An experiment of an all-optical DPSK-signal regeneration is reported. In the regenerator, incoming DPSK signals are first demodulated to on-off-keying signals, then amplitude-regenerated by a 2R regenerator, and subsequently used as control pulses for phase remodulation of clock pulses in an all-optical phase modulator. Penalty-free operation with reduced amplitude noise is demonstrated by the use of a two-stage fiber-based cascaded 2R amplitude regenerator in bidirectional configuration.
©2008 Optical Society of America
Signal regeneration can reset accumulation of signal distortion caused by various origins in long-distance transmission such as dispersion and noise, by which the reach of high-speed systems and the scale of photonic networks are significantly increased. All-optical regeneration without O/E/O conversion is particularly attractive because of its potential in high-speed and low-power-consumption operation . In future systems, various phase-shift-keying (PSK) signals in addition to conventional on-off-keying (OOK) signals will be used because of their superior transmission performance including higher sensitivity and higher spectral efficiency realized by multi-level formats . All-optical regeneration techniques for PSK signals will then be desired.
Recently a number of studies have discussed all-optical regenerators or regenerative wavelength converters for differential PSK (DPSK) signals [3–17]. Reduction of phase noise is obtained by the use of phase-sensitive amplifiers [3,4] or by converting the phase information to/from the amplitude information and performing the regeneration operation on the amplitude [5–8]. Averaging of phase fluctuations over neighboring bits can also lead to phase-noise reduction [9–12]. Amplitude-only regeneration with phase noise maintained, which is effective in reducing the nonlinear phase noise (the Gordon-Mollenauer effect), has also been reported [13–17]. Regeneration of (differential) quadrature PSK ((D)QPSK) signals is discussed in [18,19].
In this paper we report an experiment of an all-optical regenerator for DPSK signals where the noise reduction is performed in the amplitude domain by the use of a fiber-based 2R (reamplification and reshaping) amplitude regenerator . Penalty-free regeneration with suppressed amplitude noise is demonstrated.
2. DPSK signal regenerator
Figure 1 shows a block diagram of the regenerator . Incoming DPSK signals are demodulated to OOK signals by the use of a 1-bit delay interferometer (DI), by which the phase variation including noise is translated to amplitude variation. A 2R amplitude regenerator subsequently suppresses the amplitude noise both in space and mark bit slots of the OOK signal. Phase preservation is not needed in this regeneration process. The amplitude-stabilized data pulses are fed to an all-optical phase modulator  where the data pulses modulate phase of clean clock pulses generated by a local pulse source whose repetition frequency is recovered from the incoming signal. The insertion of the 2R regenerator is important because the phase noise is rather enhanced if the demodulated OOK signal is directly used as the control pulses in the phase modulator. When the input noise is circular Gaussian and the phase modulation is proportional to the energy of each control pulse, standard deviation of the signal phase is enhanced by a factor of 21/2π . Reduction of energy fluctuation of the demodulated pulses by more than this factor is thus needed. It is noted that the data encoded on the phase of the outgoing pulses are the exclusive OR of the data of adjacent bits in the incoming pulses. That is, PSK-modulated pulses having a phase pattern …, 0, 0, π, 0, π, π, π, 0,...., for example, are translated by the DI to OOK-modulated pulses having a bit pattern …, 0, 1, 1, 1, 0, 0, 1, …, which in turn yield output phase-modulated pulses having a phase pattern …, 0, π, π, π, 0, 0, π, … after the all-optical phase modulator. If the signal is repeatedly regenerated, the phase pattern will be further converted to different patterns. This conversion, however, can be reversed and the original data can be recovered by a suitable signal processing after detection at the receiver [5,6]. Precording at the transmitter is also possible to counteract the pattern alteration if the number of regenerators traversed by the signal is known beforehand at the transmitter.
3. Experimental setup
Figure 2 shows the experimental setup for the DPSK signal regeneration. Figure 2(a) is the signal source in which 10GHz short pulses at 1553nm with duration ~2.2ps are generated by a mode-locked semiconductor laser diode (MLLD). In the following highly nonlinear fiber (HNLF), the signal spectrum is broadened and sliced by an optical bandpass filter (OBPF) with 1nm bandwidth, by which wavelength-shifted pulse train at 1548.5nm is generated. The pulses are then phase-modulated with a 256-bit random pattern. In the present experiment clock pulses used in the regenerator stage are tapped from the MLLD output.
In the DPSK regenerator [Fig. 2(b)], the incoming DPSK signal is first demodulated to OOK signal by a DI. After that the OOK signal is amplitude-regenerated by cascadedMamyshev-type 2R regenerators. In this experiment two-stage regeneration is performed in bidirectional configuration using only one HNLF spool . This saves the number of costly HNLFs in exchange for using a pair of circulators. Effect of Rayleigh backscattering in the bidirectional operation is negligible at 10Gb/s signal speed and will not be significant al least up to 40Gb/s .
The HNLF (HNLF2) has zero-dispersion wavelength λ0=1560nm, dispersion slope dD/dλ=0.03ps/nm2/km, length L=1.8km, and nonlinearity coefficient γ ~12/W/km. The signal wavelength is shifted by 2.5nm at both regeneration stages but in opposite directions in wavelength. Bandwidth of the OBPFs for spectrum slicing is 1nm. The amplitude-regenerated data pulses at 1548.5nm together with the clock (probe) pulses derived from the MLLD at 1553nm are introduced to the third HNLF (HNLF3) acting as an all-optical phase modulator. HNLF3 has dispersion D=2.2ps/nm/km and L=2.4km. Walk-off time between the data and probe pulses is 24ps and the timing between the two pulse trains is adjusted by a variable delay line so that complete walk-through between the control and probe pulses takes place in the fiber. The polarizations of the data and probe signals are aligned by the use of polarization controllers (PC) and polarizers (POL) before their entering the HNLF. The power of the control pulses is chosen so that the phase shift induced to the probe pulse via the cross-phase modulation (XPM) is equal to π.
4. Results and discussion
Figure 3(a) shows an optical eye pattern of the DPSK signal generated by the setup of Fig. 2(a) and measured at location B with a sampling oscilloscope having a bandwidth of 30 GHz. Appreciable amplitude fluctuation on top of the pulses is caused by the nonlinearity in HNLF1 for the wavelength conversion that is operated near the zero-dispersion wavelength with relatively high launched power. (λ0, dD/dλ, L, and γ of HNLF1 are 1556nm, 0.026 ps/nm2/km, 1.5km, and ~12/W/km respectively, and averaged launched signal power is 11.2dBm.)
Figures 3(b) and 3(c) show the demodulated OOK signals before and after the 2R regenerator, respectively. Figure 3(c) is the waveform measured when the signal powers launched to the HNLF in the forward and backward directions in the first- and second-stage regeneration, respectively, are optimally chosen so that bit error in the final signal output from the DPSK regenerator becomes minimum. It is shown that the cascaded 2R regenerator removes amplitude noise both of pulses and in empty time slots. In Fig. 4 we plot power transfer characteristics of the 2R regenerator. Solid and dashed curves are the output power versus input power for the first- and second-stage regeneration, respectively. The power transfer of the second-stage regeneration is measured at optimal noise reduction by the first stage regeneration, i.e., the input signal power to the first-stage regenerator is chosen near the value corresponding to the peak in the first-stage power transfer curve. Shapes of the power transfer curves are remarkably different although the same fiber, but in opposite directions, is used. There are several reasons for the difference: (1) signal launch condition is different for the two directions, especially in chirp and width of the pulses, (2) fiber dispersion at the center wavelength of the input signal is different because of the presence of dispersion slope, i.e., - 0.35 and -0.27ps/nm/km for the forward and backward directions, respectively, and (3) the power transfer curve is made obscure more in the first-stage regeneration than in the second-stage regeneration because the amplitude fluctuation in the input pulses is larger in the first stage regeneration. The large oscillation in the output power near the peak of the power transfer curves is undesirable for suppression of amplitude noise. Optimization of parameters of the amplitude regenerator such as fiber dispersion, length, and frequency offset for flatter power transfer characteristics would lead to better performance of the DPSK regenerator .
Figure 3(d) shows the waveform of the probe pulse after XPM-based phase modulation in the HNLF3 and after removal of the control pulses by the subsequent OBPF. The averaged power of the control pulses launched to the HNLF3 is 11dBm. It is seen that the XPM-based phase modulation does not cause amplitude fluctuation. Comparison between Figs. 3(a) and 3(d) shows that signal amplitude is well regenerated. Actual pulse widths measured by an auto-correlator are 4.3, 4.7, 4.4, and 5.3ps for the pulses shown in Figs. 3(a), 3(b), 3(c), and 3(d), respectively.
Although short pulses with duty ratio ~ 5% are used in this experiment, the principle of the regenerator works for wider RZ pulses with duty ratio up to ~ 30%. The maximum duty ratio is mainly determined by the performance of the 2R amplitude regenerator for the noise reduction . Wider pulses, if they are isolated, can be regenerated by suitably scaling the length and dispersion of the highly nonlinear fiber, the filter offset, and the launched pulse power. For the pulses with duty ratio larger than ~ 30%, however, interaction between neighboring pulses becomes appreciable at the end of the highly nonlinear fiber before spectral filtering, which degrades the noise reduction performance. A pulse compression stage before 2R may be needed for such high duty ratio pulse trains . Low duty ratios are also preferred because they relax the accuracy of timing alignment of control and probe pulses in the XPM-based phase modulator for perfect walk-through between them to be ensured.
Figure 5 shows the result of bit-error-rate (BER) measurement for the input signal (dashed curve with triangles) and for the output signal (solid curve with circles). The receiver consists of a preamplifier, a tunable OBPF, DI, a balanced detector, RF amplifier, and a lowpass filter, followed by an error detector. Different programmed bit patterns are used for the error count of the input and output DPSK signals because of the change in logic by the regenerator. It can be seen in Fig. 5 that the regeneration gives a small negative penalty indicating successful regeneration of DPSK signals. It is noted that the negative penalty does not mean that the BER is improved by the regenerator itself. Because the receiver performance is not precisely optimized for the input and output signals at each BER measurement (because of different wavelengths and pulse widths between the input and output signals), the measured BERs are not those determined by the signals themselves. Noise in the receiver after detection also contributes to bit errors. The almost no or even slightly negative penalty shows that the incoming data encoded on the input signal pulses are correctly transferred to the output signal without introducing excess phase and amplitude noise. Dash-dotted curve with crosses in Fig. 5 is the BER measured when the demodulated OOK signal is directly fed to the phase modulator stage without amplitude regeneration. Large penalty and error floor appear, which indicates the importance of the amplitude regeneration of the demodulated OOK signal . Figure 6 shows received eye patterns of the DPSK signals after the balanced detection. Figures 6(a), 6(b), and 6(c) correspond to the input signal to the regenerator, the output signal from the regenerator, and the output signal when the 2R amplitude regenerator after the DPSK-to-OOK demodulation is removed. Figure 6(b) shows that clear eye opening is obtained after the DPSK regenerator.
Finally Fig. 7 is the receiver sensitivity versus the averaged power of the control pulses launched to the HNLF in the final phase modulator. The receiver sensitivity is defined as the received power at which BER of 10-9 is achieved. Power penalty appears as the power of control signal deviates from that corresponding to phase modulation of π in the XPM-based phase modulator. Tolerance of the control signal power of 2.1dB is seen for the power penalty less than 1dB.
An experiment of all-optical DPSK signal regeneration was reported. Cascaded fiber-based 2R amplitude regenerators were used to eliminate amplitude noise of the signal after DPSK-to- OOK format conversion by a delay interferometer. All-optical phase modulation based on XPM in a nonlinear fiber was then performed for remodulation of phase information on clean clock pulses. A small negative penalty was observed when the DPSK regenerator was inserted between the transmitter and the receiver.
In this paper we confirmed the fundamental regenerator operation of transferring phase information from input to output signals without introducing excess amplitude and phase noise. Statistics of noise on the input signal was not that of commonly encountered amplified spontaneous emission (ASE). Direct measurement of reduction of both amplitude and phase noise for input signals degraded by ASE should be performed. For the effectiveness of the regenerator to be truly demonstrated, furthermore, actual reduction of BER should be measured by an experiment including signal transmission before and after the regenerator. Performing such experiments will be a next task in this study.
In practical systems, polarization sensitivity of the XPM-based all-optical phase modulation should be avoided. The XPM operation independent of polarization of control pulses will be realized by the use of circular birefringence HNLF .
References and links
1. J. -C. Simon, M. Gay, L. Bramerie, V. Roncin, M. Joindot, T. Chartier, S. Lobo, G. Girault, Q. T. Le, T. N. Nguyen, and M. N. Ngo, “Long distance transmission using optical regeneration,” in 2008 Optical Fiber Communication Conference and Exposition and The National Fiber Optic Engineers Conference, Technical Digest (CD) (Optical Society of America, 2008), paper OWS1. [PubMed]
2. K. P. Ho, Phase-Modulated Optical Communication Systems (Springer, 2005).
3. K. Croussore, I. Kim, C. Kim, Y. Han, and G. Li, “Phase-and-amplitude regeneration of differential phase-shift keyed signals using a phase-sensitive amplifier,” Opt. Express 14, 2085–2094 (2006). [CrossRef] [PubMed]
4. A. Bogris and D. Syvridis, “RZ-DPSK signal regeneration based on dual-pump phase-sensitive amplfication in fibers,” IEEE Photon. Technol. Lett. 18, 2144–2146 (2006). [CrossRef]
5. I. Kang, C. Dorrer, L. Zhang, M. Rasras, L. Buhl, A. Bhardwaj, S. Cabot, M. Dinu, X. Liu, M. Cappuzzo, L. Gomez, A. Wong-Foy, Y. F. Chen, S. Patel, D. T. Neilson, J. Jaques, and C. R. Giles, “Regenerative all optical wavelength conversion of 40-Gb/s DPSK signals using a semiconductor optical amplifier Mach- Zehnder interferometer,” 2005 European Conference on Optical Communication, paper Th 4.3.3 (2005). [CrossRef]
6. P. Vorreau, A. Marculescu, J. Wang, G. Böttger, B. Sartorius, C. Bornholdt, J. Slovak, M. Schlak, C. Schmidt, S. Tsadka, W. Freude, and J. Leuthold, “Cascadability and regenerative properties of SOA alloptical DPSK wavelength converters,” IEEE Photon. Technol. Lett. 18, 1970–1972 (2006). [CrossRef]
7. R. Elschner, A. M. de Melo, C. -A. Bunge, and K. Petermann, “Noise suppression properties of an interferometer-based regenerator for differential phase-shift keying data,” Opt. Lett. 32, 112–114 (2007). [CrossRef]
8. M. Matsumoto, “A fiber-based all-optical 3R regenerator for DPSK signals,” IEEE Photon. Technol. Lett. 19, 273–275 (2007). [CrossRef]
9. V. S. Grigoryan, M. Shin, P. Devgan, and P. Kumar, “Mechanism of SOA-based regenerative amplification of phase-noise degradaed DPSK signals,” Electron. Lett. 41, 1021–1022 (2005). [CrossRef]
12. E. S. Awad, P. S. Cho, and J. Goldhar, “All-optical re-phasing, re-shaping, and re-amplification of RZDPSK data,” in 2007 Optical Fiber Communication Conference and Exposition and The National Fiber Optic Engineers Conference, Technical Digest (CD) (Optical Society of America, 2007), paper JThA53. [CrossRef] [PubMed]
13. A. G. Striegler, M. Meissner, K. Cvecek, K. Spnsel, G. Leuchs, and B. Schmauss, “NOLM-based RZ-DPSK signal regeneration,” IEEE Photon. Technol. Lett. 17, 639–641 (2005). [CrossRef]
14. M. Matsumoto, “Regeneration of RZ-DPSK signals by fiber-based all-optical regenerators,” IEEE Photon. Technol. Lett. 17, 1055–1057 (2005). [CrossRef]
15. S. Boscolo, R. Bhamber, and S. K. Turitsyn, “Design of Raman-based nonlinear loop mirror for all-optical 2R regeneration of differential phase-shift-keying transmission,” IEEE J. Quantum Electron. 42, 619–624 (2006). [CrossRef]
16. K. Croussore and G. Li, “Amplitude regeneration of RZ-DPSK signals based on four-wave mixing in fibre,” Electron. Lett. 43, 177–178 (2007). [CrossRef]
17. K. Cvecek, K. Sponsel, R. Ludwig, C. Schubert, C. Stephan, G. Onishchujkov, B. Schmauss, and G. Leuchs, “2R-regeneration of an 80-Gb/s RZ-DQPSK signal by a nonlinear amplifying loop mirror,” IEEE Photon. Technol. Lett. 19, 1475–1477 (2007). [CrossRef]
18. M. Matsumoto, “A fiber-based all-optical regenerator for DQPSK signals,” 2007 European Conference on Optical Communication, paper P069 (2007). [CrossRef]
19. Z. Zheng, L. An, Z. Li, X. Zhao, and X. Liu, “All-optical regeneration of DQPSK/QPSK signals based on phase-sensitive amplification,” Opt. Commun. 281, 2755–2759 (2008). [CrossRef]
20. C. Schmidt-Langhorst, R. Ludwig, M. Galili, B. Huettl, F. Futami, S. Watanabe, and C. Schubert, “160 Gbit/s all-optical OOK to DPSK in-line format conversion,” 2006 European Conference on Optical Communication, paper Th4.3.5 (2006).
21. M. Matsumoto, Y. Shimada, and H. Sakaguchi, “Wavelength-shift-free SPM-based 2R regeneration by bidirectional use of a highly nonlinear fiber,” in 2007 Optical Fiber Communication Conference and Exposition and The National Fiber Optic Engineers Conference, Technical Digest (CD) (Optical Society of America, 2007), paper OME5. [CrossRef] [PubMed]
22. L. Provost, C. Finot, P. Petropoulos, K. Mukasa, and D. J. Richardson, “Design scaling rules for 2R-optical self-phase modulation-based regenerators,” Opt. Express 15, 5100–5113 (2007). [CrossRef] [PubMed]
23. T. -H. Her, G. Raybon, and C. Headley, “Optimization of pulse regeneration at 40 Gb/s based on spectral filtering of self-phase modulation in fiber,” IEEE Photon. Technol. Lett. 16, 200–202 (2004). [CrossRef]
24. T. Tanemura, J. H. Lee, D. Wang, K. Katoh, and K. Kikuchi, “Polarization-insensitive 160-Gb/s wavelength converter with all-optical repolarizing function using circular-birefringence highly nonlinear fiber,” Opt. Express 14, 1408–1412 (2006). [CrossRef] [PubMed]