Time-sensitive networking for interlock propagation in the IFMIF-DONES facility

• Published in: Fusion engineering and design Vol. 191; p. 113774
• Main Authors: Megías, Carlos, Sánchez-Garrido, Jorge, Vázquez, Víctor, Ros, Eduardo, Cappelli, Mauro, Díaz, Javier
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.06.2023
• Subjects: Convergent networks; Determinism; IFMIF-DONES; Interlock; Machine protection system; Time-sensitive networking; Convergent networks; IFMIF-DONES; Determinism; Time-sensitive networking; Interlock; Machine protection system
• ISSN: 0920-3796; 1873-7196
• DOI: 10.1016/j.fusengdes.2023.113774

•First study of Time-Sensitive Networking for interlocks in particle accelerators.•Architecture for the propagation of interlock signals with different response times.•Promising experimental results of latency and determinism of interlock signals.•Proposed interlock propagation network for IFMIF-DONES.•Novel application of convergent networks for high-energy physics. In this study, we have proposed the use of time-sensitive networking (TSN) technologies for the distribution of the interlock signals of the machine protection system of the future IFMIF-DONES particle accelerator, required for implementing the protection mechanisms of the different systems in the facility. Such facilities usually rely on different fieldbus technologies or direct wiring for their transmission, typically leading to complex network infrastructures and interoperability problems. We provide insights of how TSN could simplify the deployment of the interlock network by aggregating all the traffic under the same network infrastructure, whilst guaranteeing the latency and timing constraints. Since TSN is built on top of Ethernet technology, it also benefits from other network services and all its related developments, including redundancy and bandwidth improvements. The main challenge to address is the transmission of the interlock signals with very low latency between devices located in different points of the facility. We have characterized our initial TSN architecture prototype, evaluated the latency and bandwidth obtained with this solution, identified applications to effectively shape the attainable determinism, and found shortcomings and areas of future improvements.