CRII: SaTC: Automated Security Analysis of Software-Based Control in Emerging Smart Transportation Under Sensor Attacks

随着智能交通信号控制、自动驾驶等一系列基于软件的智能交通解决方案的出现，交通系统正在发生深刻变革。在这些系统中，其功能智能的关键促成因素是传感能力，它可以收集必要的道路信息，以实现更好的控制决策。然而，传感器数据是从公共通道（即物理运输环境）收集的，因此不可避免地为攻击者篡改传感过程创造了机会。在这个项目中，研究者发起了对新兴智能交通系统中基于软件的控制在传感器攻击下的稳健性的系统理解的研究工作。由于交通是一项基本的城市功能，因此本项目的成功完成有望在广泛部署之前主动识别和解决这些系统中的新安全挑战，从而有助于确保日常生活的安全与安全。在这个项目中，将设计、实施和评估一种自动化方法，以便在新兴的智能交通系统中高效地发现安全问题。该方法侧重于发现语义安全问题，例如智能交通功能中改变控制决策的问题。这是因为分析它们通常会导致在一个领域甚至跨领域中发现新的和特定于领域的设计权衡，从而更有可能推进知识和指导未来的研究。为了实现这一目标，本项目将设计一个遵循进化算法方法的动态安全分析系统，并提出新的解决方案，以解决分析输入生成、语义问题发现和分析以数据流为中心的决策过程中的研究挑战。本项目将开发的系统具体应用于现实世界的智能交通系统代码库，以确保发现的安全问题、挑战和解决方向的实用性。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Transportation systems are being profoundly transformed with the emergence of a series of software-based smart transportation solutions such as intelligent traffic signal control and autonomous driving. In these systems, the key enabler of their functional intelligence is the sensing capability, which collects necessary road information to enable better control decisions. However, sensor data are collected from a public channel, i.e., the physical transportation environment, which thus inevitably creates opportunities for attackers to tamper with the sensing process. In this project, the investigator initiates research efforts towards achieving a systematic understanding of the robustness of software-based control in emerging smart transportation systems under sensor attacks. Since transportation is a basic urban function, the successful completion of this project is expected to proactively identify and address new security challenges in these systems before wide deployment, and thus help ensure the security and safety of everyday life.In this project, an automatic approach will be designed, implemented, and evaluated to enable efficient and effective discovery of security problems in emerging smart transportation systems. The approach focuses on discovering semantic security problems, which for example change control decisions in the smart transportation functions. This is because analyzing them usually leads to the discovery of novel and domain-specific design trade-offs in an area or even across areas, which thus has more potential to advance knowledge and guide future research. To achieve this goal, this project will design a dynamic security analysis system following an evolutionary algorithm approach, and propose novel solutions to address research challenges in analysis input generation, semantic problem discovery, and analyzing dataflow-centric decision process. This project will concretely apply the developed system to real-world code bases of smart transportation systems to ensure the practicality of the discovered security problems, challenges, and solution directions.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.