A proposal for master replica control in the flexible time-triggered replicated star for ethernet

Traditionally, distributed embedded systems have been designed using static approaches, i.e., assuming a static environment. However, when the environment is dynamic and imposes changing requirements on the system, flexible approaches are needed. If in addition the system must operate continuously, then high reliability is also required. The necessary flexibility for real-time requirements may be provided by Hard Real-Time Ethernet Switching (HaRTES), which is an implementation of the master/multi-slave Flexible Time Triggered (FTT) communication paradigm. To help provide high reliability, the Flexible Time-Triggered Replicated Star (FTTRS) adds network fault tolerance by duplicating the HaRTES switch and master, as well as the network links. This paper discusses how to enforce replica determinism for the masters in FTTRS by ensuring that they show corresponding outputs both in the time and value domain. Moreover, it suggests how to proceed in case replica determinism is lost despite all efforts.