Towards an Experimental Assessment of the Slave Elementary Cycle Synchronization in the Flexible Time-Triggered Replicated Star for Ethernet David Gessner, In´ es ´ Alvarez, Alberto Ballesteros, Manuel Barranco, Juli´ an Proenza DMI, Universitat de les Illes Balears, Spain {davidges, ines.alvarez.91}@gmail.com {a.ballesteros, manuel.barranco, julian.proenza}@uib.es Abstract—The communication subsystem of distributed em- bedded systems (DES) that must operate continuously and satisfy unpredictable requirement changes must be reliable and flexible. Recently the Flexible Time-Triggered Replicated Star for Ether- net (FTTRS) has been proposed as a communication subsystem that satisfies these two attributes. It is based on the master/multi- slave Flexible-Time Triggered (FTT) communication paradigm and relies on two custom switches, each with its own embedded FTT master. Both masters are active simultaneously and provide the same service. Specifically, they simultaneously and periodi- cally broadcast so-called trigger messages (TMs) in a redundant manner to make them robust to transient channel faults. One of the functions of these TMs is to divide the communication time into rounds called elementary cycles (ECs). For the correct operation of FTTRS, it is important that all slaves agree when each EC starts and ends. A mechanism to achieve this has been recently proposed. This paper presents a first implementation of this mechanism and a series of experimental tests that constitute a first step towards building a prototype of an FTTRS network. I. I NTRODUCTION A distributed embedded system (DES), to operate con- tinuously while satisfying unpredictable requirement changes, must be both highly reliable and flexible. To achieve this it requires a communication channel that satisfies those attributes as well. The goal of the Flexible Time-Triggered Replicated Star for Ethernet (FTTRS) [1] is to provide such a channel for a project called Fault Tolerance for Flexible Time-Triggered Ethernet-based systems (FT4FTT), which aims to provide high reliability and flexibility to all crucial parts of a DES. FTTRS is based on a switched Ethernet implementa- tion of the Flexible Time Triggered (FTT) communication paradigm [2], a paradigm that provides master/multi-slave communication in a way that allows the communication to adapt to changing real-time requirements. FTTRS attempts to make such communication highly reliable for switched ethernet by using fault tolerance. Its architecture is shown in Figure 1. The main components are two interconnected custom ethernet switches, each of which embeds an FTT master, and a set of FTT slaves connected to both of them. The embedded masters broadcast a periodic message called trigger message (TM), which divides the communication time into rounds of fixed duration called elementary cycles (ECs). Specifically, each EC begins with a trigger message window (TM window) in which each one of the two embedded masters broadcasts several redundant TMs to the slaves while no other traffic is exchanged on the network. The number of TMs Switch 2 (master 2) Switch 1 (master 1) Slave A Slave B Slave C ethernet link Fig. 1. FTTRS architecture. broadcast by each master in each EC is given by a parameter k, which is a function of the bit error rate of the channel. Moreover, the broadcasts are synchronized such that when one master transmits its nth TM of a given TM window, the other transmits its nth TM of the same TM window quasi- simultaneously [3]. In other words, the TM transmissions of the two masters occur in lockstep. For FTTRS to function correctly, the slaves must agree when each EC begins and ends. Since we want FTTRS to be highly reliable, we have recently proposed a mechanism to achieve this even if due to channel faults each slave fails to receive all but one TM per TM window [4]. In this paper we present a first implementation of this mechanism and a series of tests to check that the implementation is correct. Moreover, we provide some first results regarding the viability of achieving a precise EC synchronization in practice with the mechanism. The remainder of the paper proceeds as follows. Section II summarizes the EC synchronization mechanism used by the slaves. Section III describes our implementation of the EC synchronization mechanism. Section IV describes the tests we performed and the results we obtained. Finally, Section V concludes the paper and points to future work. II. THE SLAVE EC SYNCHRONIZATION MECHANISM This section summarizes the slave EC synchronization mechanism that was first presented in a previous paper [4]. As mentioned in the introduction, of the k TM replicas broadcast by each master, the corresponding slave might receive all k replicas or only a subset of them due to transient faults. Regardless of which specific replicas each slave receives on each of its links, the time instants when the slaves consider each EC to start and end must align. This can be achieved by c 2014 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprint- ing/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.