Session P4b: Timekeeping and Quantum Networking

White Rabbit Status and Plans
Javier Serrano, Amanda Díez Fernández, Maciej Lipinski, CERN; Adam Wujek, Consultant
Location: Seaview A/B
Date/Time: Thursday, Jan. 30, 11:48 a.m.

White Rabbit (WR) is an open-source networking technology which allows synchronisation of thousands of distributed nodes with an accuracy better than 1 nanosecond. We describe recent advances in the evaluation of link asymmetry and constant delays which are taking this accuracy into the realm of tens of picoseconds. In terms of precision, the lowest jitter achieved in a WR node so far is below 100 femtoseconds, integrating phase noise between 100 Hz and 20 MHz.
Originally designed as a collaborative effort including several institutes and companies, WR technology was incorporated into the IEEE-1588 standard in its latest 2019 edition, under the so-called High Accuracy PTP Profile. As a result of this standardisation effort, we have seen quite an increase in uptake of the technology. WR is present in sectors as diverse as science, telecom, data centres, electric power distribution and finance, with different degrees of maturity, ranging from laboratory tests to full operational deployment.
WR is also being tested by several research groups to synchronise quantum networks. CERN, the European Laboratory for Particle Physics, recently launched its Quantum Technology Initiative (QTI) with Quantum Communication as one of its Centres of Competence (CoC). Work has started to evaluate ways in which WR can contribute to this CoC, with a first goal of demonstrating the distribution of entangled photon pairs which coexist with WR signals in the same optical fibre. We present our initial analysis, surveying areas of potential application of WR within CERN’s QTI programme.
WR technology is based on fully open-source components: software, firmware, gateware and hardware. WR-compliant hardware (switches and nodes) can be purchased from several independent vendors. This makes WR a natural candidate for the establishment of an international time transfer network. We describe existing links and future plans in Europe to build such a network, both at the level of bilateral agreements and within larger transcontinental initiatives.
With the proliferation of actors and applications in the WR ecosystem, it soon became apparent that a coordinating body was needed. The White Rabbit Collaboration (WRC) was created to fulfil this mission. In its first year of operation, over 15 members joined the WRC, including research institutes, universities and companies, and membership is growing fast. Its main goal is to ensure good support for WR developers and the maintenance of the basic building blocks of WR technology: the WR Switch and WR PTP Core (WRPC) instantiated in each WR Node. In addition, it provides training and a test and qualification framework, and it hosts projects which rely on WR to fulfil their mission. One example is the development of an add-in card for the new WR Switch (version 4) to ensure good holdover in the absence of an operational WR link. Plans for other projects include seamless redundancy and enhancements for long-distance links. We describe the modus operandi of the WRC, in itself an innovative knowledge and technology transfer initiative exploring new forms of public-private partnerships, and we present the current state of projects, along with ideas for the future.
WR started as a technology whose main domain of application was scientific experiments. It is an exercise in maximising performance while remaining fully standard-compatible and using low-cost, scalable technology. In particular, the use of Field-Programmable Gate Arrays (FPGAs) imposes some fundamental limits in low-frequency phase noise or, conversely, long-tau Allan Deviation. These limits are typically not an issue, except for the most stringent synchronisation needs. There are many applications in which the combination of performance, price and vendor-independence makes WR the most suitable solution. This was already true in science a decade ago, and with time many non-scientific applications requiring sub-ns synchronisation have benefited from this technology as well.
To summarise, this paper is an overview of the complete WR ecosystem, including a current status of technical and non-technical work, and plans for further development in the short and medium term.