A proposal for managing the redundancy provided by the flexible time-triggered replicated star for ethernet

Distributed embedded systems that must operate continuously and satisfy unpredictable requirement changes must be reliable and flexible. Flexibility for changing realtime requirements may be provided by Hard Real-Time Ethernet Switching (HaRTES), an implementation of the master/multi-slave Flexible Time Triggered (FTT) communication paradigm. HaRTES relies on a single switch that embeds the FTT master and thus constitutes a single point of failure. Moreover, the trigger messages (TMs), which are periodic polling messages sent by the master to the slaves, are critical for the correct operation of any FTT implementation. Nevertheless, HaRTES has no mechanism to prevent their loss. To provide high reliability through fault tolerance, the Flexible Time-Triggered Replicated Star (FTTRS) duplicates the HaRTES switch and master, and modifies the masters to proactively retransmit the TMs. This spatial and temporal redundancy requires an appropriate management by the slaves. This paper discusses how this management can be performed.