Functional Safety Evaluation for Cybersecurity Measures in Low-End Automotive Control Units

Cybersecurity and functional safety are currently quite complementary disciplines for the development of future electronic and electrical components in vehicles. To protect vehicles from emerging adversaries, novel cybersecurity measures must be deployed to vehicle components. These measures must not only address potential threats adequately but must also ensure safe functional operation when random faults occur to reduce hazards to passengers and traffic participants. While high-end control units have sufficient processing capabilities to integrate established cybersecurity measures, low-end processing units also carrying out safety-relevant functions present a gap to cybersecurity. It is in these low-end control units where less established procedures such as lightweight authentication schemes, as proposed in our earlier work, can play a crucial role in providing adequate cybersecurity measures. By applying the lightweight authentication with our previously elaborated parity prediction scheme for low-end control units, soft errors caused by single event upsets can be detected, ensuring the reliability of cybersecurity functions and safety-relevant functions. In this work, our contribution presents a quantitative functional safety evaluation of the parity prediction enabled lightweight authentication scheme used as a safety mechanism improving functional safety random fault metrics of an automotive control unit in a rear lighting system. This evaluation demonstrates the effectiveness of integrating novel safety measures for cybersecurity functionality into existing functional safety evaluation schemes. Finally, we provide an outlook on how the impact of soft errors will become more relevant for denser IC technology sizes in electronic and electrical components designed for future vehicles.