Long-haul White Rabbit Transmission

Long-haul White Rabbit Transmission Josef Vojtech, Michal Spacek, Jaromir Sima,Tomas Novak, Vladimir Smotlacha, Ondrej Havlis, Elisabeth Andriantsarazo, Rudolf Vohnout, Martin Slapak, Lada Altmannova, Radek Velc, Petr Pospisil, Jan Kundrat dept. Optical networks CESNET Prague, Czechia [email protected] Michal Spacek, Tomas Novak, Elisabeth Andriantsarazo, Jaroslav Roztocil Czech Technical Univerzity Prague, Czechia Abstract— This paper presents a comparison of the benefits of optically amplified time transfer using the White Rabbit (WR) protocol compared to regeneration-based methods utilizing White Rabbit Switches (WRS). We highlight the advantages in terms of simpler way of calibration, lower jitter, and cost efficiency compared to conventional regeneration methods. Keywords—precise time, stable frequency, fiber transfer, White Rabbit I. INTRODUCTION The distribution of precise time and frequency is crucial in many fields and industries. Specifically, in the case of comparing accurate time scales that are implemented, for example, using caesium atomic clocks, ensuring the transfer of precise time is absolutely essential. Currently, this is mainly accomplished using Global Navigation Satellite Systems (GNSS). The common-view method is one of the methods that utilize GNSS and allows for comparing distant time scales when direct local comparison is not feasible. This method achieves uncertainties on the order of a few nanoseconds. The resulting uncertainty is influenced by several factors, such as uncertainties caused by multipath propagation and receiver noise, uncertainties in the calibration of receiver delays, antennas, and antenna cables, uncertainties in antenna positioning, and ionospheric and tropospheric delays. Recently, GNSS techniques have faced increasing threats from spoofing and jamming, revealing vulnerabilities to the injection of false timing information or disruption of signals. These threats can have severe consequences in navigation, compromise the integrity of data in astronomy and physics research, and interfere with time-stamping services essential for financial transactions. In particular, they pose security risks for critical infrastructure, such as power grids and communication networks, which depend on precise timing. Additionally, they can have economic repercussions for industries reliant on GNSS for logistics and transportation. Therefore, there is a need for resilient time dissemination technology that complements satellite-based standards while maintaining equivalent levels of stability and accuracy. Higher precision in the transfer of precise time and frequency can be achieved using optical fibre links. Currently, the White Rabbit (WR) system [1]. II. LONG-HAUL IMPLEMENTATIONS Despite the emerging interest in WR networks, majority of implementations of long-haul/country wide WR networks, prefers utilization of telecom lambdas or Optical Electrical— Optical (OEO) regeneration. In first mentioned case calibration (briefly described in next chapter) is very complicated makes referenceability to an official time scale UTC(k) almost impossible. In second case it seems attractive to utilize simple protocol agnostic and very affordable regenerators. Unfortunately, as shown eg. [2], due to absence of retiming function time and frequency transfer is becoming unstable after few hundreds of second. III. NECESSITY OF CALIBRATION The synchronization between individual WR nodes can be significantly affected by asymmetry in the optical transmission path. Asymmetry refers to the difference in signal propagation time in the forward and reverse directions, both in electronics and the optics. Asymmetry in the optical path despite it bidirectional use, is among other factors, in majority caused by chromatic dispersion. Since the WR system uses bidirectional communication and typically different wavelengths for the forward and backward channels, calibrating the optical path’s asymmetry is crucial for correct Fig. 1. WR time transfer over 400km of fibre - optically amplified setup and accurate synchronization of the nodes. When two synchronized nodes are far apart, the calibration becomes a relatively complicated matter. Calibration consists of local asymmetry calibration and fibre delay asymmetry. Principles of White Rabbit calibration are discussed in [3] and practically evaluated in [4]. Method for the automatic evaluation of asymmetry in the optical transmission path is shown in [5]. IV. LONG HAUL FRIENDLY SOLUTION Chapter II indicated also possibility of regenerated and retimed solution, it has been tested e.g. here [6] and also deployed here [7]. However this solution requires to perform calibration of each segment and its recalibration in case of any service works having significant influence on fibre length or other source of asymmetry. This can be avoided by deployment optically amplified solution. Such solution has been shown by [8] or [9]. To compare two previously mentioned scenarios we set in laboratory setup consisting of 4 times 100 km of G.652D fibre on spools, each spool 50km. Laboratory setup consists of two WR switch devices, configured as grand master and slave, as shown in Fig. 1. The WR system utilizes commercial Small Form Pluggable (SFP) transceivers. Therefore, we have used the pair of 1.25 Gbps SFPs designed for channels 8 and 9 of Dense Wavelength Division Multiplex (DWDM) (wavelengths 1571.24 and 1570.42 nm) featuring extended input sensitivity of -32 dBm and +1 dBm or higher output power. At first, the attenuation of the whole line was compensated by three fully bidirectional EDFA amplifiers CzechLight with single path for signal in both directions placed in between spans. Determined Maximal time interval error (MTIE) is shown in Fig 2 with orange line. Later we replace these amplifiers with WRSes. Measured MTIE is shown in Fig 2. by blue line. To achieve reliable operations, we will consider dual power supply solutions both for WRs and also for bidirectional EDFA amplifiers with remote management. Setting cost of such amplifier to about 0.2 of cost fully featured WRs we are getting cost comparison of regenerated (blue) and amplified solution (orange). Number of inline huts is on X axis. CONCLUSION We continue with verification of propriate setups od long haul WR lines with excessive attenuations (over optical budget of SFP transceivers). On laboratory 400 km fibre setup we measured that both regenerated and amplifies scenarios will keep target sub ns MTIE. However amplified setup overperforms regenerated setup in terms of MTIE and CAPEX and OPEX costs. As our measurement still continuous we are planning to verify behaviour over real fibres buried in ground. ACKNOWLEDGMENT This works was supported by Ministry of education, Youth and Sport of the Czech Republic as a part of the QUEENTEC project, reg. nr. CZ.02.01.01/00/22 008/0004649 Quantum Engineering and Nanotechnology. Authors would like to thank members of WR community a GEANT SIG-TFN community for very valuable, open and friendly discussions. We also would like to thank our colleagues for very useful support in WR transfer deployments within operational network. Fig. 2. MTIE[s] for regenerated solution in Blue and amplified solution in Orange. Fig. 3. Arbitrarry cost comparison regenerated in Blue and amplified solution in Orange. Number of inline huts on X axis. REFERENCES [1] M. Lipinski et al.,“White rabbit: A ptp application for robust subnanosecond synchronization,” IEEE International Symposium on Precision Clock Synchronization for Measurement, Control, and Communication, ISPCS , 25–30 (2011). [2] J. Vojtech et al., "Long Haul Single Fiber Reamplified-Reshaped White Rabbit Transmission," 2021 Joint Conference of the European Frequency and Time Forum and IEEE International Frequency Control Symposium (EFTF/IFCS), Gainesville, FL, USA, 2021, pp. 1-3, doi: 10.1109/EFTF/IFCS52194.2021.9604245. [3] G. Daniluk, “White Rabbit calibration procedure,” 2015. [4] E. Dierikx, E., Y. Xie, “White Rabbit Good Practice Guide,” Dutch Metrology Institute, May 2019. [5] Michal Spacek, et al., “System for continuous evaluation of optical path asymmetry,” Proc. SPIE 12686, Infrared Remote Sensing and Instrumentation XXXI, 126860Q (20 October 2023); https://doi.org/10.1117/12.2678936. [6] Ebenhag, S.-C., Hedekvist, P. O., Jarlemark, P. & Sundblad, R. (2019). Redundant Distributed Timescale Traceable to UTC(SP). In: IFCS/EFTF 2019 - Joint Conference of the IEEE International Frequency Control Symposium and European Frequency and Time Forum, Proceedings. [7] A. Jallageas, “Long-haul dissemination of the Swiss national time scale UTC(CH) in a telecommunication fiber network via White Rabbit” sent for publication. [8] E. F. Dierikx et al., "White Rabbit Precision Time Protocol on LongDistance Fiber Links," in IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control, vol. 63, no. 7, pp. 945-952, July 2016, doi: 10.1109/TUFFC.2016.2518122. [9] J. Vojtěch et al., "Precise time transfer on the IPE - VUGKT line - a detailed analysis," 2016 39th International Conference on Telecommunications and Signal Processing (TSP), Vienna, Austria, 2016, pp. 757-760, doi: 10.1109/TSP.2016.7760986.