Schedulability Analysis for Messages in Gateway-Interconnected Controller Area Network

Abstract

Controller Area Network (CAN) is widely used inside the automobiles. To decrease design complexity and cost, several CAN are utilized in different sub systems and gateway is employed to realize the communication between them. But the employment of gateway brings many challenges for schedulability analysis of CAN messages. We first analyzed the key challenges for schedulability analysis of messages. And then, based on existing method proposed for one single CAN, a new analysis method that is based on interference analysis for dynamically arriving gateway messages is proposed for non-gateway messages. Furthermore, a division approach that transforms the end-to-end schedulability analysis of gateway messages into the similar situation with that of non-gateway messages is proposed for gateway messages. The proposed method is proved to be safe, and by comparing it with a full space searching based simulator and applying to a real message set, its effectiveness is demonstrated as well.