Fault recovery of 10BASE-T1S automotive ethernet with bus/ring hybrid topology

Abstract

Automotive Ethernet has emerged as a communication protocol to meet the escalating demands in applications such as advanced driver assistance systems (ADAS), autonomous driving, and connected vehicles. This technology surpasses the capabilities of the traditional controller area network (CAN) by providing higher bandwidth and supporting bidirectional communication, which is essential for real-time applications and the easy expansion of networks through Ethernet switches. Furthermore, automotive Ethernet guarantees packet transmission within the bounded latency, achieving low packet latency and jitter, and low packet loss through time-sensitive networking (TSn), a set of standards within the IEEE 802 network, which ensures the operation of time-critical applications. TSN also helps for automotive Ethernet, which requires reliability to deliver messages from time-critical applications without errors or loss. Among TSN standards, frame replication and elimination for reliability (FRER) stands out as a technology that provides redundancy. Although FRER is useful in detecting data loss, it consumes significant bandwidth and works only in a mesh topology. Therefore, it is difficult to be used in the bus type automotive Ethernet such as 10BASE-T1S. The bus topology presents a problem wherein if a segment of the cable becomes damaged, communication throughout the entire network becomes impossible. In this paper, we propose a fault-tolerant redundancy protocol for within the 10BASE-T1S bus topology. Our method involves adapting the physical layer collision avoidance (PLCA) protocol, originally utilized in the 10BASE-T1S bus topology, for operation in a ring-type connection. This approach offers the advantage of maintaining compatibility with the existing PLCA protocol in use. We verify the effectiveness of our proposed method through simulations and hardware emulation, confirming its ability to restore network functionality in the event of hard faults, such as cable disconnections and node failures.