Cargando…

A Functional Verification Methodology for Highly Parametrizable, Continuously Operating Safety-Critical FPGA Designs: Applied to the CERN RadiatiOn Monitoring Electronics (CROME)

Electronic systems that are related to human safety need to comply to strict international standards such as the IEC 61508. We present a functional verification methodology for highly parametrizable, continuously operating, safety-critical real-time systems implemented in FPGAs. It is compliant to I...

Descripción completa

Detalles Bibliográficos
Autores principales: Ceesay-Seitz, Katharina, Boukabache, Hamza, Perrin, Daniel
Lenguaje:eng
Publicado: 2020
Materias:
Acceso en línea:https://dx.doi.org/10.1007/978-3-030-54549-9_5
http://cds.cern.ch/record/2740501
Descripción
Sumario:Electronic systems that are related to human safety need to comply to strict international standards such as the IEC 61508. We present a functional verification methodology for highly parametrizable, continuously operating, safety-critical real-time systems implemented in FPGAs. It is compliant to IEC 61508 and extends it in several ways. We focus on independence between design and verification. Natural language properties and the functional coverage model build the connection between system safety requirements and verification results, providing forward and backward traceability. Our main verification method is Formal Property Verification (FPV), even for Safety Integrity Level 1 and 2. Further, we use constrained-random simulation in SystemVerilog with the Universal Verification Methodology and a design independent C reference model. When faults are discovered, the coverage model is extended to avoid regressions. Automation allows the reproduction of results and the reuse of verification code. We evaluate our methodology on a subset of the newly developed CERN RadiatiOn Monitoring Electronics (CROME). We present the challenges we faced and propose solutions. Although it is impossible to simulate the full design exhaustively, several formal properties have been fully proven. With FPV we found some safety-critical faults that would have been extremely hard to find in simulation.