Open Access
Add to BibSonomy
Abstract
We experimentally demonstrate a total net-rate of 27.88 Tb/s for C-band wavelength-division multiplexing (WDM) transmission over an ultralong span-length of 150 km. It is the largest net capacity × span-length product of 4182 Tb/s·km for C-band, single-core, standard single-mode optical fiber transmission over a length of more than 3,000 km. A total of 99 channels, spaced at 50 GHz intervals, are employed for transmitting 32 GBaud probabilistically constellation-shaped (PCS) 64QAM signals with an information entropy of 5.5. High gain amplifiers can achieve wavelength-division multiplexing (WDM) transmission with a bandwidth of 6.25 THz, at a noise figure below 4.3 dB, without the assistance of distributed Raman amplification.
© 2024 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
In recent years, global network traffic has increased significantly due to the rapid development of various multimedia and telecommunications service technologies such as big data, cloud computing, artificial intelligence and machine-to-machine communication. Optical fiber communication, serving as the indispensable infrastructure of the global communication network, is facing pressure to expand capacity. At present, transoceanic transmission systems play a crucial role in the worldwide exchange of data, serving as vital connectors between countries and transporting over 95% of international data traffic [1–3]. These systems persistently evolve towards achieving ultra-high speed, ultra-large capacity, and ultra-long distances (transatlantic-range or even transpacific-range) to meet the demands caused by the ever-growing volume of global network traffic [4,5].
Significant advancements have been made in recent years to improve the achievable information rates of single-mode fibers through the expansion of their transmission frequency bands. This has enabled not only the utilization of the C-band, but also the utilization of the L-band and even the S-band [6,7]. In conjunction with technological breakthroughs in integrated optical transceiver [8,9], gain-flattened ultra-wideband low-noise erbium-doped fiber amplifier (EDFA), ultra-low-loss large-effective-area single-mode fiber, advanced constellation shaping, nonlinear compensation (NLC) techniques etc., these advancements continuously renew the transmission capacity records of optical fiber communication systems. Meanwhile, system cost is also an important factor limiting the practical application of optical fiber communication systems. Therefore, the trade-off between capacity and cost has always been a compelling research focus. The comprehensive expense of these systems encompasses cable, fiber, repeater, transponder, and marine deployment costs. With cable and fiber costs being directly tied to transmission length, diminishing the count of repeaters and transponders (linked with the fiber span length) can markedly decrease the system’s overall cost. Consequently, the required number of EDFAs for a specific link length rises as the fiber span length diminishes, leading to elevated system expenses. Conversely, system costs can be significantly reduced by a longer span length. However, the majority of demonstrations for long-distance transmission systems exceeding three thousand kilometers have typically used short fiber spans ranging from 50 to 80 km to alleviate the decrease in optical signal-to-noise ratio (OSNR) for improving the transmission distance, and even Raman amplification with low noise figure is utilized to further boost the transmission distance. Reducing transmission system costs by increasing span length has received significant attention [10–13].
Figure 1 provides an overview of various C-band transmission experiments with different span lengths for long-distance (≥3000 km) and high-capacity (≥15Tbit/s) scenarios [14–23]. Note that these C-band transmission experiments are achieved using single-core, single-mode optical fiber and the capacity is calculated by net channel rate × channel number. Our work achieves the largest net capacity × span-length product of 4182 Tb/s·km for C-band, single-core, standard single-mode optical fiber transmission over a length of more than 3,000 km as shown in Fig. 1. The key to achieving long-span lengths is to alleviate the decrease in OSNR. The received OSNR after multiple EDFA can be estimated based on the launching power (P), the loss of fiber span (Loss), and the noise figure (NF) of the EDFA:
where N is considered as the number of EDFAs. Therefore, when the transmission distance is fixed, there are some strategies to slow down the rapid decrease in the system’s OSNR. One key point is to reduce the number N of EDFAs by applying longer span length in the transmission system, while longer span length may be an effective method to reduce system costs, but can result in higher fiber loss (so we have adopted ultra-low loss fiber in the transmission system to mitigate this impact) which may cause the degradation of OSNR. Specifically, it is possible to achieve higher system OSNR in the fixed optical fiber transmission system by reducing the noise figure of EDFA or allowing higher channel launch powers, or a combination of both. Therefore, it is desirable to minimize fiber loss, the number and the noise figure of EDFAs, while maximizing the fiber effective area to reduce the fiber nonlinearity effect [24–26].
Fig. 1. The C-band transmission experiments with different span lengths for long-distance(≥3000 km) and high-capacity (≥15Tbit/s).
In this work, we experimentally validate a net capacity of 27.88 Tbit/s (a line capacity of 34.85 Tbit/s) for a 4550 km single-mode optical fiber, utilizing predominantly 150 km span lengths, C-band low-noise coefficient and high-gain EDFA, PDM-PS-64QAM, and ultra-low loss large effective area single-mode fibers. The 150 km fiber span length is the longest reported span length to date for C-band, single-core, standard single-mode optical fiber transmission over a length of more than 3,000 km and ultra-large capacity (≥15Tb/s). One important reason for applying the long span length in this system is that our team designs and optimizes the parameters of C + band EDFA (Accelink) such as gain, flatness, and noise index to match fiber span loss and wideband amplification.
2. Experimental setup
Figure 2 illustrates the experimental setup of the C-band EDFA-only long-haul and ultra-large capacity transmission system. The transmitter terminal comprises two components, including a C-band ASE noise source, and a test channel. For channel emulation, the non-measurement band signals consist of the channelized amplified spontaneous emission (ASE) noise devoid of modulation. In the test channel, the digital signals are shaped through a square root raised cosine (SRRC) pulse with a roll-off factor of β = 0.9, followed by conversion to analog signals using a DAC device operating at 256 GSa/s with a 70 GHz analog bandwidth. The test channel light source from a tunable external-cavity laser with a linewidth < 10 kHz is modulated by a IQ modulator to generate a 32GBaud PCS-64QAM optical signal with an entropy of 5.5. Subsequently, polarization multiplexing operation is applied to achieve the line rate of 352Gbit/s PDM-PCS-64QAM. The transmission link uses a loop test bed which comprises three spans. The first two spans consist of 150 km ultra-low-loss large-effective-area single-mode fibers and an ultra-wideband EDFA with wavelength range from 1524 nm to 1572.3 nm and variable amplification gain from 22 dB to 32 dB. Subsequently, the third part consists of a 50 km ultra-low-loss large-effective-area single-mode fiber, a wavelength selective switch (WSS) and an acousto-optic modulator (AOM). We chose 50 km ultra-low-loss large-effective-area single-mode fiber in the third span mainly to balance the power of the loop test bed. In this experimental system, the loop test bed is controlled by the on/off time of two AOMs. The WSS (II-VI 1000B) is applied to balance gain spectra and filter out amplified spontaneous emission (ASE) noise in the loop test bed. The total output power of C-band EDFA (the launching power of the transmission link) is ∼24dBm. Table 1 presents the key parameters (e.g. modulation format, symbol rate, the parameters of optical fiber and EDFA) of the long-haul transmission system. After transmission over 4,550 km, the test channel is filtered out at the receiving end through a WSS, detected by a coherent receiver with an electro-optic bandwidth of ∼40 GHz and sampled by a real-time digital oscilloscope with 59 GHz electrical bandwidth and 256GSa/s sampling rate. Finally, the received digital signal is processed by offline digital signal processing (DSP) module including chromatic dispersion and nonlinear compensation, polarization mode dispersion compensation, carrier frequency offset estimation (FOE) and carrier phase estimation (CPE). Here, digital backpropagation (DBP) is executed for compensating chromatic dispersion and nonlinear effects in the transmission link. Subsequently, multi-modulus algorithm (MMA) is employed for polarization de-multiplexing and equalization. Then, a 4th power fast Fourier transform (FFT) operation is employed to estimate the frequency offset between the local oscillator and the test channel, and the blind phase search (BPS) algorithm is utilized to compensate carrier phase noise. Finally, standard decision-directed least-mean-square (DD-LMS) algorithm is introduced for further achieving better performance.
Fig. 2. The experimental setup. TLS: tunable laser source; ASE: amplifier spontaneous emission; DAC: digital-to-analog converter; WSS: wavelength selective switch; EDFA: Erbium-doped fiber amplifier; AWG: arbitrary waveform generator; WDM: wavelength division multiplexing; AOM: acousto-optical modulators; PC: polarization controller; VOA: variable optical attenuator; DSP: digital signal processing; LO: local oscillation; PBS: polarization beam splitter; PBC: polarization beam combiner; DSO: digital storage oscilloscope.
Table 1. System primary parameters.
In this work, C-band (1524 nm to 1572.3 nm) ultra-wideband (∼48 nm) EDFA designed by our team plays a crucial role in long-distance/high-capacity transmission and achieving ultra-long span-length imposes higher requirements on the gain and noise figure (NF) of EDFA. Simultaneously, to mitigate the degradation of output spectrum flatness caused by amplifier cascading, a single amplifier is required to have ultra-low gain fluctuations (gain flatness). Figure 3 illustrates the measurement results of gain and noise figure for the C-band ultra-wideband EDFA with different gains (e.g. 32 dBm, 27 dBm, and 22 dBm). It can be observed that the C-band ultra-wideband EDFA has the lowest noise figure (<4.3dBm) under 32dBm gain. Therefore, the C-band ultra-wideband EDFA is very suitable for ultra-long span-length (low-cost) optical fiber transmission systems.
3. Experiment results
The C-band optical spectrums before and after 4,550 km transmission are depicted in Fig. 4(a) and (b), illustrating the OSNR performance. Before transmission, the C-band ASE noise source is divided into 99 spurious optical channels with 0.4 nm interval by a WSS with a resolution of 0.025 nm, which also can be applied as a band-stop filter to create a spectral gap at the location where the test channel can be inserted. [27] provides a detailed investigation into the application of spectrally shaped amplified spontaneous emission noise (SS-ASE) loading in the design and evaluation of broadband (up to the entire C-band) optical fiber transmission systems. The C-band optical spectrum of the transmitter is shown in Fig. 4(a). After 4,550 km (350 km × 13 loops) transmission, the optical spectrum is depicted in Fig. 4(b). The observation indicates that the OSNR of the short-wavelength segment is significantly lower than that of the long-wavelength segment, which can be attributed to the higher gain in the lower wavelength band. Consequently, the WSS needs to make more cuts to the spectrum of the lower wavelength band during each loop for spectral shaping. This undoubtedly leads to a weakening of the signal-to-noise ratio of the lower wavelength band. Another contributing factor is the influence of the Raman scattering effect.
The Q performance with different OSNR values under back-to-back condition is illustrated in Fig. 5(a). The Q factors are calculated under a 352 Gb/s PCS-64QAM (H = 5.5) signal at 1549.325 nm. To attain the 25% soft decision forward error correction (SD-FEC) threshold of Q = 4.69, an OSNR of 24 dB is necessary, representing a margin of approximately 0.8 dB from the theoretical limit. Next, we further demonstrate the transmission performance at different fiber transmission lengths, as depicted in Fig. 5(b). We observe that the signal, under a transmission distance of 4900 km, precisely meets the transmission conditions for a 25% FEC, yet the spectrum displayed in Fig. 4(b) reveals poorer OSNR performance in the lower frequency band, constituting a primary constraint leading to a final transmission distance limitation of 4550 km.
Fig. 5. (a) Theoretical and measured back-to-back performance; (b) Q-factor vs. transmitted distance at 1549.325nm
Figure 6 presents the Q factors between 4.86 and 5.61 for all 99 channels within the C band (ranging from 1528.425 nm to 1567.625 nm), which below the 25% threshold of SD-FEC at a Q of 4.69 after 4,550 km transmission. The net data rate of 32GBaud PDM-PCS-64QAM (H = 5.5) signals with a 25% overhead of SD-FEC is 281.6 Gbit/s (the line rate is 352 Gbit/s). The net capacity amounts to 99 × 281.6Gbit/s = 27.88Tbit/s (the line capacity is 34.58 Tbit/s), with a corresponding capacity-distance product of approximately 126.85Pbit/s × km. In the experimental system, characterized by white Gaussian noise distribution, SD-FEC decoding is proficient in supporting signal demodulation. It is well known that chromatic dispersion (CD) varies with wavelength. In our wavelength division multiplexing systems with a wavelength span exceeding 37 nm, we use our previously proposed CD estimation algorithm to accurately estimate the accumulated CD value of each channel. The CD coefficient of ultra-low loss large-effective-area fiber is 20.88ps/nm/km at 1550 nm. The CD values of all C-band 99 channels after 4550 km transmission range from 0.01975 ps/nm.km to 0.02217 ps/nm.km as shown in Fig. 6.
4. Conclusions
In our experimental investigation, we experimentally demonstrate a total net-rate of 27.88 Tb/s over 4550 km for C-band wavelength-division multiplexing (WDM) transmission over an ultralong span-length of 150 km. It is the largest net capacity × span-length product of 4182 Tb/s·km for C-band, single-core, standard single-mode optical fiber transmission over a length of more than 3,000 km. The transmission link uses a loop test bed which comprises three spans, consisting of two 150 km spans and 50 km span. Note that the 50 km span is to balance the gain and loss of the optical fiber link. As has been reported, more than 85% (3900 km) of the 4550 km transmission links in our system use 150 km spans, accounting for the leading role in the transmission of OSNR. The utilized fiber was predominantly a prototype featuring ultra-low loss and a very large effective area, specifically designed for submarine systems. And the employed ultra-wideband EDFA is characterized by high gain and low noise figure. In future, the combination of the C-band EDFA and L-band ultra-wideband EDFA is expected to achieve 100 nm ultra-wideband amplification. At the same time, the use of high-gain fiber amplifiers is expected to reduce the cost of medium and long-distance fiber optic transmission systems.
Funding
National Key Research and Development Program of China (2021YFB2206303); National Natural Science Foundation of China (62005228); Key Technology Research and Development Program of Shandong (2023CXPT100); National Student Research Training Program of China (20230613037).
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. W. He and S. Yu, “Latest survey on optical fiber communication,” Sci. Sin.-Inf. 50(9), 1361 (2020). [CrossRef]
2. W. Klaus, P.J. Winzer, and K. Nakajima, “The role of parallelism in the evolution of optical fiber communication systems,” Proc. IEEE 110(11), 1619–1654 (2022). [CrossRef]
3. P. J. Winzer, David T. Neilson, Andrew R. Chraplyvy, et al., “Fiber-optic transmission and networking: the previous 20 and the next 20 years,” Opt. Express 26(18), 24190–24239 (2018). [CrossRef]
4. S. Yu, M. Luo, X. Li, et al., “Recent progress in an ‘ultra-high speed, ultra-large capacity, ultra-long distance’ optical transmission system,” Chin. Opt. Lett. 14(12), 120003 (2016). [CrossRef]
5. L. Yan, Alan E. Willner, Xiaoxia Wu, et al., “All-optical signal processing for ultra-high speed optical systems and networks,” J. Lightwave Technol. 30(24), 3760–3770 (2012). [CrossRef]
6. F. Pittalà, G. Böcherer, P. Schulte, et al., “72.64 Tb/s DWDM transmission over 100 km G.654D fiber using super C-band erbium-doped fiber ampliffcation,” Optical Fiber Communication Conference (OFC), San Diego, CA, USA, 2022, paper W3C.4.
7. F. Hamaoka, K. Minoguchi, T. Sasai, et al., “150.3-Tb/s Ultra-wideband (S, C, and L Bands) single-mode fibre transmission over 40-km using >519Gb/s/A PDM-128QAM Signals,” European Conference on Optical Communication (ECOC), Rome, Italy, 2018, pp. 1–3.
8. L Di and B John E, “Recent progress in heterogeneous III-V-on-silicon photonic integration,” Light: Advanced Manufacturing 2(1), 59 (2021). [CrossRef]
9. H. Liu, B. Pan, Y. Huang, et al., “Ultra-compact lithium niobate photonic chip for high-capacity and energy-efficient wavelength-division-multiplexing transmitters,” Light: Advanced Manufacturing 4(2), 1 (2023). [CrossRef]
10. M. Salsi, H. Mardoyan, P. Tran, et al., “155 × 100Gbit/s coherent PDM-QPSK transmission over 7,200 km,” in Proceedings of European Conf. Opt. Commun. (2009), paper PD2.5
11. John D. Downie, Jason Hurley, John Cartledge, et al., “112 Gb/s PM-QPSK transmission up to 6000 km with 200 km amplifier spacing and a hybrid fiber span configuration,” European Conference on Optical Communication (ECOC), Geneva, Switzerland, 2011. [CrossRef]
12. John D. Downie, Jason Hurley, Dragan Pikula, et al., “Ultra-long-haul 112 Gb/s PM-QPSK transmission systems using longer spans and Raman amplification,” Opt. Express 20(9), 10353–10358 (2012). [CrossRef]
13. G. Charlet, M. Salsi, P. Tran, et al., “72 × 100Gb/s transmission over transoceanic distance, using large effective area fiber, hybrid Raman-erbium amplification and coherent detection,” Optical Fiber Communication Conference and Exposition (OFC) (Optical Society of America, Washington, DC, 2009), paper PDPB6.
14. A. Ghazisaeidi, Ivan Fernandez de Jauregui Ruiz, Rafael Rios-Muller, et al., “65Tb/s transoceanic transmission using probabilistically-shaped PDM-64QAM,” European Conference on Optical Communication (ECOC),2016, pp. Th.3.C.4.
15. A. Ghazisaeidi, Laurent Schmalen, Patrice Tran, et al., “54.2 Tb/s transoceanic transmission using ultra low loss fiber multi-rate FEC and digital nonlinear mitigation,” in ECOC’2015, pp. Th.2.2.3.
16. A. Ghazisaeidi, Laurent Schmalen, Ivan Fernandez de Jauregui, et al., “52.9 Tb/s transmission over transoceanic distances using adaptive multi-rate FEC,” in ECOC’2014, pp. PD.3.4.
17. J.-X. Cai, Hussam G. Batshon, Matthew V. Mazurczyk, et al., “51.5 Tb/s capacity over 17,107 km in C + L bandwidth using single mode fibers and nonlinearity compensation,” in ECOC’2017, pp. Th.PDP.A.2
18. S. Zhang, Fatih Yaman, Yue-Kai Huang, et al., “50.962Tb/s over 11185 km Bi-directional C + L transmission using optimized 32QAM,” in Conference on Lasers and Electro-Optics, OSA Technical Digest (online) (Optica Publishing Group, 2017), paper JTh5A.9. [CrossRef]
19. M. Ionescu, Polina Bayvel, Domanic Lavery, et al., “74.38 Tb/s transmission over 6300 km single mode fibre enabled by C + L amplification and geometrically shaped PDM-64QAM,” J. Lightwave Technol. 38(2), 531–537 (2020). [CrossRef]
20. L. Jiang, “54.07 Tb/s Transmission over 12,057 km single mode fiber using C + L Band EDFA amplification with 75.36 km span length,” in ACP’2022, pp. 33–37.
21. M. Salsi, “38.75 Tb/s transmission experiment over transoceanic distance”, Proc.ECOCPD3E.2 (2013). [CrossRef]
22. B. Zhu, J. Zhang, J. Yu, et al., “Transmission of 200Gb/s PM-16QAM and 150Gb/s PM-8QAM DWDM Signals over Long-haul and Transoceanic Distance at 100 km Span Length with EDFA-only,” 2017 European Conference on Optical Communication (ECOC), Gothenburg, Sweden, 2017, pp. 1–3.
23. Dayou Qian, Ming-Fang Huang, Shaoliang Zhang, et al., “30Tb/s C- and L-bands bidirectional transmission over 10,181 km with 121 km span length,” Opt. Express 21(12), 14244–14250 (2013). [CrossRef]
24. Y. Yamamoto, M. Hirano, K. Kuwahara, et al., “OSNR-enhancing pure-silica-core fiber with large effective area and low attenuation,” in Optical Fiber Communication Conference and Exposition (OFC) (Optical Society of America, Washington, DC, 2010), paper OTuI2. [CrossRef]
25. R. Chen, M. O’Sullivan, C. Ward, et al., “Next generation transmission fiber for coherent systems,” in Optical Fiber Communication Conference and Exposition (OFC) (Optical Society of America, Washington, DC, 2010), paper OTuI1. [CrossRef]
26. J. D. Downie, J. E. Hurley, J. Cartledge, et al., “Transmission of 112 Gb/s PM-QPSK signals over 7200 km of optical fiber with very large effective area and ultra-low loss in 100 km spans with EDFAs only,” in Optical Fiber Communication Conference and Exposition (OFC) (Optical Society of America, Washington, DC, 2011), paper OMI6 [CrossRef]
27. Daniel J. Elson, Gabriel Saavedra, Kai Shi, et al., “Investigation of bandwidth loading in optical fibre transmission using amplified spontaneous emission noise,” Opt. Express 25(16), 19529–19537 (2017). [CrossRef]
