Integrating OSEK RTOS in L2 virtual ECU for simulating timing behavior of automotive systems

Abstract

Modern vehicles rely on Electronic Control Units (ECUs) to manage safety-critical functions such as engine control and braking. To reduce hardware dependency and accelerate development, Virtual ECUs (vECUs) are used for early testing, though many lack accurate real-time behavior. This thesis upgrades a Level 2 vECU to Level 3 by integrating Trampoline, an open-source OSEK-compliant real-time operating system (RTOS), enabling fixed-priority preemptive scheduling. The modified vECU runs in a real-time Linux environment and is evaluated against a production ECU using timing metrics like execution time and response time. Two response time analysis (RTA) models—classic and offset-aware—are applied to predict worst-case response times, which are then validated against empirical measurements. Since RTA is essential for ensuring tasks meet deadlines in automotive systems, validating these models enhances their practical utility. The results show that the upgraded vECU can simulate timing behavior with reasonable accuracy, supporting cost-effective, early-stage validation in automotive software development.