https://www.mdu.se/

Controller Area Network (CAN) is an ISO-standardized protocol that is extensively used as the onboard real-time network in vehicles. Evaluating and mitigating responsetime jitter experienced by CAN messages is paramount, particularly in real-time applications where timing predictability is crucial. This paper presents a jitter analysis framework for CAN messages. Within this framework, the worst-case responsetime jitter of CAN messages is calculated through pre-runtime analysis. Additionally, a post-runtime analysis of message traces helps ascertain the maximum response-time jitter these messages encounter at runtime. The framework performs a comparative evaluation of the pre-runtime and runtime response-time jitter and then back-propagates insights and guidelines to the systems' designers to mitigate the jitter. To validate this approach, we implement the framework using two prominent industrial tools: Rubus-ICE and CANalyzer. The effectiveness of the framework is further substantiated by modeling and analyzing an industrial use case provided by Volvo Construction Equipment.