Redundant Time Synchronization Mechanism Using Kalman filter in Train Communication Network

The time sensitive network(TSN) has been implemented in the commercial high-speed railway in the current years. Robust time synchronization is one of the key TSN technologies to fulfill the requirement of huge data traffic. In this paper, the fault-tolerant mechanism for master-slave time offset has been analyzed. To address this issue, an improved Kalman-based redundant system is proposed to optimize the time offset. Furthermore, the validation is delivered in OMNeT++ with a real topology recommended by Shift2rail. The simulation results indicate that the predictor can detect fault from the master within one cycle. It illustrates that the system in principle conforms to the time synchronization process defined by the gPTP protocol and guarantees the fault tolerance of the master-slave clock synchronization process.