A Joint Routing and Time-Slot Scheduling Load Balancing Algorithm for In-Vehicle TSN

To address the real-time transmission challenges of multiple types of data frames in Advanced Driver Assistance Systems, this paper proposes an in-vehicle Ethernet transmission technology based on Time-Sensitive Networking. In the ADAS environment, existing shortest-path-based routing algorithms often lead to load imbalance, and time-slot scheduling based on fixed routes fails to ensure the reliable transmission of diverse data streams in vehicular networks. To overcome these issues, we propose a joint routing and time-slot scheduling algorithm that effectively expands the feasible space for route planning, better balances network load, and improves time-slot allocation. This enhances the reliability and stability of network transmissions. Firstly, we construct various network topologies and load scenarios based on the complexity of ADAS systems. Secondly, we design a routing algorithm based on the K Shortest Path algorithm, where the path length and load are used to form a cost function to evaluate candidate transmission paths, selecting the optimal one. Finally, we develop an integer linear programming model based on constraint space and use a genetic algorithm to solve the time-slot allocation problem, ensuring deterministic data stream transmission. Experimental results show that compared to the shortest-path-based time-slot scheduling schemes, the proposed joint routing and time-slot scheduling algorithm can effectively achieve network link load balancing and end-to-end delay optimization for Time-Sensitive Networking in various network topologies and load scenarios. The proposed algorithm ensures the deterministic delivery of diverse data streams in TSN for in-vehicle scenarios, and through integration with 5G, it provides precise timing control and highly reliable data transmission for future vehicle connectivity and autonomous driving technologies. An open-source project is available on GitHub. You can view it by clicking here: https://github.com/mrxu190/TSN.git.