Open Access
Add to BibSonomyAbstract
We generate a 105.7-Gb/s signal by directly modulating a 1.5-µm VCSEL with a 33.35-Gbaud 3-level signal and polarization multiplexing. By using digital coherent detection, we successfully transmit the 105.7-Gb/s line rate (88.10 Gb/s net bit rate) signal over 960-km standard single-mode-fiber (SSMF) at a 20% hard-decision forward-error correction (FEC) threshold, which is at bit-error ratio (BER) of 1.5x10−2.
©2013 Optical Society of America
1. Introduction
The rapid growth of Internet and cloud computing applications drives a huge demand for the capacity of communication networks. With the commercialization and deployment of 100-Gb/s technologies using polarization-division-multiplexed quadrature-phase-shift-keying (PDM-QPSK) and digital coherent detection in optical transport networks and the development of higher bit rates such as 400-Gb/s and 1-Tb/s technologies, there is also an urgent need to upgrade metro networks from 10 Gb/s to 100 Gb/s and beyond in the near future [1]. Digital coherent detection is one way to achieve high spectral efficiencies and networking flexibilities. Compared with optical transport networks, metro networks are much more sensitive to cost, footprint, and power consumption. Significant developments have to be done to achieve small-form-factor, low-power consumption and low-cost coherent transceivers [2,3].
Vertical-cavity surface-emitting lasers (VCSELs) dominate short-reach and low-data-rate applications due to their cost, energy efficiency, and footprint [4–11]. Most VCSELs work at 850 nm and the highest bit rate achieved has reached 55 Gb/s, targeting at data center applications with link lengths of less than 100 meters [4–6]. Significant progress has also been made on single-mode 1.5-µm VCSELs [7–11]. Recent new developments of single-mode 1.5-µm VCSELs have enabled 40-Gb/s operation and the maximum transmission distance of 100 km at 10 Gb/s over standard single-mode fibers (SSMFs) [7–11]. A 100-Gb/s short reach link using VCSELs with direct modulation has also been demonstrated recently [10], with 4-level pulse-amplitude modulation (PAM), polarization-division multiplexing and direct detection, but only 100-m transmission was achieved. Directly modulated VCSELs are generally not believed to be suited for metro networks with transmission distances between 100 and 1000 km at high data rates.
In this paper, we demonstrate the transmission of a 100-Gb/s signal over 960-km SSMF using 1.5-µm directly modulated VCSELs, aiming at the application in next-generation 100-Gb/s metro networks. The achieved transmission distance at such high data rate is enabled by digital coherent detection, 3-PAM modulation, and the use of direct-modulation induced chirp. To the best of our knowledge, this is the longest transmission distance achieved by directly-modulated VCSELs. The signal was generated by driving a single VCSEL with a 33.35-Gbaud 3-level signal (i.e., achieving 52.86 Gb/s), and subsequently polarization-division multiplexing to emulate two VCSELs transmitting 105.70 Gb/s at the same wavelength channel. At the receiver, coherent detection is used to improve system performance, including chromatic dispersion (CD) compensation, polarization-mode dispersion (PMD) compensation, polarization demultiplexing, and distortion compensation in the electronic domain with digital signal processing (DSP). In this paper, the coherent detection of the 3-PAM signal is achieved with offline processing of the data captured by real-time digital sampling oscilloscopes. This can be easily realized in real systems by modifying and current mature coherent PDM-QPSK receivers. Compared with an inphase/quadrature modulator based coherent transmitter, the VCSEL-based transmitter has a smaller form factor, lower power consumption and lower cost, as it requires lower driving voltages and does not need external modulators and related optical components. Meanwhile, as we are using pure intensity modulation, the large linewidth of the VCSEL (> 500 MHz) has little effect on system performance, and no carrier frequency and phase recoveries are needed in the DSP, which further reduces complexity and power consumption of the coherent receiver.
2. Experimental setup
The experimental setup is shown in Fig. 1. The VCSEL is a high-speed short-cavity VCSEL with buried tunnel junction (BTJ) aperture diameter of 4.5 µm. It operates in a single mode and emits linearly polarized light along a well-defined polarization axis. Its emission wavelength and 3-dB modulation bandwidth are 1.5 μm and 18 GHz, respectively. Detailed descriptions of the specific VCSEL characteristics can be found in [11]. Considering the bandwidth of the VCSEL and performance of multi-level PAM, we chose 3-PAM in our experiment, which carries 1.585 (log23) bits per symbol per polarization, corresponding to 3.17 bits per symbol when using polarization-division multiplexing. At 33.35-Gbaud, the raw line rate is 105.7195 Gb/s.
Fig. 1 Experimental setup. PolMux: polarization multiplexer, OF: optical filter, LO: local oscillator, DGEF: dynamic gain equalizer filter. The insets are the eye-diagrams of the electrical driving signal, output optical signal from the VCSEL, and the recovered constellations after offline processing.
Two bits of a 3-bit high-speed digital-to-analog converter (DAC) was used to generate a 3-level 33.35-Gbaud signal, which were fed by 48-symbol delay-decorrelated versions of and of a pattern generator with a 33.35-Gb/s 215-1 pseudo-random bit-sequence (PRBS). The driving signal from the DAC had a peak-to-peak amplitude of about 800 mV. No additional driver/amplifier was used. The bias of the VCSEL was set to 8 mA and the temperature was set at 25°C. The emitting wavelength after modulation was ~1528 nm and the output power was 0 dBm. The eye-diagrams of the 3-level electrical and optical signals are shown in the insets of Fig. 1. The output from the VCSEL was sent to a polarization multiplexer. The length difference between x and y polarization branches was about 15 meters, which introduced a time delay that is longer than the coherence length of the VCSEL, thus emulating two such high-speed modulated VCSELs at the same wavelength channel, as needed in such a transponder. The signal was amplified by an erbium-doped-fiber amplifier (EDFA) and then sent to a JDSU TB9 optical grating filter with a 3-dB bandwidth of 0.52 nm. The function of this filter will be explained below.
Transmission experiments were performed in a 4x80-km EDFA amplified SSMF recirculating loop without any dispersion compensation. A dynamic gain equalizing filter (DGEF) after each loop was used to block amplified spontaneous emission (ASE) noise and the losses of switches and DGEF were compensated by an EDFA. At the receiver, the signal was mixed with a free-running tunable external-cavity laser (ECL) local oscillator (LO) in a polarization diversity 90° hybrid, followed by four balanced detectors with bandwidths of 40 GHz. The performance does not change when we move the LO frequency a few GHz away from the transmitter VCSEL. The four signal components were captured by two 2-channel 80-GSamples/s real-time digital sampling oscilloscopes with 30-GHz bandwidths. The captured signal was digitally processed offline. For offline DSP, the sampling skew was first corrected and the signal was synchronously re-sampled to 2 samples per symbol. After CD compensation, a butterfly equalizer with 9 taps, adapted via a least-mean-square (LMS) algorithm, was used for polarization demultiplexing and inter-symbol interference compensation. Symbol identification was performed right after the equalizer and no carrier frequency and phase estimation was used. Bit-error ratios (BERs) were calculated using direct error-counting. As shown in the inset of Fig. 1, the recovered signal constellations have three rings.
The optical filter following the directly modulated VCSEL significantly improved system performance, as illustrated by the captured signal clouds taken by a digital sampling oscilloscope shown in Fig. 2 in back-to-back operation. The optical filter (which may equally well be implemented by DSP at transmitter or receiver) suppressed the lower-level amplitude and increased the amplitude difference between different levels. This can be explained by Fig. 3, which shows the spectra of the signal before and after the filter together with the filter transfer function. In a directly modulated laser, higher-intensity symbols are blue shifted relative to lower-intensity symbols. When we align filter and signal wavelengths in the way shown in Fig. 3, the red shifted part (lower-intensity symbols) of the signal experiences a higher attenuation than the blue shifted part (higher-intensity symbols). This frequency modulation (FM) to amplitude modulation (AM) conversion increases the eye-opening of the signal and thus the performance of the system.
3. Experimental results
We first measured back-to-back performance, and the results are given in Fig. 4, which shows the BER versus optical signal-to-noise ratio (OSNR) in back-to-back operation. It shows that there is an error floor at a BER of about 2.0x10−3. With 7% overhead hard-decision forward-error-correction (FEC) code (resulting in a net bit rate of 98.80 Gb/s), we can achieve error-free operation with an OSNR larger than 26 dB; if a 20% overhead hard-decision FEC code (net bit rate of 88.10 Gb/s) is used [12], error-free operation can be achieved with an OSNR larger than 20.3 dB.
Fig. 4 BER versus OSNR in back-to-back operation. The BERs at 7% and 20% hard-decision FEC are 3.8x10−3 and 1.5x10−2, respectively.
We then measured the transmission performance of the signal. We varied the launch power to each span and measured BER at three transmission distances, 320 km, 640 km, and 960 km. The results are plotted in Fig. 5. The optimum launch powers are about 2~3 dBm, similar for all the three transmission distances. At 3-dBm launch power, the delivered OSNRs are about 29.5 dB, 26.5 dB and 24.0 dB for 320 km, 640 km and 960 km, respectively. With 7% overhead hard-decision FEC, we can achieve 320-km transmission distance, and with 20% hard-decision FEC, 960-km distance can be reached.
4. Conclusion
Using 3-PAM modulation, polarization-division multiplexing, and digital coherent detection, we successfully transmitted a 105.7-Gb/s (raw line rate) signal generated by direct VCSEL modulation over 320-km SSMF at 7% hard-decision FEC threshold (98.80 Gb/s net bit rate) and 960-km SSMF at 20% hard-decision FEC threshold (88.10 Gb/s net bit rate), respectively. Compared with a coherent transmitter based on inphase/quadrature modulators, the VCSEL-based transmitter promises a smaller form factor, lower power consumption and lower cost. Meanwhile, the DSP power may also be reduced at the coherent receiver side through elimination of frequency and phase recoveries. By combining the advantages of VCSELs from short-reach communications and powerful coherent detection from long-haul transmission, a VCSEL transmitter in combination with a coherent receiver may well be suited for 100G metro networks.
References and links
1. P. J. Winzer, “High-spectral-efficiency optical modulation formats,” J. Lightwave Technol. 30(24), 3824–3835 (2012). [CrossRef]
2. P. Evans, M. Fisher, R. Malendevich, A. James, P. Studenkov, G. Goldfarb, T. Vallaitis, M. Kato, P. Samra, S. Corzine, E. Strzelecka, R. Salvatore, F. Sedgwick, M. Kuntz, V. Lal, D. Lambert, A. Dentai, D. Pavinski, B. Behnia, J. Bostak, V. Dominic, A. Nilsson, B. Taylor, J. Rahn, S. Sanders, H. Sun, K.-T. Wu, J. Pleumeekers, R. Muthiah, M. Missey, R. Schneider, J. Stewart, M. Reffle, T. Butrie, R. Nagarajan, C. Joyner, M. Ziari, F. Kish, and D. Welch, “Multi-channel coherent PM-QPSK InP transmitter photonic integrated circuit (PIC) operating At 112 Gb/s per wavelength,” in Optical Fiber Communication Conference and Exposition and The National Fiber Optic Engineers Conference, OSA Technical Digest Series (Optical Society of America, 2011), paper PDPC7 (2011).
3. P. Dong, C. Xie, L. Chen, L. L. Buhl, and Y.-K. Chen, “112-Gb/s monolithic PDM-QPSK modulator in silicon,” Opt. Express 20(26), B624–B629 (2012). [CrossRef] [PubMed]
4. M. C. Amann, E. Wong, and M. Müller, “Energy-efficient high-speed short-cavity VCSELs,” in Optical Fiber Communication Conference and Exposition and The National Fiber Optic Engineers Conference, OSA Technical Digest Series (Optical Society of America, 2012), paper OTh4F.1 (2012).
5. D. M. Kuchta, A. V. Rylyakov, C. L. Schow, J. E. Proesel, C. Baks, C. Kocot, L. Graham, R. Johnson, G. Landry, E. Shaw, A. MacInnes, and J. Tatum, “ A 55Gb/s directly modulated 850nm VCSEL-based optical link” in Proc. IEEE Photonics Conference (San Francisco, CA, USA, 2012), paper PD 1.5 (2012). [CrossRef]
6. J. S. Gustavsson, A. Larsson, Å. Haglund, J. Bengtsson, P. Westbergh, R. Safaisini, and E. Haglund, “High speed 850nm VCSELs for > 40Gb/s transmission,” in Optical Fiber Communication Conference and Exposition and The National Fiber Optic Engineers Conference, OSA Technical Digest Series (Optical Society of America, 2013), paper OTh4H.4 (2013). [CrossRef]
7. W. Hofmann, M. Müller, P. Wolf, A. Mutig, T. Gründl, G. Böhm, D. Bimberg, and M.-C. Amann, “40 Gbit/s modulation of 1550 nm VCSEL,” Electron. Lett. 47(4), 270–271 (2011). [CrossRef]
8. R. Rodes, J. B. Jensen, D. Zibar, C. Neumeyr, E. Roenneberg, J. Rosskopf, M. Ortsiefer, and I. T. Monroy, “All-VCSEL based digital coherent detection link for multi Gbit/s WDM passive optical networks,” Opt. Express 18(24), 24969–24974 (2010). [CrossRef] [PubMed]
9. Y. Rao, C. Chase, M. C. Y. Huang, S. Khaleghi, M. R. Chitgarha, M. Ziyadi, D. P. Worland, A. E. Willner, and C. J. Chang-Hasnain, “Continuous tunable 1550-nm high contrast grating VCSEL,” in CLEO Technical Digest 2012, paper CTh5C.3 (2012).
10. R. Rodes, J. Estaran, B. Li, M. Müller, J. B. Jensen, T. Gründl, M. Ortsiefer, C. Neumeyr, J. Rosskopf, K. J. Larsen, M.-C. Amann, and I. T. Monroy, “100 Gb/s single VCSEL data transmission link,” Optical Fiber Communication Conference and Exposition and The National Fiber Optic Engineers Conference, OSA Technical Digest Series (Optical Society of America, 2012), paper PDP5D.10 (2012). [CrossRef]
11. M. Müller, W. Hofmann, T. Gründl, M. Horn, P. Wolf, R. D. Nagel, E. Rönneberg, G. Böhm, D. Bimberg, and M.-C. Amann, “1550-nm high-speed short-cavity VCSELs,” IEEE J. Sel. Top. Quantum Electron. 17(5), 1158–1166 (2011). [CrossRef]
12. F. Yu, C. Xie, and L. Zeng, “Application aspects of enhanced FEC for 40/100G systems,” in Proc. European Conference on Optical Communication (Torino, Italy, 2010), workshop 11, (2011).
