Towards secure cyber-physical systems for autonomous vehicles
Document typeConference report
PublisherBarcelona Supercomputing Center
Rights accessOpen Access
All rights reserved. This work is protected by the corresponding intellectual and industrial property rights. Without prejudice to any existing legal exemptions, reproduction, distribution, public communication or transformation of this work are prohibited without permission of the copyright holder
Cyber-Physical systems have become ubiquitous. These systems integrate different functionalities to satisfy the performance requirements and take advantage of the available processing power of multi-core systems. Safety critical applications such as autonomous vehicles or medical devices rely not only on proving correct functionality of cyber-physical systems as essential certification criteria but they must also satisfy other design constraints such as energy efficiency, low power consumption and reliability. Their need to connect to the internet have created new challenges which means addressing the security vulnerabilities has become as the first-class design concern. In this talk, first a hardware/software co-design approach for two critical tasks, real-time pedestrian and vehicle detections, which are essential in advanced driving assistance systems (ADAS) and autonomous driving systems (ADS) is presented. We use partial dynamic reconfiguration on FPGA for adaptive vehicle detection. In the second part of this talk, a system-level security-aware design approach is presented to avoid or confine the impact of security compromises on the critical components of the cyber-physical systems implemented in multiprocessor systems on chip. Our system-level security approach considers the described system architecture for a specific application and analyzes its security vulnerability based on the specified security rules to generate an impact analysis report. Then, it creates a new system architecture configuration to protect the critical components of the system by providing isolation of tasks without the need to trust a central authority at run-time for heterogeneous multiprocessor system. This approach allows safe use of shared IP with direct memory access, as well as shared libraries by regulating memory accesses and the communications between the system components.
CitationBiglari-Abhari, M. Towards secure cyber-physical systems for autonomous vehicles. A: . Barcelona Supercomputing Center, 2020, p. 65-66.