Reliability in Real-Time: Solving Timing Latency in SDV Networks
Researchers have introduced a cross-validated framework for analyzing the timing of Automotive CAN networks. This breakthrough addresses the critical need for deterministic communication in increasingly complex software-defined vehicle architectures.
As vehicles transition into software-defined entities (SDVs), the internal communication networks that link sensors, actuators, and electronic control units (ECUs) are under unprecedented strain. Researchers from National Yang Ming Chiao Tung University (NYCU) have published a groundbreaking paper on timing analysis for Controller Area Networks (CAN), the backbone of automotive communication.
The central challenge in SDVs is ensuring that safety-critical messages—such as braking commands or steering adjustments—reach their destination with absolute determinism. In older architectures, this was manageable. However, modern SDVs feature a flood of data from ADAS sensors and infotainment systems that can "clog" the network. The researchers' new framework utilizes Deterministic Stochastic Petri Nets (DSPN) combined with Worst-Case Response-Time (WCRT) analysis to predict and prevent network bottlenecks before they occur.
This framework allows SDV engineers to simulate how software updates will affect the physical performance of the car. By providing a cross-validated method to ensure message reliability, the research provides a vital safety net for the industry. As manufacturers roll out over-the-air (OTA) updates that alter vehicle behavior, having a mathematical guarantee that the underlying CAN bus can handle the bandwidth is not just a technical requirement—it is a safety imperative.
Source: Semiconductor Engineering
