Grant Call 2a - List of Projects

Project Title:
Attack-Resilient AI-Empowered Autonomous Cyber-Physical Systems
Lead Principal Investigator:
Assistant Professor Tan Rui

Abstract:
Recent advances of artificial intelligence technologies enable autonomous cyber-physical systems (e.g., autonomous vehicles and unmanned aerial vehicles) to accomplish sophisticated tasks in real-world environment. However, the cyber components of these systems face increasing threats from adversaries. We will study both the internal and external attacks on sensor data, i.e., sensor data corruption by attackers who have obtained access to the system’s sensors and adversarial examples that introduce tiny changes to the system’s environment. We will design attack detection and thwarting approaches and develop their efficient implementations suitable for embedded hardware accelerators. Moreover, we will construct a testbed to capture representative cyber components of autonomous cyber-physical systems. It facilitates the design, evaluation, and demonstration of the threats and countermeasures. The testbed’s simulation environment allows others to test against the attack scenarios programmed in this project. The demonstration will reinforce the relevant industry’s awareness on the criticality of the studied threats.

Project Title:
Uncovering Vulnerabilities in Machine Learning Frameworks via Software Composition Analysis and Directed Grammar-Based Fuzzing
Lead Principal Investigator:
Associate Professor David Lo

Abstract:
Smart systems are increasingly dependent on machine learning (ML) frameworks, e.g., TensorFlow, for their feature implementation. These frameworks are built on top of many third-party libraries, which depend on many others. Simply trusting and reusing a framework poses a security risk as the framework and its direct and transitive dependencies can contain exploitable vulnerabilities. To mitigate this risk, this project will create an advanced software composition analysis solution that scans dependency hierarchies and builds new deep learning architectures to analyse code and document repository data and flag vulnerabilities. Further, the flagged vulnerabilities will be verified if it can be reached via our new directed grammar-based fuzzing solution that generates valid test cases (following predefined grammars) and drives test executions to vulnerable code. Our solution targets vulnerabilities hidden deep in ML framework dependencies, which are hard for a classic fuzzer to uncover and for framework developers to recognize as they appear in third-party code.

Project Title:
Over-the-air Security Testing of Wireless Protocol Implementations
Lead Principal Investigator:
Assistant Professor Sudipta Chattopadhyay

Abstract:
Wireless technology is one of the key factors for the progress towards a smart nation. Thus, it is crucial to investigate and improve the security of wireless devices for a secure smart nation. The project aims to develop technologies to systematically discover security vulnerabilities in current and next-generation wireless protocol implementations. Subsequently, it develops technologies for defending against unknown security attacks to protect a large number of smart devices employed in sectors such as medical, logistics, smart home and wearables, among others.