Safety-oriented mixed-signal verification of automotive power devices in a UVM environment

The complexity of automotive applications is continuously increasing, leading to a growing demand for methodologies that offer comprehensive mixed-signal verification. However, compared to the highly automated verification methodologies in the digital domain, pre-silicon verification in the analog domain still implies a substantial amount of manual work and computational effort. Apart from this, automotive applications most often have to comply with functional safety standards and therefore their robustness concerning safety-critical faults needs to be proven. This is normally ensured by performing safety verification with faults being purposefully injected into the designs. In this paper we present a methodology that enables a regression-based mixed-signal verification combined with an existing approach for analog safety analysis. Both concepts are applied to an automotive design, in which faults have been injected, in order to demonstrate their capabilities.

[1]  Wolfgang Ecker,et al.  A transaction-oriented UVM-based library for verification of analog behavior , 2014, 2014 19th Asia and South Pacific Design Automation Conference (ASP-DAC).

[2]  Cristian Tepus,et al.  Constraint random stimuli and functional coverage on mixed signal verification , 2014, 2014 International Semiconductor Conference (CAS).

[3]  Peter R. Wilson,et al.  The Circuit Designer's Companion , 2012 .

[4]  L. Devroye Non-Uniform Random Variate Generation , 1986 .

[5]  Linus Maurer,et al.  Towards simulation based evaluation of safety goal violations in automotive systems , 2014, Proceedings of the 2014 Forum on Specification and Design Languages (FDL).