Optimized scheduling of multicore ECU architecture with bio-security CAN network using AUTOSAR

Abstract

Abstract In recent days, the developments in automobiles are a tremendous one. Especially, in the cars, the developments are increasing day by day. Initially, the operations of the car are using manual one but now various features are incorporated and moving towards the autonomous driving mode. For all these developments and working a control unit is needed to operate properly which is called Electronic control units in an automobile. These ECUs are different for different operations. The ECUs are increasing in parallel along with the developments in the car. But, the more number of ECUs implementations results in a high overhead communication and device is large in nature. To minimize the ECUs as well as to perform the operations effectively the core concept is introduced. In the core, the more no of electronic control units are replaced by runnable and it is operated in a real-time operating system. The single core effectively operated but when a high priority operation or interrupt occurs it performance reduces due to the single core architecture. To improve the performance of device a migration of single core architecture to multicore architecture takes place. The multicore architecture provides a parallel execution of different tasks along with the interrupts without any interventions. Several algorithms like First in First out, Round robin, etc., were introduced to improve the scheduling of operations. All these concepts were based on the task periods or priority. Hence to improve and provide a better scheduling algorithm this paper introduced an optimization concept based on the utilization factor using the Satin Bowerbird algorithm. The performance is analysed through the controller wait time and utilization time.