https://www.mdu.se/

Too large amount of jitter in a Controller Area Network in a vehcle can be very problematic since these systems are real-time embedded systems and therefore are time critical. It is a very important issue to investigate since message delays can be very catastrophic. We are performing pre-runtime and post-runtime analysis on Volvo's CAN network to find potential causes of jitter. We develop a jitter analysis framework for analysis with the help of a System Development method. In pre-runtime we are performing Response Time Analysis with the help of the industrial tool Rubus-ICE to find the worst-case response-time and the worst-case jitter. A pre-runtime analysis is done because the Controller Area Network is a hard embedded real-time system and therefore is vastly time-critical and a real-time system needs to be predictable. In post-runtime we are calculating simulated meassured response times and jitter with the assistance of CANalyzer and Python. Post-runtime is to analyze the results from runtime - in the case of this thesis it is simulated meassured response times, both best-case and worst-case, and also the maximum amount of jitter. Lastly, we are evaluating the findings and providing conclusions on how to mitigate the jitter and come around the problem of messages missing deadlines.