Simultaneous QPSK-to-\(2\times\) BPSK Wavelength and Modulation ...

IEEE PHOTONICS TECHNOLOGY LETTERS, VOL. 26, NO. 12, JUNE 15, 2014

1207

Simultaneous QPSK-to-2×BPSK Wavelength and Modulation Format Conversion in PPLN Francesco Da Ros, Kjeld Dalgaard, Yutaka Fukuchi, Jing Xu, Michael Galili, and Christophe Peucheret

Abstract— Phase-sensitive cascaded second-harmonic generation and difference-frequency generation in a periodically poled lithium niobate waveguide allow converting two orthogonal quadratures of an optical field to different wavelengths, thus enabling simultaneous quadrature phase-shift keying-to2×binary phase-shift keying modulation format and wavelength conversions. Static phase-sensitive extinction ratios above 20 dB are obtained for both quadratures, resulting in error-free dynamic operation with low penalty (BER 10−9 ) at 10 Gbaud. Index Terms— Phase shift keying, nonlinear optics, periodically poled lithium niobate (PPLN), phase-sensitive amplification.

I. I NTRODUCTION

I

N RECENT years, the renewed interest in phase-sensitive amplification for all-optical signal processing has been combined with efforts towards improving the spectral efficiency through the use of higher order modulation formats such as quadrature phase-shift keying (QPSK) and quadrature amplitude modulation (QAM), leading to various schemes for phase regeneration of QPSK [1], 8-QAM [2], and 16-QAM signals [3] being demonstrated. Among demonstrations of all-optical signal processing targeting advanced modulation formats, R. P. Webb et al. have proposed the use of phase-sensitive four-wave mixing (FWM) for converting the two complex quadratures of an optical signal to different wavelengths. This functionality is suitable for QPSK-to-2×BPSK modulation format and wavelength conversion [4] and could be employed to enhance the phase noise tolerance of a conventional QPSK balanced receiver [5], or to provide simultaneous Manuscript received January 29, 2014; revised March 20, 2014; accepted April 15, 2014. Date of publication April 18, 2014; date of current version May 16, 2014. This work was supported by the Danish Research Council for Technology and Production Sciences under Project 09-066562. F. Da Ros, K. Dalgaard, and M. Galili are with the Department of Photonics Engineering, Technical University of Denmark, Kgs. Lyngby DK-2800, Denmark (e-mail: [email protected]; [email protected]; [email protected]). Y. Fukuchi is with the Department of Electrical Engineering, Faculty of Engineering, Tokyo University of Science, Tokyo 125-8585, Japan, and also with the Department of Photonics Engineering, Technical University of Denmark, DK-2800 Kgs. Lyngby, Denmark (e-mail: [email protected]). J. Xu was with the Department of Photonics Engineering, Technical University of Denmark, DK-2800 Kgs. Lyngby, Denmark. She is now with the School of Optical and Electronic Information, Huazhong University of Science and Technology, Wuhan 430074, China (e-mail: [email protected]). C. Peucheret was with the Department of Photonics Engineering, Technical University of Denmark, Kgs. Lyngby DK-2800, Denmark. He is now with FOTON Laboratory, University of Rennes 1, Lannion 22305, France (e-mail: [email protected]). Color versions of one or more of the figures in this letter are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/LPT.2014.2318992

quadrature demultiplexing and phase regeneration in a scenario of a middle node in a network where the two binary phaseshift keying (BPSK) quadratures are to be directed to different destinations. In contrast with another recently demonstrated phase-sensitive technique relying on orientating the phasesensitive gain axis to demultiplex the desired quadrature of a QPSK signal [6], this method enables the simultaneous recovery of both BPSK quadratures. Static operation (i.e. using continuous wave (CW) signals) of this scheme has also been demonstrated using semiconductor optical amplifiers (SOAs) as nonlinear media [4]. Furthermore, dynamic operation has been reported in SOAs for a 10.65-Gbaud QPSK signal [7], even though the lack of a phase-locking scheme limited the investigations of the performances. In spite of the operation of the scheme having been numerically predicted at a symbol rate as high as 40 Gbaud, and positive conversion efficiencies having been reported with pump spacings of 600 GHz [8], pattern effects need to be addressed when processing high bit rate signals in SOAs, due to a relatively slow carrier recovery time. To fully exploit the benefits of all-optical signal processing, bit rate transparent operation is a desirable condition. Along such a direction, we have shown the potential for implementing this scheme using the Kerr nonlinearity in highly nonlinear optical fibers (HNLFs) and experimentally investigated its performances statically [9] and dynamically [10]. However, HNLFs are limited by a relatively low nonlinear coefficient, which results in long fibers being required to achieve the necessary nonlinear phase shift. Furthermore, suppressing stimulated Brillouin scattering (SBS) in silica HNLFs is a major challenge when implementing phase-sensitive functionalities. Among other nonlinear media, periodically-poled lithium niobate (PPLN) waveguides offer a more compact implementation, as well as a better SBS immunity, and have been proven effective for achieving phase-sensitive processing exploiting cascaded second-order nonlinearities [3], [11]. In this work, we experimentally investigate the implementation of the orthogonal quadratures converter in PPLN waveguides, showing static phase-sensitive extinction ratios (ERs) in excess of 20 dB for both converted idlers. Furthermore, dynamic operation is also successfully demonstrated, converting the orthogonal quadratures of a 10-Gbaud QPSK signal to two 10-Gbps BPSK signals located at different wavelengths. The converted quadratures can be recovered error-free (i.e. with a bit-error-ratio (BER) better than 10−9 ) with low power penalty compared to standard demodulation

1041-1135 © 2014 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.

1208


TABLE I O PTIMIZATION R ESULTS : P OWER L EVELS AND P HASES OF THE F IVE WAVES AT THE PPLN WAVEGUIDE I NPUT

Fig. 1. Operation principle sketching the waves allocation at the input and output of the PPLN waveguide, together with the phase dependence of the cascaded SHG/DFG enabling the conversion of the two orthogonal quadratures (I and Q) of the optical signal to different wavelengths. s represents the signal phase at the input of the waveguide while 0 is an arbitrary initial phase.

of a differential QPSK signal using a 1-symbol delay interferometer (DI) and balanced detection. II. O PERATION P RINCIPLE AND E XPERIMENTAL S ETUP The converter relies on the use of four phase-coherent CW pumps (denoted as P1-P4) that are injected together with a phase-coherent signal (denoted as S) into a PPLN waveguide (NEL, WH-0780-000-F-B-C), as illustrated in Fig. 1. Phasesensitive processes based on cascaded second-harmonic generation (SHG) and difference-frequency generation (DFG) in the waveguide are functionally equivalent to FWM and can be used to generate two idlers (denoted as I and Q), whose conversion efficiencies are determined by the phase relation between signal and pumps. By optimizing the pumps phases and power levels together with the signal power, the conversion efficiency versus signal phase curves of the two idlers can be phase-shifted by 90◦, resulting in the conversion of the in-phase and quadrature components of S to I and Q, respectively. The experimental set-up is represented in Fig. 2. An optical frequency comb with 40-GHz line spacing was obtained by phase modulating (PM) with modulation index of 4.3 a CW signal generated from a narrow linewidth (100 kHz) external cavity laser at 1562.4 nm with a 40-GHz radio frequency signal. Frequency comb generation was used to guarantee a stable phase relation between the five waves involved in the phase-sensitive process. Alternatively, for practical applications, the pumps could be generated through an additional mixing pre-stage as discussed in [1], [2], and [11]. An optical processor (Finisar, Waveshaper) was used to select four 80-GHz spaced pumps and a signal, located in-between the shortest wavelength pumps, out of the frequency comb and to adjust their power levels and relative phases as indicated in Table I. The values have been optimized numerically and fine tuned experimentally. For the static characterization of the underlying phasesensitive process, all the selected waves were directed towards the same output port of the processor, therefore injecting into the PPLN waveguide five phase-locked CW signals. At the device output, two CW idlers (I and Q) were generated in the empty frequency slots between the pumps, as shown in the spectra of Fig. 3(a). Their normalized conversion efficiencies as a function of the signal phase, reported in Fig. 3(b), have been obtained by sweeping the signal phase with the optical processor and measuring the relevant idler output power using an optical spectrum analyzer. The phase-sensitive

ERs exceed 20 dB for both idlers and the conversion efficiency curves are (90° ± 5°)-shifted in signal phase. For the system experiment, the signal was sent to a different output port of the processor than the pumps, QPSK-modulated at 10 Gbaud using a standard IQ modulator, and recombined with the pumps. The pumps and signal paths have been balanced using 11 m of standard single mode fiber (SMF) to ease the operation of the phase control loop aiming at compensating slow thermal drifts. The states-of-polarization of the waves were aligned to the TE mode of the waveguide and an erbium-doped fiber amplifier (EDFA) was used to couple a total power of 25.6 dBm in the waveguide. The length, fiber-to-fiber insertion loss, conversion efficiency and operation temperature of the waveguide are 3 cm, 3.3 dB, 286%/W and 50◦ C, respectively. The temperature was tuned to set the quasi-phase matching (QPM) wavelength to 1562.4 nm (i.e. P3), thus equalizing the conversion efficiencies of the two idlers within 0.5 dB (Fig. 4). The polarization sensitivity of the waveguide could limit practical applications. However, polarization diversity schemes relying on the use of either two identical devices in parallel [12] or the same device in bidirectional operation [12], [13] have already been reported. At the waveguide output, a pair of tunable optical bandpass filters (TOBPFs) with 0.5-nm and 0.3-nm full-width at halfmaximum (FWHM) bandwidths were used to select one of the idlers at a time. The filtered idler was input to a pre-amplified BPSK balanced receiver for BER testing. An EDFA located between the TOBPFs was used to compensate their insertion losses. Finally, phase-to-intensity demodulation was performed in the receiver by a 1-bit (100 ps) delay interferometer (DI) followed by a balanced photodiode with cut-off frequency of 45 GHz. The splitting of pumps and signal and their propagation along different paths inevitably results in a loss of phase coherence due to thermal effects, even when balancing the paths lengths. In order to lock the waves in phase, 10% of the waveguide output power was taped out for a phase stabilization feedback loop. The idler Q was selected through two fixed OBPFs with 0.3-nm and 0.8-nm FWHM bandwidths, detected by a low-speed avalanche photodiode (APD) and used as reference for a feedback loop based on a piezoelectric actuator (PZT) with a 15-kHz bandwidth. It should be noted that the same reference idler and feedback loop conditions have been used for both received idlers, showing the simultaneous conversion of the two orthogonal quadratures of the original QPSK signal. The phase stabilization mechanism relies on tracking the average power variations of idler Q as the relative phase

DA ROS et al.: SIMULTANEOUS QPSK-TO-2×BPSK WAVELENGTH AND MODULATION FORMAT CONVERSION

Fig. 2.

1209

Experimental setup for dynamic phase-sensitive wavelength and format conversion of a 10-Gbaud QPSK signal.

Fig. 4. Optical spectra at the input and output of the waveguide for a 10-Gbaud QPSK-modulated input signal.

Fig. 3. (a) Optical spectra at the output of the waveguide for different input signal phase values corresponding to maximum conversion efficiency for Q (s = 30°) and I (s = 120°) and quadrature point (s = 75°). (b) Normalized conversion efficiencies for the two idlers (I and Q) as a function of the input signal phase.

minima in the CE curve would be converted into two on-off keying (OOK) modulated idlers. The information carried by both I and Q would then be the result of exclusive OR operation between the datastreams carried by the two quadratures of the original signal, with consequent loss of information. Notice that the conversion efficiencies can be considered linear around the quadrature points, therefore no dithering is required unlike for locking to a maximum or minimum point. More complex phase locking schemes making use of both I and Q could be implemented, potentially providing tighter locking operation. However, such schemes would require extra care taking into account the difference in conversion efficiencies between the two idlers. III. R ESULTS AND D ISCUSSION

between signal and pumps drifts. The phase-sensitive cascaded SHG/DFG directly maps the phase drifts into power variations of the idlers. For instance, applying a low frequency (below 10 kHz) linear phase modulation to the signal through the PZT, the average power of idler Q detected through the APD (3-dB bandwidth of 50 MHz) varies according to the sine square transfer functions reported in Fig. 3(b). The feedback was tuned to lock the original QPSK signal such that the constellation points matched the quadrature points of the conversion efficiency versus input signal phase, i.e. 3 dB below the maximum. The locking point is critical in order to successfully convert the two orthogonal quadratures to I and Q. A misalignment would result in excess phase noise added to the BPSK idlers through the conversion. Furthermore, a QPSK signal aligned to match the position of maxima and

The spectra at the input and output of the PPLN waveguide are reported in Fig. 4. The existence of two modulated idlers is clearly confirmed in the output spectrum, showing output conversion efficiencies above −6 dB. The performances of the two converted idlers have been evaluated through BER measurements. In this analysis, decorrelated pseudo-random bit sequences (PRBSs) of length 27 −1 have been used to drive the IQ modulator. It should be noted that, for the back-to-back QPSK quadratures, the demodulated signal is no longer a PRBS, since no differential encoding had been applied at the transmitter. Instead, the error analyzer was programmed with the expected patterns and the memory available limited the length of the PRBS that could be used. In contrast, the investigated scheme allows direct recovery of the original PRBS by simply differentially detecting the

1210


both idlers. The conversion of both complex quadratures of a 10-Gbaud QPSK signal to two distinct BPSK idlers I and Q has been experimentally demonstrated, reporting error-free performances (BER better than 10−9 ) with low power penalty compared to balanced detection of the QPSK signal using a 1-symbol delay interferometer. The fixed conditions used for the phase-sensitive stabilization scheme confirm the possibility for simultaneous conversion of the two idlers. ACKNOWLEDGMENT Fig. 5. (a) BER performances for the idlers together with back-toback QPSK and BPSK as references. Eye diagrams for (b) one quadrature of the demodulated back-to-back QPSK, and the converted (c) Q and (d) I quadratures (−36 dBm received power).

The authors would like to thank H. C. H. Mulvad for fruitful discussions.

idlers I and Q. The resulting BER curves as a function of the average received power are shown in Fig. 5(a) and compared with back-to-back performances for both a BPSK signal at 10 Gbps and a QPSK signal at 10 Gbaud. The reference BPSK signal was demodulated using the 100-ps DI while, for the reference QPSK, the two quadratures were demodulated one at a time using the same 100-ps DI and tuning its bias point to different values. Both idlers could be detected error-free, with sensitivities of −36 dBm and −33.5 dBm, for I and Q, respectively. A comparison between the detected BPSK idlers and the demodulated back-to-back QPSK signal shows a better sensitivity at low average received power for both idlers. At higher power levels, however, the sensitivity improvement of Q decreases while the performances of I degrade, reaching a power penalty of 2 dB at a BER of 10−9 . The degradation is mainly caused by a sub-optimum locking position for the phase stabilization. As the locking scheme tracks the variations of Q, a slight mismatch in the orthogonality between the conversion efficiencies of the two idlers could explain the difference in performance. A more thorough tuning of the pumps phases and power levels as well as the QPM wavelength is believed to allow achieving similar performances for the two idlers, as shown in [10]. The worsening of the performances is more severe for lower BER values when a longer gating time, hence longer-lasting stability is required. Nevertheless, errorfree performances were obtained for both converted idlers with clear and open eye diagrams (Fig. 5(c) and (d)).

[1] J. Kakande et al., “First demonstration of all-optical QPSK signal regeneration in a novel multi-format phase sensitive amplifier,” in Proc. 36th ECOC, 2010, paper PD 3.3. [2] T. Richter, R. Elschner, and C. Schubert, “QAM phase-regeneration in a phase-sensitive fiber-amplifier,” in Proc. 39th ECOC, 2013, paper We.3.A.2. [3] T. Umeki, O. Tadanaga, M. Asobe, Y. Miyamoto, and H. Takenouchi, “First demonstration of high-order QAM signal amplification in PPLN-based phase sensitive amplifier,” in Proc. 39th ECOC, 2013, paper PD 1.C.5. [4] R. P. Webb, J. M. Dailey, R. J. Manning, and A. D. Ellis, “Phase discrimination and simultaneous frequency conversion of the orthogonal components of an optical signal by four-wave mixing in an SOA,” Opt. Exp., vol. 19, no. 21, pp. 20015–20022, Oct. 2011. [5] F. Da Ros, J. Xu, L. Lei, and C. Peucheret, “Phase noise tolerant QPSK receiver using phase sensitive wavelength conversion,” in Proc. 18th OECC/PS, 2013, paper TuS2-5. [6] M. Gao, T. Kurosu, T. Inoue, and S. Namiki, “Low-penalty phase de-multiplexing of QPSK signal by dual pump phase sensitive amplifiers,” in Proc. 39th ECOC, 2013, paper We.3.A.5. [7] M. J. Power, R. P. Webb, and R. J. Manning, “All-optical phase discrimination using SOA,” Opt. Exp., vol. 21, no. 22, pp. 25664–25669, Nov. 2013. [8] R. P. Webb, M. Power, and R. J. Manning, “Phase-sensitive frequency conversion of quadrature modulated signals,” Opt. Exp., vol. 21, no. 10, pp. 12713–12727, May 2013. [9] F. Da Ros, P. M. Calabrese, N. Kang, E. Palushani, and C. Peucheret, “Orthogonal phase quadratures conversion to different wavelengths through phase-sensitive four wave mixing in an highly nonlinear fiber,” in Proc. OFC, 2013, paper OW4C.3. [10] F. Da Ros, K. Dalgaard, L. Lei, J. Xu, and C. Peucheret, “QPSK-to2×BPSK wavelength and modulation format conversion through phasesensitive four-wave mixing in a highly nonlinear optical fiber,” Opt. Exp., vol. 21, no. 23, pp. 28743–28750, Nov. 2013. [11] B. Puttnam, D. Mazroa, S. Shinada, and N. Wada, “Large phase sensitive gain in periodically poled lithium niobate with high pump power,” IEEE Photon. Technol. Lett., vol. 23, no. 7, pp. 426–428, Apr. 1, 2011. [12] S. L. Jansen, D. van den Borne, P. M. Krummrich, S. Spälter, G.-D. Khoe, and H. de Waardt, “Long-haul DWDM transmission systems employing optical phase conjugation,” IEEE J. Sel. Topics Quantum Electron., vol. 12, no. 4, pp. 505–520, Jul. 2006. [13] A. Albuquerque et al., “Phase-sensitive amplification in a single bidirectional PPLN waveguide,” Opt. Exp., vol. 21, no. 19, pp. 22063–22069, Sep. 2013.

R EFERENCES

IV. C ONCLUSION We have demonstrated simultaneous QPSK-to-2×BPSK modulation format and wavelength conversion using phasesensitive cascaded SHG and DFG processes in a PPLN waveguide. The scheme has been optimized under CW conditions, achieving phase-sensitive ERs in excess of 20 dB for