Improper timing has severe safety repercussions in modern vehicles due to the use of mixed-criticality systems. Human lives and increased material costs might be at stake due to faulty configuration or incorrect startup of the timing subsystem. TTEthernet is a communication infrastructure that enables combined use of distributed applications with mixed-criticality requirements over a single physical network. TTEthernet provides a fault-tolerant, self-stabilizing synchronization that establishes a temporal partitioning and ensures isolation of high and low criticality dataflows. The cornerstone of robust synchronization is the cold start under faulty nodes since it is performed once before establishing the timing schedule. We propose a genetic simulation strategy applied to the single-fault analysis of TTEthernet synchronization protocol. Simulation measurements demonstrate the effectiveness of our approach and reveal the worst-case startup time of the TTEthernet network under a single-fault hypothesis. Furthermore, we reveal that the median of the worst-case startup time is almost 500 times larger than the integration cycle duration.

Finally, we argue that our algorithm has further application in the analysis of TTEthernet synchronization, like multiple fault scenarios and generalization to a more complex multi-hop TTEthernet topologies.

由于使用了混合临界系统，不正确的计时对现代车辆的安全性产生了严重影响。由于时序子系统的错误配置或不正确的启动，可能会危及生命和增加的材料成本。TTEthernet是一种通信基础结构，可在单个物理网络上组合使用具有混合关键性要求的分布式应用程序。以太网提供了一个容错，自稳定的同步，该同步建立了时间分区，并确保了高和低临界数据流的隔离。稳健同步的基石是故障节点下的冷启动，因为它在建立时序计划之前执行了一次。我们提出了一种遗传仿真策略，该策略适用于TTEthernet同步协议的单故障分析。仿真测量证明了我们方法的有效性，并揭示了在单故障假设下TTEthernet网络的最坏情况下的启动时间。此外，我们发现最坏情况下的启动时间的中位数几乎比积分周期的持续时间大500倍。

最后，我们认为我们的算法在TTEthernet同步分析中有进一步的应用，例如多个故障场景以及对更复杂的多跳TTEthernet拓扑的推广。