估，被认为值得支持。

项目成果

ATTRITION: Attacking Static Hardware Trojan Detection Techniques Using Reinforcement Learning

• DOI: 10.1145/3548606.3560690
• 发表时间: 2022-08
• 期刊: Proceedings of the 2022 ACM SIGSAC Conference on Computer and Communications Security
• 作者: Vasudev Gohil;Hao Guo;Satwik Patnaik;Jeyavijayan Rajendran

DETERRENT: detecting trojans using reinforcement learning

威慑：使用强化学习检测木马

• DOI: 10.1145/3489517.3530518
• 发表时间: 2022
• 期刊: IEEE Design Automation Conference
• 作者: Gohil, Vasudev;Patnaik, Satwik;Guo, Hao;Kalathil, Dileep;Rajendran, Jeyavijayan
• 通讯作者: Rajendran, Jeyavijayan

Reinforcement Learning for Hardware Security: Opportunities, Developments, and Challenges

硬件安全的强化学习：机遇、发展和挑战

• DOI: 10.1109/isocc56007.2022.10031569
• 发表时间: 2022
• 期刊: IEEE International SoC Design Conference (ISOCC
• 作者: Patnaik, Satwik;Gohil, Vasudev;Guo, Hao;Rajendran, Jeyavijayan JV
• 通讯作者: Rajendran, Jeyavijayan JV