Safety Analysis of the Local Interconnect Network Protocol in the Context of ISO26262 2016-01-0095
In this work, we analyze the use of the Local Interconnect Network (LIN) bus (and some of its potential variants) as Safety Element out of Context (SEooC) from an ISO-26262 perspective and provide the reader with an analysis methodology to compare between a range of different LIN protocol configurations and benchmark them against Automotive Safety Integrity Level (ASIL) targets as defined in ISO-26262.
A methodology for a quantitative residual failure probability analysis is shown before applying it to the standard LIN protocol. The residual failure rate in time (RF) of LIN (compliant with ISO26262) has been investigated with a range of reasonable application assumptions. This paper shows that a high bit error probability assumption of 3e-5 yields an RF of 3e-4/h which is too high to satisfy the assumed ASIL-B target (1e-7/h) or higher functional safety requirements in noisy application. However, an additional safety mechanism such as the ISO-26262 proposed CRC-8 with 0×97 polynomial is studied and is found to yield a reduced RF of ∼3e-19/h. As well, change in the assumptions such as a lower bit error probability of 1e-6 is found to yield a reduced RF of 4e-10/h. These results indicate that, based on the assumptions, the standard LIN may not meet the ASIL-B RF targets. Refining the assumptions and conducting a system level analysis is definitely required as future work to establish a more accurate residual fault assessment.