SHF: Medium: Simulation-Based Analysis of EM Side Channels in Embedded Systems: From Software to Fields

嵌入式计算系统代表了社会日常生活所依赖的基础设施的关键骨干。它们通常在安全关键环境中对敏感数据进行操作，例如运输，医疗保健和工业控制。保护嵌入式电子系统的一个重要方面是防止系统行为级别的信息泄漏。存在多个场所，攻击者可以通过这些场所获得这些信息，即使没有完全访问内部系统组件。具体地，在电子系统中的软件活动期间发出的电磁（EM）场包含关于底层系统行为的重要信息。该项目将开发系统设计人员可以使用的工具，以确保对通过EM侧渠道泄露的信息的弹性。作为推广和教育部分的一部分，该项目将开发一个关于物理和物联网安全的短期课程，该课程将使用跨学科的网络安全挑战，涵盖密码学，电磁学，嵌入式系统和硬件/软件设计主题，以激发追求科学和技术职业的兴趣。EM侧通道分析的唯一现有能力是基于繁琐的实验室测量。这是一个根本问题，因为大多数嵌入式软件开发人员无法对EM漏洞进行基于实验室的表征。这种表征需要典型软件工程师缺乏的仪器专业知识，价格昂贵，提供有限的可观察性，并且与计算工具流相比缓慢。该项目将开发新的基于仿真的方法，用于表征在复杂的嵌入式处理器上运行的软件所产生的电磁场。这将需要快速而准确地跟踪时空EM辐射，从电路到处理器微架构再到在其上运行的软件。该仿真工具流程将允许设计改进的系统级防御恶意软件，以及通过编译器和微架构级别的嵌入式硬件/软件协同设计最大限度地减少侧通道泄漏。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。

Embedded computing systems represent a critical backbone of infrastructure that society relies on in daily life. They often operate on sensitive data in safety-critical environments, such as transportation, health care, and industrial control. An important aspect of securing embedded electronic systems is to prevent information leakage at the level of system behavior. There exist multiple venues by which an attacker can get such information even without having complete access to the internal system components. Specifically, electromagnetic (EM) fields emanated during software activity in electronic systems contain important information about the underlying system behavior. This project will develop tools that system designers can use to ensure resilience against information being leaked through EM side channels. As part of outreach and education components, this project will develop a short course on physical and IoT security that will use interdisciplinary cybersecurity challenges spanning cryptography, electromagnetics, embedded system, and hardware/software design topics to generate interest in pursuing careers in science and technology.The only existing capability for EM side-channel analysis is based on cumbersome laboratory measurements. This is a fundamental problem, as most embedded software developers are not in a position to carry out lab-based characterization of EM vulnerabilities. Such characterizations require instrumentation expertise that a typical software engineer lacks, are expensive, provide limited observability, and are slow compared to what a computational tool flow enables. This project will develop novel simulation-based approaches for characterization of EM fields emanated by software running on complex embedded processors. This will require fast yet accurate tracking of spatio-temporal EM emanations all the way from circuits to processor micro-architectures to software running on them. The simulation tool flow to be developed will permit design of improved system-level defenses against malware as well as minimization of side-channel leakage through embedded HW/SW co-design at both compiler and micro-architecture levels.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.