Recent years, transformation in automotive industry have a huge technology growth. Flexray, Control Area Network (CAN), Local Interconnect Network (LIN), CAN-FD, Automotive Ethernet are the principal automotive communication protocols in industry exercise. CAN protocol is primely used among the automotive communication protocols. But such predominantly used CAN protocol doesn‘t have clear appropriateness of overhead, latency and deterministic performance behavior. In order to analyze these complex uncertainty behavior, Electronic Control Unit’s (ECU) dynamic architecture of communication system and the CAN bus load behavior should be identified. Here, we used AUTOSAR based communication stack. This abstract deals with deterministic timing analysis of CAN protocol by which we can optimize it for having potential End-to-End (E2E) runtime communication for a dedicated event chain. The analysis was targeted by two approaches – (a) Timing Simulation and (b) prove of simulation results measurement of a real implementation. In a simulated environment, the complete dynamic behavior of the ECU, the CAN bus behavior, the trigger and flow of event chain was simulated. For measurement in the real implementation, time bases of the involved ECUs had been synchronized using the “synchronized time base manager” feature of AUTOSAR, the relevant software events (interrupts, task, function calls …) contributing to the event chain had been identified and measured , CAN bus load was injected by appropriate CAN bus simulators. From these two approaches, detailed time analysis was made to understand the communication architecture behavior in ECUs and the CAN bus matrix system involved. This timing analysis were plotted for several iterations to understand the mutual pattern and to derive exact behavior of our need. This analysis results helped to create an optimized E2E system and improve the simulation results.
Keywords: Bus load, Latency, Communication Overhead, Dynamic Architecture, Determinism