A “black-box” phase sensitive amplifier is presented achieving simultaneous suppression of deterministic phase distortion on two independent 42.66 Gbit/s DPSK modulated signal wavelengths.
©2011 Optical Society of America
The exponential growth of internet data traffic is driving enormous research efforts towards optical technologies that will be able to serve the associated capacity demands and increase system’s reach . In this respect, signal regeneration plays a key role in preventing the accumulation of transmission impairments and thus supporting the development of flexible and scalable physical layer infrastructures for high capacity networking [2–4]. The high maturity level of opto-electronic approaches may be combated by the low cost and low power consumption potential of their all-optical counterparts. However, to become viable and competitive candidates for next generation networks, all-optical regenerators must also support the recently emerged phase encoded modulation formats  providing efficient compensation of the accumulated phase distortions [3,6]. Furthermore, to minimize the cost per transported bit, solutions that simultaneously process multiple wavelength channels, e.g. through the use of a common nonlinear medium , are preferred.
Until recently, all-optical regenerators were mostly designed to suppress amplitude distortion  and initial schemes for phase encoded signals focused on phase preserving amplitude regeneration [8,9] or phase-to-amplitude format conversion and subsequent amplitude regeneration [10–12]. Recently, direct removal of the phase distortion by exploiting the unique phase squeezing capabilities of a phase sensitive amplifier (PSA), has been demonstrated . Operating the PSA in the saturation regime enables the amplitude distortion to be reduced as well . Initial regenerative schemes with PSAs were based on single pump degenerate four-wave mixing (FWM) in a fiber interferometer [13,15]. Phase regeneration using dual pump degenerate FWM process was shown initially in . By adding the phase locking scheme proposed in , the first practical “black-box” PSA based regenerator for differential phase shift keying signals (DPSK) has been demonstrated in . Finally, phase noise suppression on a DPSK modulated signal channel using dual pump non-degenerate FWM has been shown in .
Multi-wavelength regeneration has been demonstrated to date only for on-off keying (OOK) signal formats  and mostly with fiber based schemes. Amplitude noise suppression on multiple channels has been proposed through soliton compression enabled by suitable dispersion management  or quasi-continuous filtering . Proper dispersion management can be also introduced to extend Mamyshev’s technique , for multi-wavelength operation [22,23]. Recently, the principle was applied in a polarization maintaining fiber, with bidirectional propagation, to regenerate four channels simultaneously . On the other hand, PSAs have demonstrated multi-wavelength performance only as linear low noise amplifiers achieving sub 3 dB noise figure performance, with a single pump non-degenerate FWM based scheme .
Recently, we experimentally reported simultaneous phase regeneration on two 42.66 Gbit/s DPSK input signals using a phase sensitive amplifier (PSA) that is based on multi-pump degenerate four-wave mixing (FWM) . In this paper, the PSA construction as a “black-box” subsystem and use as an in-line module is discussed along with details of the device optimization. Our scheme offers independent processing of each signal wavelength using a single nonlinear medium. Performance evaluation with periodic input phase distortion showed improvement in terms of receiver sensitivity by more than 10 dB, for both channels.
2. Principle and experiment setup
The experimental setup of the proposed regeneration scheme is illustrated in Fig. 1 . At the input we have considered two independently modulated DPSK signals at 42.66Gbit/s. Each signal was produced by modulating separate laser sources, at 1549.91 nm and 1550.71 nm, with de-correlated pseudorandom binary sequences (PRBS) of 231-1 pattern length encoded using separate Mach Zehnder modulators. Pure phase distortion was added to both channels using a single phase modulator driven by a 14 GHz square wave from an independent device. The two channels were subsequently amplified to 14 dBm by an Erbium doped fibre amplifier (EDFA) and launched into the phase regenerator.
An in-line or “black-box” PSA based regenerator requires phase synchronization of its local pumps to the incoming signals. The synchronization stage should involve extraction of the signal carrier phase, translation to the required pump wavelengths and subsequent phase locking of the various signals . The two DPSK channels were combined with a strong CW pump signal at 1547.91 nm. The three signals were amplified to a total power of 30 dBm and launched into a highly nonlinear fibre (HNLF). The HNLF used in this section was a strained aluminous-silicate highly nonlinear fibre whose increased stimulated brillouin scattering (SBS) threshold alleviated the need for active SBS suppression. The fibre had a length of 192 m and at 1550 nm the dispersion parameter was −0.2 ps/(nm·km), the nonlinear coefficient 7.4 W−1·km−1 and the attenuation 14 dB/km. The optical spectrum at the output of the HNLF, taken at point B, is depicted in Fig. 1(b). The FWM interaction of the pump signal with each one of the two DPSK channels extracted a corresponding modulation stripped carrier at symmetric wavelength position. Each of the extracted carriers was self-locked to its data signal and the common local pump, maintaining the required frequency condition for PSA performance among the interacting waves [16,28].
After the carrier extraction stage the signal was split, and in one path, the generated carriers were selected by an optical filter and were used to optically injection lock two additional local pump lasers. The injected power of each carrier was kept below −35 dBm, at a level where the limited bandwidth (< 0.5 GHz) of the injection locking process rejected any residual phase modulation on the input signal . Operating the synchronization units at such low bandwidths is a necessary condition for a black-box PSA to achieve broadband phase noise suppression .The emitted power of each local pump laser was 8 dBm. The two injection locked pump lasers from one path and the signal and common pump from the other were directed to a wavelength selective switch (WSS). The WSS suppressed the unwanted spectral components that had been generated from the FWM process in the carrier extraction stage, see Fig. 1(b). The filtered spectrum at the output of the WSS is depicted in Fig. 1(c). It includes the common CW pump (pump 1), the two propagated DPSK signal channels and their corresponding phase locked CW pump waves (pump 2, pump 3) at the symmetric wavelength positions of 1551.91 nm and 1553.52 nm, respectively.
After the WSS the combined signals were amplified to 30 dBm and launched into a second HNLF for the phase sensitive regeneration process, which was similar to the fibre of the carrier extraction stage but with a length of 210 m. The interferometric nature of the scheme made it vulnerable to slow temperature induced path length variations which accordingly affected the phase sensitive gain of the PSA. Since the input DPSK channels were not phase correlated and their corresponding pumps, out of the injection locking process, followed different paths in the interferometer, separate compensation against phase drifts was provided. This was achieved with a feedback circuit setup that used both channels at the output of the PSA as error signals to control two different piezoelectric fibre stretchers (PZTs), one for each pump path. A 20 kHz dithering tone was added onto the control signals of the PZTs, and subsequently extracted with corresponding lock-in amplifiers, in order to lock each channel at the maximum or minimum phase sensitive gain states of the PSA. Figure 1(d) depicts the resulting spectra at the output of the regeneration stage of the PSA, taken at point D, when the channels were simultaneously locked at the maximum or minimum phase sensitive gain, respectively. A gain contrast ratio of 11 dB was achieved for both channels. Figure 1(e) depicts the output spectra when the two channels were locked at two different phase sensitive gain states.
The interaction of the three pumps in the HNLF created strong FWM products that could potentially overlap with the incoming signals. To avoid such scenario a careful selection of the pump wavelengths with respect to the incoming signal locations was required, see Fig. 1(d),(e). More specifically, the frequency distance of the common pump with each signal should be a non-integer multiple of their channel spacing. Additional data signals from the Bragg scattering and phase conjugation processes within the HNLF were created, spanning a large wavelength spectrum. A final WSS was required at the output of the PSA to select the two signals and remove the unwanted wavelengths. Finally, at the receiver section an 1-bit delay interferometer was used for demodulating the channels. The output of the interferometer was connected to a balanced photodiode and phase locked loop clock recovery circuit for detecting the signal and for performing the BER analysis.
The high complexity of the system required rigorous optimization of each stage to maximize the overall performance. At the carrier extraction stage the optimum power level of the local pump with respect to the power of the input signal had to be optimized so that the carriers could be extracted without degrading the signal quality. This was achieved by measuring the bit error rate (BER) of both channels at the output of the 1st HNLF and characterizing the penalty in their received sensitivity (defined at BER: 10−9) as a function of the pump-to-signal power ratio Ppump1/Pchannel(n), where Ppump1 is the power level of the pump-1 and Pchannel(n) (n = 1, 2) the power level of either of the two DPSK channels, Pchannel(1) = Pchannel(2), measured also at the output of the 1st HNLF. The results of this analysis are depicted in Fig. 2 . The power level of the EDFA at the input of the HNLF was always kept constant and equal to 30 dBm, therefore, lower values of the pump-to-signal power ratio effectively corresponded to higher launched signal powers. This caused degradations due to inter-channel nonlinearities and led to penalties that approached 2 dB. The penalty was reduced to less than 0.5 dB when the pump-to-signal power ratio became larger than 3 dB. For the remainder of our study this ratio was set at 10 dB.
At the second stage of the PSA two critical parameters had to be optimized, namely, the powers of the local pumps as well as the relative level of the signal power with respect to the pumps. We characterized the sensitivity penalty of the two channels at the output of the PSA by varying both parameters with the help of the first WSS, see Fig. 3(a-b) . In addition, we evaluated the phase suppression capabilities of the PSA by measuring the introduced improvement in terms of receiver sensitivity when the input channels were distorted by 3 dB at the input of the PSA.
Figure 3(a) shows the measured penalty for both channels as a function of the ratio of the pump-1 power P'pump1 to the power level P'pump(n) (n = 2, 3) of either of the two pump waves (pump −2, pump −3), i.e. P'pump1/P'pump(n), where P'pump(2) = P'pump(3). These power levels were measured at the output of the 2nd HNLF fibre, whilst at its input an EDFA set the power to a constant total level of 30 dBm. The common pump (pump-1) interacted with pump-2 and pump-3, therefore at low power ratios it was strongly depleted. As a result, the effective noise figure of the PSA was degraded and an additional sensitivity penalty was introduced. On the other hand, when pump 1 became too large the phase sensitive gain of the amplifier was reduced due to the low power of pumps 2 and 3. The optimum operating region was achieved for pump-to-pump power ratios of around 5 dB, where the sensitivity penalty was less than 0.5 dB for both channels. Subsequently, the two channels were subjected to phase distortion, which degraded their received sensitivities at the input of the PSA by 3 dB. The PSA improved the channels’ performance by suppressing the distortion. The corresponding improvement in receiver sensitivity was defined as regeneration and it has been measured for both channels as a function of the pump-to-pump power ratio, see Fig. 3(a). Best performance occurs at the optimum operating point of the undistorted case, where the regeneration approaches 2 dB.
Having optimized the pump power profile of the PSA the next step was to identify the optimum power level of the two signals. The results of this optimization process are depicted in Fig. 3(b). More specifically, the figure shows the sensitivity penalty for both channels as a function of the ratio of the pump-1 power P'pump1, to the power level of either of the two DPSK signals P'channel(n) (n = 1, 2) i.e. P'pump1/ P'channel(n), where P'channel(1) = P'channel(2)These power levels were defined at the output of the 2nd HNLF. At low ratios the signal power level effectively increased and the regenerator operated under saturation. This gave rise to stronger nonlinear interactions between the two data signals as well as among the data signals and the common pump-1, which degraded by a relatively small amount the performance of the PSA. However, much higher penalties appeared when the pump-to-signal power ratio was set beyond 12 dB. This was due to the excess signal attenuation that was required by the WSS to set this power ratio and which increased the effective noise figure of the PSA. The optimum point appeared at the value of 8 dB, and resulted in a penalty of 0.5 dB. The figure shows also the results of a similar optimization process when the input channels have experienced a phase degradation that corresponded to 3 dB sensitivity penalty at the input of the PSA. At the optimum operating point a sensitivity improvement of 2 dB was achieved for both signals.
We finally assessed the regenerative capabilities of the PSA against various levels of input phase distortion, as shown in Fig. 4 . Despite the presence of a small offset attributed to OSNR degradation and inter-channel crosstalk, Fig. 4 demonstrates that the achieved regeneration nearly tracks the penalty induced by the phase distortion, offering over 10 dB improvement in the receiver sensitivity, when the applied phase distortion induces a 12 dB penalty. Figure 5(a)-(b) depict measured BER curves for both channels taken against total received power, for input degradations of about ~10 dB at 10−9. It is also shown that in the absence of any input degradation the receiver sensitivity penalty of each channel at 10−9 was less than 1 dB, and error free operation was achieved. Furthermore, when introducing significant phase distortion on the input channels, the PSA demonstrated the anticipated phase squeezing capabilities. The restored eye diagrams are illustrated in Fig. 5(c) and Fig. 5(d). Although balanced detection has been used, the monitored eye diagrams refer only to the duo-binary output of the one-bit delay interferometer. Therefore they are asymmetric and their upper rails carry more noise with respect to the lower rails as they result from signal-to-spontaneous emission beating at the photodiode. However, considering this asymmetry, we believe that both rails are regenerated. Finally, we have to point out that if random phase noise is considered at the input, the regenerative performance of the PSA is expected to be less efficient. This could be attributed partially to a non-fragmented distribution of the introduced phase noise , as well as to an enhanced amplitude distortion from phase-to-amplitude conversion in the HNLF .
We have proposed and experimentally implemented, for the first time to our knowledge, an in-line PSA with multichannel regenerative capabilities. The regenerative performance of our scheme has been evaluated against periodic phase distortion on two 42.66 Gbit/s DPSK signals. Significant recovery of the undistorted signal quality has been achieved for both channels, quantified by more than 10 dB improvement in terms of receiver sensitivity.
This work has been supported and by European Communities Seventh Framework Programme FP/2007-2013 under grant agreement 224547 (PHASORS) and by Science Foundation Ireland under Grant 06/IN/I969
References and links
1. A. D. Ellis, J. Zhao, and D. Cotter, “Approaching the non-linear Shannon limit,” J. Lightwave Technol. 28(4), 423–433 (2010). [CrossRef]
2. O. Leclerc, B. Lavigne, E. Balmefrezol, P. Brindel, L. Pierre, D. Rouvillain, and F. Seguineau, “Optical regeneration at 40 Gbit/s and beyond,” J. Lightwave Technol. 21(11), 2779–2790 (2003). [CrossRef]
3. S. J. Ben Yoo, “All-optical regeneration for ultra-long fiber links and its prospects for future applications in new modulation formats,” in Proc OFC 2009, paper OThS4.
4. S. Sygletos, I. Tomkos, and J. Leuthold, “Technological challenges on the road toward transparent networking,” J. Opt. Netw. 7(4), 321–350 (2008). [CrossRef]
5. P. J. Winzer and R.-J. Essiambre, “Advanced optical modulation formats,” Proc. IEEE 94(5), 952–985 (2006). [CrossRef]
6. M. Matzumoto, “Optical regenerators for novel modulation schemes,” in Impact of Nonlinearities on Fiber Optic Communications, Optical and Fiber Communication Reports 7, S. Kumar, ed. (Springer 2011).
7. F. Parmigiani, L. Provost, P. Petropoulos, D. J. Richardson, W. Freude, J. Leuthold, A. D. Ellis, and I. Tomkos, “Progress in multichannel all-optical regeneration based on fiber technology,” IEEE J. Sel. Top. Quantum Electron. (to be published), DOI: 10.1109/JSTQE.2011.2126040.
9. K. Cvecek, K. Sponsel, R. Ludwig, C. Schubert, C. Stephan, G. Onishchukov, B. Schmauss, and G. Leuchs, “2R-regeneration of an 80 Gbit./s RZ DPSK signal by a nonlinear amplifying loop mirror,” IEEE Photon. Technol. Lett. 19(19), 1475–1477 (2007). [CrossRef]
11. 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. Jacques, and C. R. Giles, “Regenerative all optical wavelength conversion of a 40 Gb/s DPSK signals using a SOA MZI,” in Proc. ECOC 2005, Thu 4.3.3.
12. 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 all-optical DPSK wavelength conversion,” IEEE Photon. Technol. Lett. 18(18), 1970–1972 (2006). [CrossRef]
13. K. Croussore, C. Kim, and G. Li, “All-optical regeneration of differential phase-shift keying signals based on phase-sensitive amplification,” Opt. Lett. 29(20), 2357–2359 (2004). [CrossRef] [PubMed]
14. K. Croussore and G. Li, “Phase Regeneration of NRZ-DPSK Signals Based on Symmetric-Pump Phase-Sensitive Amplification,” IEEE Photon. Technol. Lett. 19(11), 864–866 (2007). [CrossRef]
15. M. E. Marhic and C. H. Hsia, “Optical amplification and squeezed-light generation in fibre interferometers performing degenerate four-wave mixing,” IEEE J. Sel. Top. Quantum Electron. 3(6), 341–358 (1991). [CrossRef]
16. R. Weerasuriya, S. Sygletos, S. K. Ibrahim, R. Phelan, J. O’Carroll, B. Kelly, J. O’Gorman, and A. D. Ellis, “Generation of frequency symmetric signals from a BPSK input for phase sensitive amplification,” in Proc. OFC 2010, OWT6,San Diego CA, USA, 21–25 March 2010.
17. R. Slavík, F. Parmigiani, J. Kakande, C. Lundström, M. Sjödin, P. Andrekson, R. Weerasuriya, S. Sygletos, A. D. Ellis, L. Grüner-Nielsen, D. Jakobsen, S. Herstrøm, R. Phelan, J. O’Gorman, A. Bogris, D. Syvridis, S. Dasgupta, P. Petropoulos, and D. Richardson, “All-optical phase and amplitude regenerator for next generation telecommunications systems,” Nat. Photonics 4(10), 690–695 (2010). [CrossRef]
18. P. Frascella, S. Sygletos, F. C. G. Gunning, R. Weerasuriya, L. Grüner-Nielsen, R. Phelan, J. O’Gorman, and A. D. Ellis, “DPSK signal regeneration with a dual-pump nondegenerate phase sensitive amplifier,” IEEE Photon. Technol. Lett. 23(8), 516–518 (2011). [CrossRef]
19. T. Ohara, H. Takara, A. Hirano, K. Mori, and S. Kawanishi, “40-Gb/s × 4-channel all-optical multichannel limiter utilizing spectrally filtered optical solitons,” IEEE Photon. Technol. Lett. 15(5), 763–765 (2003). [CrossRef]
21. P. V. Mamyshev, “All-optical data regeneration based on self-phase modulation effect,” in Proc ECOC 1998, 475–476.
23. C. Kouloumentas, L. Provost, F. Parmigiani, S. Tsolakidis, P. Petropoulos, I. Tomkos, and D. J. Richardson, “Four-channel all-fiber dispersion managed 2R regenerator,” IEEE Photon. Technol. Lett. 20(13), 1169–1171 (2008). [CrossRef]
24. L. Provost, F. Parmigiani, K. Mukasa, M. Takahashi, J. Hiroishi, M. Tadakuma, P. Petropoulos, and D. J. Richardson, “Simultaneous all-optical 2R regeneration of 4×10 Gbit/s wavelength division multiplexed channels,” in Proc. ECOC 2007, Di 4.5.1.
25. Z. Tong, C. Lundström, P. A. Andrekson, C. J. McKinstrie, M. Karlsson, D. J. Blessing, E. Tipsuwannakul, B. J. Puttnam, H. Toda, and L. Grüner-Nielsen, “Towards ultrasensitive optical links enabled by low-noise phase-sensitive amplifiers,” Nat. Photonics 5(7), 430–436 (2011). [CrossRef]
26. P. Frascella, S. Sygletos, S. K. Ibrahim, L. Grüner-Nielsen, R. Phelan, J. O’Gorman, and A. D. Ellis, “Two channel phase regeneration based on a “black-box” Phase Sensitive Amplifier,” in Proc. ECOC 2011, Th. 11. LeCervin.2.
27. S. Sygletos, R. Weerasuriya, S. K. Ibrahim, F. Gunning, R. Phelan, J. O’Gorman, J. O’Carrol, B. Kelly, A. Bogris, D. Syvridis, C. Lundström, P. Andrekson, F. Parmigiani, D. J. Richardson, and A. D. Ellis, “Phase locking and carrier extraction schemes for phase sensitive amplification,” in Proc ICTON (2010).
29. E. K. Lau, L. J. Wong, X. Zhao, Y.-K. Chen, C. J. Chang-Hasnain, and M. C. Wu, “Bandwidth enhancement by master modulation of optical injection-locked lasers,” J. Lightwave Technol. 26(15), 2584–2593 (2008). [CrossRef]
30. R. Slavik, A. Bogris, F. Parmigiani, J. Kakande, M. Westlund, M. Sköld, L. Grüner-Nielsen, R. Phelan, D. Syvridis, P. Petropoulos, and D. J. Richardson,“Bandwidth enhancement by master modulation of optical injection-locked lasers,” to appear in IEEE J. Selected Topics in Quantum Electron., DOI: . [CrossRef]