Configuration and Timing Analysis of TSN-based Distributed Embedded Systems

Sammanfattning: The set of IEEE Time-Sensitive Networking (TSN) standards is an emerging candidate for backbone communication in modern applications of real-time distributed embedded systems. TSN provides various traffic shaping mechanisms that aim at managing the timing requirements of traffic. Emerging applications of these systems, particularly in the automotive domain, often run complex distributed software that requires low-latency and high-bandwidth communication across multiple onboard electronic control units. Using TSN in these systems introduces multiple challenges. Specifically, the developers of these systems face a lack of development techniques and tools, as TSN standards only offer general recommendations for the use of its features and mechanisms. There is an urgent need for development techniques, tools, and methods to assist the developers in effectively leveraging the features outlined in TSN standards. In this thesis, we identify and address several challenges encountered in the development of TSN-based distributed embedded systems, particularly focusing on the stages of system and software modeling, network configuration, and timing analysis. The overall goal of this thesis is to support the development of these systems in the aforementioned stages while considering the Quality of Service (QoS) requirements of all traffic classes in TSN. We present techniques to facilitate the system and software modeling of TSN-based distributed embedded systems. These techniques enable performing timing analysis in the early stages of system and software development. In the stage of network configuration, we propose techniques for managing the configuration complexity and supporting the automatic configuration of mechanisms in TSN. The proposed configuration techniques consider achieving acceptable QoS in various traffic classes. In the stage of timing analysis, we address the challenges of incorporating various TSN traffic classes and mechanisms by extending the existing timing analyses. The results indicate that the proposed techniques effectively facilitate the system and software modeling, network configuration, and timing analysis of TSN-based distributed embedded systems.