We demonstrated the simultaneous amplification of a coherent multi-carrier signal using a χ(2)-based non-degenerate phase sensitive amplifier (PSA). The signal-to-noise ratio (SNR), which is degraded by the additional amplified spontaneous emission (ASE) noise, can be recovered due to the gain difference between a phase-correlated signal-idler pair and uncorrelated excess noise. Utilizing the second harmonic pumping of a χ(2)-based PSA enables us to observe the SNR recovery directly by comparing the SNR for the input with that for the PSA output. A 3-dB optical-SNR (OSNR) improvement was obtained as a result of the gain difference. We also achieved a 3-dB SNR improvement in the electric domain by reducing the signal-ASE beat noise. The receiver sensitivity for a 10 Gbit/s phase shift keying signal was clearly improved with the PSA.
©2012 Optical Society of America
The recent rapid growth in data traffic over the internet has led to the need for new technologies capable of achieving higher bandwidth capacities. Recently, ultra-high-speed channel transmission based on a coherent multi-carrier signal has been widely studied for future spectrally efficient high-capacity optical transport systems [1–3]. Such high-capacity systems require the low-noise generation and amplification of coherent multi-carrier signals. An optical phase-sensitive amplifier (PSA) is capable of realizing low noise amplification with breaking the 3-dB quantum-limited noise figure (NF) of a conventional phase-insensitive amplifier (PIA) . In principle, a PSA has a 0-dB NF, which means that there is no degradation of the signal-to-noise ratio (SNR) after signal amplification.
Both χ(2)- and χ(3)-based PSAs have been investigated with two types of parametric processes, namely frequency-degenerate (signal and idler frequencies are identical) [5–8] and non-degenerate (signal and idler frequencies are different) [9–12] PSAs. With the degenerate PSA, only a single channel signal is amplified for a fixed pump configuration. In contrast, the non-degenerate process can achieve simultaneous multi-channel amplification. Furthermore, the non-degenerate PSA has the additional advantage of SNR recovery due to the difference between the gains of correlated signal-idler pairs and uncorrelated noise. The quantum limited NF is defined by the noise of the clean coherent input and amplified signals. However, in actual operation, the input is accompanied by excess noise. A PSA exhibits a higher gain for correlated signal-idler pairs than for uncorrelated noise due to the constructive interference of the correlated signal idler pairs. Consequently, the gain difference enhances the NF performance in a non-degenerate PSA. Recently, it has been experimentally demonstrated that a fiber-based non-degenerate PSA link exhibits nearly 6 dB better NF performance than a conventional erbium-doped fiber amplifier (EDFA) amplified link , which corresponds theoretically to 3 dB better performance than an all-PSA link . It has also been reported that a fiber-based PSA consisting of optical combs has superior noise performance to a PIA .
Although a fiber-based PSA has superior noise performance to a PIA, SNR recovery with respect to the input signal has never been directly observed. The fiber-based PSA can provide a high gain by using a pump laser at the 1.5-μm telecom-band wavelength. The compatibility with an optical fiber communication system is an advantage. However the separation of the signal, idler, and pump is fundamentally difficult. Extrinsic spontaneous emission generated by an EDFA for the pump may degrade the PSA performance. In contrast, a χ(2) -based PSA utilizes the second harmonic of the signal as a pump. Therefore, the signal and pump can be separated in a simple configuration. This enables the direct observation of the noise level after the PSA without extrinsic noise.
In this paper, we propose the simultaneous amplification of coherent multi-carrier signals using a χ(2)-based non-degenerate PSA with periodically poled LiNbO3 (PPLN) ridge waveguides. We confirmed experimentally, for the first time, a 3-dB SNR improvement through the phase sensitive amplification of a multi-carrier pair by reducing the signal-ASE beat noise. By using the proposed PSA, we successfully demonstrate the amplification of a multi-carrier binary phase-shift keying (BPSK) signal with an improved power penalty.
2. Principle and experimental setup
We use a non-degenerate optical parametric process to achieve the phase-sensitive amplification of a multi carrier signal. Figures 1 (a) and 1(b) show an illustration of the operating principle. A χ(2)-based PSA utilizes the second harmonic of the signal as a pump. The pump frequency (2ω0) is exactly double the signal frequency (ω0) for the degenerate parametric process. With the non-degenerate parametric process, phase sensitive operation is achieved by the interaction between a correlated signal-idler pair. The frequencies of the signal (ω0 + Δ) and idler (ω0-Δ) are symmetrically detuned from the center frequency of the degenerate signal (ω0). The non-degenerate PSA gain for the phase correlated and power equalized signal-idler pair will be the same as the degenerate PSA gain. In general, it is difficult to lock the frequency and phase of the signal-idler pair in a non-degenerate PSA. However a coherent multi-carrier signal, which consists of an optical comb generated from a single light source, automatically stabilizes its frequency and the phase of all channels without a complex phase-locking scheme .
Here, we consider that the signal and idler pair is accompanied by uncorrelated noise such as amplified spontaneous emission (ASE) noise. The uncorrelated ASE will be amplified with a smaller gain than that of the correlated signal-idler pair because the ASE phase components include both in-phase (amplified) and quadrature-phase (de-amplified) fields equally. The gain difference means that a better OSNR can be obtained at the PSA output than at the input for both degenerate and non-degenerate signals.
However, the reasons for the reduction in ASE power are different for the two types of signals. With the degenerate configuration, the difference in gain arises from the de-amplification of the quadrature electric field in the ASE. This means that the ASE noise power will be amplified with a 3 dB lower gain than in the PSA. However the in-phase field of the ASE will be amplified with the same gain as the signal and thus have the same intensity beat noise before and after the PSA. In contrast, with the non-degenerate configuration, only a phase matched photon pair in the ASE will be amplified while a mismatched pair will be de-amplified. Since the photons in the ASE have random phases, only half of the in-phase field and half of the quadrature-phase field are amplified. Then, the ASE noise power will amplified with 3 dB lower gain than the PSA, and the gain for the in-phase electric field is half the signal gain. Consequently, the SNR of each carrier is improved for a non-degenerate PSA.
Figure 2 shows the experimental setup. We used a 1.54-µm-band external cavity laser diode (ECLD). A continuous wave (CW) was divided with a 3-dB coupler to provide the pump and signal lights. CW optical carriers with a 20-GHz spacing were generated with a Mach-Zehnder modulator (MZM) . We used a push-pull type MZM to obtain signal and idler pairs with the same phase. The subcarriers were simultaneously modulated by a differential phase shift keying (DPSK) MZM driven with a 10 Gb/s pseudo-random bit sequence (PRBS) pattern. Then, ASE light was intentionally combined with the optical coupler to add uncorrelated noise. The bandwidth of the ASE was adjusted with a band-pass filter (BPF). At the PSA, we used two PPLN ridge waveguides with a high conversion efficiency of over 2000%/W . The PPLN waveguides were assembled into fiber-pigtail modules with four input/output ports to allow SH pumping. The pump light was amplified with an EDFA up to 31.5 dBm and injected into PPLN module 1 to generate an SH wave around 770 nm at a power of about 300 mW. The SH wave and the multi-carrier signal were injected into PPLN module 2 for the optical parametric amplification (OPA). The dichromatic mirrors in the two modules effectively suppressed the unwanted output of the 1.54-µm-band pump and the ASE generated by the EDFA for the pump. To achieve a stable PSA output, a piezoelectric transducer (PZT) based optical phase-locking loop (PLL) was used to compensate for the slow relative phase drifts between the signal and SH-pump lights induced by temperature variations and acoustic vibrations. The input signal and the PSA output signal were compared at the receiver using three different kinds of measurements. The OSNR of the multi-carrier signal was measured by using an optical spectrum analyzer (OSA). Each sub-carrier was divided by using an optical filter with a 20 GHz spacing. The differences in the noise levels after optical-to-electric conversion for degenerate and non-degenerate signals were examined with an electrical spectrum analyzer (ESA). The bit error rate was measured with a differential receiver, which consisted of a delay interferometer (DI), a balance PD, and an error detector (ED).
3. Experimental results and discussion
First, we checked the improvement in the OSNR. The ASE from the EDFA used for the pump was effectively suppressed by the dichromatic mirrors in the two PPLN waveguides. The leakage power of the ASE into the PSA output was negligible in this experiment. This enabled a direct comparison of the OSNR for the input with that for the PSA output. Figure 3(a) shows the optical spectra of the input signal and the PSA output without the intentional addition of ASE noise. The center carrier is the degenerated signal and the pairs that are symmetrically detuned from the center carrier are non-degenerate signals. The simultaneous amplification of a modulated multi-carrier signal was achieved. The phase matching bandwidth of the PPLN waveguide for OPA was approximately 60 nm, which enabled the simultaneous amplification of a multi-carrier signal. In this experiment, the channel number was limited by the configuration of the optical comb generation. In addition, the phase matching bandwidth of the PPLN waveguide for SHG was approximately 0.2 nm at the fundamental wavelength. The pump wave was CW in this experiment, so the SHG bandwidth was also wide enough and there was no adverse effect on the PSA output. The gain of the non-degenerate signals was the same as that of the degenerate signal, namely an external gain of about 7 dB. The gain of the PSA includes the internal parametric gain of the PPLN waveguide ( + 13 dB), the insertion loss of the OPA module (‐5 dB), and the insertion loss of the tap coupler (‐1 dB) for the PLL. This result indicates that the fully correlated and equalized signal-idler pair of the coherent multi-carrier interacts in the PSA. Note that the PSA also generated an ASE, and so we intentionally added ASE noise that was sufficiently larger than the intrinsic ASE of the PSA to confirm the SNR benefit for the degraded signal.
Figure 3(b) shows optical spectra for the intentionally degraded signal with the accompanying excess noise. The intentionally added ASE level is about −40 dBm at 0.01 nm resolution which is sufficiently larger than the intrinsic ASE level of −67 dBm. The optical spectra for the PSA output exhibit a clearly higher OSNR than that at the input. The ASE gain is 6 dB lower than the PSA gain. It appears that there is a 6-dB OSNR improvement, however a 3-dB improvement results from the polarization dependence of the optical parametric amplification. In this experiment, the added ASE had both polarizations but the PPLN amplified only the TM polarization component.
For a valid comparison of the OSNR, we employed an input signal with intentional ASE only for the TM polarization. We used a CW multi-carrier signal to evaluate the gain differences between signal and ASE for each sub carrier. Figure 4(a) shows the optical spectra of the input signal and the PSA output. The simultaneous amplification of the multi-carrier signal and an OSNR improvement were observed. Figure 4(b) shows the gain for the correlated signals and for the uncorrelated ASE noise as a function of the added ASE power level. A 3-dB gain difference, which corresponds to a 3-dB OSNR improvement, was achieved for both the degenerate and non-degenerate signals.
Note that the ASE levels of the input in Fig. 3(b) and Fig. 4(a) have a wavelength-dependent slope. This originates from the wavelength dependence of the EDFA used to generate the intentional ASE. In contrast, the ASE levels of the PSA output flattened the wavelength dependence. If there is a power difference between the signal and idler, each wave has a different gain for a non-degenerate OPA process. In a non-linear medium, the idler power is first increased by the difference frequency generation (DFG) from the higher intensity signal. Once the idler power approaches the signal power, the signal is also amplified by the DFG from the idler. Then, the signal-idler power difference is reduced if the OPA gain is sufficiently high.
Next, we examined whether or not the PSA improved the intensity beat noise. The signal was significantly degraded. Therefore, the noise power was dominated by the signal-ASE intensity beat noise in this experiment. Figures 5 (a) and 5(b) show the noise spectra of the input and PSA output for degenerate and non-degenerate signals, respectively. Here, we used a CW multi-carrier signal. The input and PSA output powers were both equalized using an optical attenuator. The noise level of the PSA output for a degenerate signal is the same as the noise level of the input, as shown in Fig. 5(a). The in-phase electric field of the ASE noise, which contributes to the signal-ASE beat noise, is included at the same rate in the PSA output. This result reveals that the SNR did not change after passing through the PSA. In contrast, the noise level of the PSA output for a non-degenerate signal was lower than that of the input, as shown in Fig. 5(b). We obtained a 2.8 dB average noise reduction below the carrier frequency of 10 GHz. The in-phase electric field of the ASE noise, which contributes to the signal-ASE beat noise, is reduced by half for the PSA output. This result shows that the SNR is improved compared with that of the input signal by the reduction in the signal-ASE beat noise that is realized after the signal has been amplified by the PSA.
As shown in Fig. 6 (for an input signal OSNR of 12 dB) the intentionally added ASE noise caused a 3-dB power penalty at a BER of 10−9 for both degenerate and non-degenerate input signals. For the non-degenerate signal, the power penalty was improved more than 2.5 dB at a BER of 10−9 after the signal had passed through the PSA. Then, error-free operation was achieved with < 0.5 dB power penalty. In contrast, only a slight improvement of 1.0 dB at a BER of 10−9 was obtained for the degenerate signal after the PSA. We believe that the improvement was the result of another effect such as the phase regeneration of the PSA.
Eye diagrams of the input and PSA output for a non-degenerate signal are shown in Fig. 7 for received powers of −39 and −37 dBm, respectively. Clearer eye opening was obtained for the PSA output than for the input signals.
Figure 8 shows the results of BER measurements for severely degraded signals by means of the intentionally added ASE noise (for an input signal OSNR of 10.5 dB). The degraded OSNR of the input signal caused a BER floor with large power penalties. The BER floor for the non-degenerate signal was improved with a power penalty improvement of 4 dB at a BER of 10−7 after the PSA. In contrast, only a slight improvement of about 1.5 dB at a BER of 10−7 was obtained for the degenerate signal.
We demonstrated the simultaneous amplification of a coherent multi-carrier signal using a χ(2)-based non-degenerate PSA. A 3-dB reduction in the signal-ASE beat noise of a coherent multi-carrier pair was achieved experimentally for the first time using a χ(2)-based non-degenerate PSA in the telecommunication wavelength range. A second harmonic pumped χ(2)-based PSA enables us to compare the input and output SNRs directly. A 3-dB OSNR improvement was obtained for both degenerate and non-degenerate signals due to the gain difference between the signal and the noise. An almost 3-dB reduction in the signal-ASE beat noise in electrical domain was achieved only for the non-degenerate signal. In addition, the receiver sensitivity for a BPSK signal was clearly improved for the non-degenerate signal. This noise reduction capability will be useful for the generation and amplification of coherent multi-carrier signals for future high capacity optical transport systems.
References and links
1. T. Kobayashi, A. Sano, A. Matsuura, M. Yoshida, T. Sakano, H. Kubota, Y. Miyamoto, K. Ishihara, M. Mizoguchi, and M. Nagatani, “45.2 Tb/s C-band WDM transmission over 240 km using 538 Gb/s PDM-64QAM single carrier FDM signal with digital pilot tone,” In Proceedings of the European Conference and Exhibition on Optical Communication (ECOC 2011, Geneva, Switzerland) PDP paper Th.13.C.6.
3. A. D. Ellis and F. C. G. Gunning, “Spectral density enhancement using coherent WDM,” IEEE Photon. Technol. Lett. 17(2), 504–506 (2005). [CrossRef]
4. C. M. Caves, “Quantum limits in noise in linear amplifiers,” Phys. Rev. D Part. Fields 26(8), 1817–1839 (1982). [CrossRef]
5. J. A. Levenson, I. Abram, T. Rivera, and P. Grangier, “Reduction of quantum noise in optical parametric amplification,” J. Opt. Soc. Am. B 10(11), 2233–2238 (1993). [CrossRef]
6. M. Asobe, T. Umeki, and O. Tadanaga, “Phase sensitive amplification with noise figure below the 3 dB quantum limit using CW pumped PPLN waveguide,” Opt. Express 20(12), 13164–13172 (2012). [CrossRef] [PubMed]
7. W. Imajuku, A. Takada, and Y. Yamabayashi, “Low-noise amplification under the 3-dB noise figure in a high-gain phase-sensitive fiber amplifier,” Electron. Lett. 35(22), 1954–1955 (1999). [CrossRef]
8. R. Slavík, F. Parmigiani, J. Kakande, C. Lundström, M. Sjödin, P. A. 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. J. Richardson, “All-optical phase and amplitude regenerator for next-generation telecommunications system,” Nat. Photonics 4(10), 690–695 (2010). [CrossRef]
9. K. J. Lee, F. Parmigiani, S. Liu, J. Kakande, P. Petropoulos, K. Gallo, and D. Richardson, “Phase sensitive amplification based on quadratic cascading in a periodically poled lithium niobate waveguide,” Opt. Express 17(22), 20393–20400 (2009). [CrossRef] [PubMed]
11. R. Tang, P. Devgan, V. S. Grigoryan, and P. Kumar, “Inline frequency-non-degenerate phase-sensitive fibre parametric amplifier for fibre-optic communication,” Electron. Lett. 41(19), 1072–1074 (2005). [CrossRef]
12. J. Kakande, C. Lundström, P. A. Andrekson, Z. Tong, M. Karlsson, P. Petropoulos, F. Parmigiani, and D. J. Richardson, “Detailed characterization of a fiber-optic parametric amplifier in phase-sensitive and phase-insensitive operation,” Opt. Express 18(5), 4130–4137 (2010). [CrossRef]
13. Z 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]
14. Z. Tong, C. J. McKinstrie, C. Lundström, M. Karlsson, and P. A. Andrekson, “Noise performance of optical fiber transmission links that use non-degenerate cascaded phase-sensitive amplifiers,” Opt. Express 18(15), 15426–15439 (2010). [CrossRef] [PubMed]
15. F. Kakande, P. Petropoulos, and D. J. Richardson, “Fiber optical parametric amplification of optical combs for enhanced performance and functionality,” In Proceedings of the European Conference and Exhibition on Optical Communication (ECOC 2011, Geneva, Switzerland) paper Th.11.LeCervin.5.
16. S. Oikawa, F. Yamamoto, J. Ichikawa, S. Kurimura, and K. Kitamura, “Zero-Chirp Broadband Z-Cut Ti:LiNbO3 Optical Modulator Using Polarization Reversal and Branch Electrode,” J. Lightwave Technol. 23(9), 2756–2760 (2005). [CrossRef]
17. T. Umeki, O. Tadanaga, and M. Asobe, “Highly efficient wavelength converter using direct-bonded PPZnLN ridge waveguide,” IEEE J. Quantum Electron. 46(8), 1206–1213 (2010). [CrossRef]