Open Access
Add to BibSonomyAbstract
We propose and demonstrate error-free conversion of a 40 Gbit/s optical time division multiplexed signal to 4 × 10 Gbit/s wavelength division multiplexed channels based on cascaded second harmonic and difference frequency generation in a periodically poled lithium niobate waveguide. The technique relies on the generation of spectrally (and temporally) flat linearly chirped pulses which are then optically switched with short data pulses in the nonlinear waveguide. Error-free operation was obtained for all channels with a power penalty below 2dB.
©2010 Optical Society of America
1. Introduction
All-optical signal processing has the potential to play a critical role in future high-speed and large-capacity optical networks in order to reduce both power demands and latency related to optical to electrical conversion and electronic processing [1]. Current optical networks are based on either time division multiplexing (TDM) or wavelength division multiplexing (WDM), which are individually mature technologies. However, methods for the conversion of TDM signals to mixed TDM-WDM format will be required in future state-of-the-art digital communication systems in order to combine WDM network topologies with ultra-high speed optical TDM (OTDM) networks and to facilitate signal grooming at the edge of the network [2]. Various approaches for such format conversion have been previously reported including: cross-gain compression in a semiconductor optical amplifier [3]; cross-phase modulation in a nonlinear optical loop mirror [4]; self-phase modulation (SPM) followed either by optical time gating [5] or four-wave mixing [6] in highly nonlinear fibers. Other applications for TDM to WDM format conversion, such as OTDM add-drop multiplexing [2, 4] and packet compression /expansion [7], have also been demonstrated previously.
In recent years, the use of cascaded second-order nonlinear processes in periodically poled lithium niobate (PPLN) waveguides has attracted considerable interest as a promising route to realize all-optical signal processing. The technology provides for high nonlinear coefficients, an ultra-fast optical response, bit rate and modulation format transparency, no spontaneous emission noise, low cross talk, and no intrinsic frequency chirp [8]. Cascaded second-harmonic and difference-frequency generation (cSHG/DFG) and cascaded sum- and difference-frequency generation (cSFG/DFG) have both been exploited in various all-optical signal processing applications such as all-optical wavelength conversion [9, 10], format conversion [11], logic gates [12], tunable optical time delay generation [13], and phase sensitive amplification [14].
In this paper, we propose a novel scheme to achieve OTDM to mixed TDM-WDM format conversion using the cSHG/DFG process in a fiberized PPLN waveguide [15]. The technique relies on nonlinear switching method using a linearly chirped rectangular pulse [16]. PPLN devices offer a number of attractive features relative to implementations relying on nonlinear fibers, including the prospect of far more compact devices and significantly reduced sensitivity to both external environmental perturbations and parasitic nonlinear effects, such as stimulated Brillouin scattering (SBS), the mitigation of which would either impose serious performance limitations or add significant complexity to the system.
2. OTDM to WDM format conversion via χ(2) cascading
Figure 1 illustrates our OTDM to mixed TDM-WDM conversion scheme based on cSHG/ DFG in a PPLN waveguide. For a more quantitative evaluation of the device performance, we also plot the results of numerical simulations in the same figure [Fig. 1(a) – (d)], which correspond to the system experiments we performed with the available PPLN waveguide device (these will be described in Section 3 below).
2.1 Principles of operation
The PPLN waveguide operates with two inputs:
- ❖ An OTDM data signal [Fig. 1(b)], at the aggregate bit rate of N (= 4) temporally interleaved tributary channels, having a bit period Tp and optical frequency ωp which is within the acceptance bandwidth of the PPLN.
- ❖ A train of synchronized linearly chirped rectangular pulses [Fig. 1(a)], with a duration spanning across the N tributary channels (Δτcp ~ NTp) and a repetition rate equal to that of the individual TDM tributaries (1/Tp).
- ❖ SHG from ωp, which creates an up-converted OTDM data stream at 2ωp;
- ❖ DFG between the 2ωp signal and the input rectangular pulse train, which makes each tributary channel interact with a different portion of the linearly chirped signal spectrum, thereby acquiring a different frequency shift as indicated in Fig. 1 (ω'k = 2ωp – ωk).
2.2 Theoretical model
The plots in Fig. 1 illustrate in more detail the expected results for the uniform PPLN waveguide we used in the experiments, characterized by a length L = 30 mm, a normalized SHG conversion efficiency ηnor = 60% W−1cm−2, and 80% input coupling losses. In the model we used the dispersion parameters of bulk LiNbO3 [17]. For typical fabrication conditions of buried waveguides in this wavelength range [18], the main effect of waveguide dispersion is a shift of 2~3 μm of the optimum period with respect to the bulk, while group velocity mismatch and group velocity dispersion are not significantly affected.
The results shown in Fig. 1 were calculated for the following inputs [electric field envelopes A(z = 0, t)]:
- ❖ OTDM signal at the SHG phase matching wavelength λp, with average power Pp and a Gaussian field profile [Fig. 1(b)] of the form: Ain(0, t) = Ap⋅exp{–[(t–nTp)/τp]2,} with n = ±1/2, ±3/2, Tp = 25 ps, τp = 6 ps, corresponding to a full width at half maximum (FWHM) pulse duration of 7 ps (power) [τ = τp (2·ln2)1/2], Ap = [PpTp/(τp√2π)]1/2;
- ❖ Chirped pulse with a spectrum spanning Δλcp = 6 nm around λcp, average power Pcp and a field profile of the form: Acp(0, t) = Acp0⋅exp{–(ln2)/2[2(t–δt)/Δτcp]18}⋅exp[iCp(t–δt) 2], with Δτcp = 90 ps, Acp = (PcpTp/Δτcp)1/2, a chirp Ccp = 0.023 ps−2, and a time delay δt with respect to the OTDM input (δt = 2 ps for Fig. 1, to optimize the response with respect to SHG walk-off).
- ❖ the nonlinear coupling coefficients: Γ1 = ηnor 1/2, Γ2 = (λp/λcp)Γ1, Γ3 = (λp/λout)Γ1, Γ4 = 2Γ1;
- ❖ the mismatches: Δβ1 = 2π [2 (nsh–np)/λp–1/Λ] and Δβ2 = 2π [nsh/λsh–ncp/λcp–ncp/λc–1/Λ];
- ❖ the group velocity terms: δν1 = νin−1, δν2 = νcp−1, δν3 = νout−1, δν4 = νsh−1,
Figure 1(c) plots the calculated amplitude and chirp of the generated output OTDM-WDM pulses, for λp = 1546 nm, Pp = 16 dBm, λcp = 1551 nm, Pcp = 14 dBm, and λout~2λp–λcp = 1541 nm, simulating the experiments described in the following section. In Fig. 1(d), we show also the expected OTDM pump throughput and SH output (solid and dotted lines, respectively). No distortion is apparent in the format converter output [Fig. 1(c)], despite the presence of SHG walk-off [Fig. 1(d)]. This confirms the capability of the cSHG/DFG scheme to work with pulses (7-ps FWHM in the simulations) shorter than the SHG walk-off limit (~10 ps for our device).
3. Experiment and discussion
Figure 2(a) shows the corresponding experimental setup used to convert the OTDM signal to a mixed TDM-WDM signal. A 30-mm-long fiber pigtailed PPLN waveguide (HC Photonics Corp.) was used for the cSHG/DFG process, and its phase matching wavelength for SHG was 1546 nm at 50°C. A 10 GHz, 1.5 ps mode-locked erbium glass oscillator (ERGO) was first split into two separate paths using a 3-dB coupler to generate the data signal and the linearly-chirped rectangular pulses, respectively. The pulses in the data path were modulated by a 231 – 1 pseudorandom bit sequence (PRBS) using a lithium niobate modulator and then filtered using a 0.5 nm-bandpass filter (yielding a temporal width of 7 ps) to match the PPLN SHG acceptance bandwidth, and then multiplexed up to 40Gbit/s to form a ~ 33% duty cycle return-to-zero on-off-keyed (RZ-OOK) signal as shown in Fig. 2(b). In order to generate the 10 GHz-linearly chirped rectangular-like pulses, the initial pulses were amplified to 21 dBm and launched into a 490-m long highly nonlinear fiber (HNLF) with a nonlinear coefficient of ~ 20 /W/km, a dispersion of –0.64 ps/nm/km at 1550 nm, a dispersion slope of +0.030 ps/nm2/km and an attenuation of 0.49 dB/km. The generated SPM spectral bandwidth was 25 nm, and the wavelength range between 1548 nm and 1554 nm was subsequently selected using a fiber Bragg grating (FBG) filter with a ~ 40dB out of band extinction ratio. These pulses were stretched in time by propagation through 140 m of dispersion compensating fiber (DCF), resulting in a rectangular-like envelope of ~ 85 ps FWHM as shown in Fig. 2(c). The pulses were observed to have sharp trailing and leading edges and a good flat-top section.
Fig. 2 (a) Experimental setup used to convert the OTDM to mixed TDM-WDM signal. MOD: modulator, EDFA: erbium-doped fiber amplifier, PC: polarization controller, DCA: digital communication analyzer (detection bandwidth of 32GHz). (b) Eye diagram of the 40-Gbit/s data signal. (c) Temporal trace of linearly-chirped rectangular pulse (for illustrative purposes, when taking this measurement the repetition rate was gated down to 5 GHz). (d) Spectral trace after PPLN. (e) Eye diagram and spectral trace of converted mixed TDM-WDM signal.
Figure 3(a) shows the spectrogram of the linearly chirped pulses measured using an electro-optic modulation-based linear frequency resolved optical gating (FROG) [19]. The measured chirp rate parameter was + 0.023 ps−2. A variable delay line was used to adjust the relative delay between the OTDM and linearly-chirped rectangular pulses. The two signals were then combined with a 3-dB coupler and launched into the PPLN waveguide. As discussed previously, the 40 Gbit/s data signal generated the second harmonic (at 773 nm) via the SHG process in the PPLN, which then interacted with the linearly chirped pulses via the DFG process to produce the mixed TDM-WDM-format signal as a replica of the original data, as shown in Fig. 1. The measured output spectrum of the PPLN waveguide is shown in Fig. 2(d). The 3-dB bandwidth of each of the four converted WDM channels and the spacing between them were 0.57 nm and 1.42 nm, respectively and their optical signal to noise ratio was more than 22dB.
Fig. 3 Measured FROG traces for (a) the linearly chirped pulse and (b) the converted mixed TDM-WDM signals.
A filter, tunable both in bandwidth and central wavelength (Alnair Labs.), was used after the PPLN device to extract either the full converted mixed TDM-WDM signal or one of the four WDM channels separately, and the corresponding eye-diagrams and spectral traces are shown in Fig. 2(e) and Fig. 4 .
Fig. 4 (a) Filtered and amplified spectra and corresponding eye diagrams of each switched tributary channel. (b) BER curves for each tributary channel and the back-to-back signal.
As can be seen, all of the four converted channels show a clear open eye-diagram. The individual pulses were also characterized using the FROG technique and similar envelopes and pulse widths (8 ~9 ps) to the original pulses were obtained, suggesting that the OTDM to WDM conversion has not significantly affected the quality of the signal (the slight increase in the signal pulse width is attributed to the choice of a tight filter bandwidth at the system output). The measured spectrogram of the converted mixed TDM-WDM output is shown in Fig. 3(b), where it can be appreciated that it has an opposite slope to that of the original linearly chirped signal [Fig. 3(a)]. This is because the higher frequency components of the linearly chirped pulse interact first with the input OTDM pulses which are then converted to lower frequency components in the mixed TDM/WDM output as sketched in Fig. 1.
We also assessed the performance of the conversion system through bit-error rate (BER) measurements, properly filtering and amplifying each WDM channel in the process. Error-free operation (BER=10−9) was achieved for all four WDM channels and the power penalty of the four channels was between 1.5dB and 2dB as compared to the back-to-back measurements at (10 Gbit/s).
4. Conclusion
We have successfully demonstrated the conversion of a 40 Gbit/s OTDM signal to 4 × 10 Gbit/s WDM channels using cSHG/DFG in a fully fiberized 30-mm-long PPLN waveguide. This process generates a spectral representation of data packets, which can then be further processed either temporally or spectrally. Error-free operation was obtained for all channels with a power penalty below 2dB.
Acknowledgements
The research leading to these results has received funding from the UK EPSRC under grant agreement EP/F032218/1. Katia Gallo gratefully acknowledges support from the EU under grant agreement PIEF-GA-2009-234798.
References and links
1. A. E. Willner, “All-optical signal processing in next-generation communication systems,” presented at the Conference on Optical Fiber Communication (OFC), San Diego, CA, 24–28 Feb. 2008.
2. K. Uchiyama and T. Morioka, “All-optical time-division demultiplexing experiment with simultaneous output of all constituent channels from 100Gbit/s OTDM signal,” Electron. Lett. 37(10), 642–643 (2001). [CrossRef]
3. D. Norte and A. E. Willner, “All-optical data format conversions and reconversions between the wavelength and time domains for dynamically reconfigurable WDM networks,” J. Lightwave Technol. 14(6), 1170–1182 (1996). [CrossRef]
4. P. J. Almeida, P. Petropoulos, F. Parmigiani, M. Ibsen, and D. Richardson, “OTDM add-drop multiplexer based on time-frequency signal processing,” J. Lightwave Technol. 24(7), 2720–2732 (2006). [CrossRef]
5. H. Sotobayashi, W. Chujo, and K.-I. Kitayama, “Photonic gateway: TDM-to-WDM-to-TDM conversion and reconversion at 40 Gbit/s (4 channels × 10 Gbits/s),” J. Opt. Soc. Am. B 19(11), 2810–2816 (2002). [CrossRef]
6. H. Sotobayashi, W. Chujo, and T. Ozeki, “80Gbit/s simultaneous photonic demultiplexing based on OTDM-to-WDM conversion by four wave mixing with supercontinuum light source,” Electron. Lett. 37(10), 640–642 (2001). [CrossRef]
7. P. J. Almeida, P. Petropoulos, B. C. Thomsen, M. Ibsen, and D. Richardson, “All-optical packet compression based on time-to- wavelength conversion,” IEEE Photon. Technol. Lett. 16(7), 1688–1690 (2004). [CrossRef]
8. C. Langrock, S. Kumar, J. E. McGeehan, A. E. Willner, and M. M. Fejer, “All-optical signal processing using χ(2) nonlinearities in guided-wave devices,” J. Lightwave Technol. 24(7), 2579–2592 (2006). [CrossRef]
9. K. Gallo, G. Assanto, and G. I. Stegeman, “Efficient wavelength shifting over the erbium amplifier bandwidth via cascaded second order processes in lithium niobate waveguides,” Appl. Phys. Lett. 71(8), 1020–1022 (1997). [CrossRef]
10. M. H. Chou, I. Brener, M. M. Fejer, E. E. Chaban, and S. B. Christman, “1.5-μm-band wavelength conversion based on cascaded second-order nonlinearity in LiNbO3 waveguides,” IEEE Photon. Technol. Lett. 11(6), 653–655 (1999). [CrossRef]
11. J. Wang, J. Sun, X. Zhang, D. Huang, and M. M. Fejer, “All-optical format conversions using periodically poled lithium niobate waveguides,” IEEE J. Quantum Electron. 45(2), 195–205 (2009). [CrossRef]
12. J. E. McGeehan, M. Giltrelli, and A. E. Willner, “All-optical digital 3-input AND gate using sum- and difference-frequency generation in a PPLN waveguide,” Electron. Lett. 43(7), 409–410 (2007). [CrossRef]
13. Y. Wang, C. Yu, L. Yan, A. E. Willner, R. Roussev, C. Langrock, M. M. Fejer, J. E. Sharping, and A. L. Gaeta, “44-ns continuously tunable dispersionless optical delay element using a PPLN waveguide with two-pump configuration, DCF, and a dispersion compensator,” IEEE Photon. Technol. Lett. 19(11), 861–863 (2007). [CrossRef]
14. 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]
15. K. Gallo, J. Prawiharjo, F. Parmigiani, P. Almeida, P. Petropoulos, and D. Richardson, “Processing ultrafast optical signals in broadband telecom systems by means of cascaded quadratic nonlinearities,” presented at the 8th International Conference on Transparent Optical Networks (ICTON), Nottingham, UK, 18–22 Jun. 2006.
16. T. Morioka, S. Kawanishi, H. Takara, and M. Saruwatari, “Multiple-output, 100 Gbit/s all-optical demultiplexer based on multichannel four-wave mixing pumped by a linearly-chirped square pulse,” Electron. Lett. 30(23), 1959–1960 (1994). [CrossRef]
17. D. H. Jundt, “Temperature-dependent Sellmeier equation for the index of refraction, ne, in congruent lithium niobate,” Opt. Lett. 22(20), 1553–1555 (1997). [CrossRef]
18. K. Gallo, A. Pasquazi, S. Stivala, and G. Assanto, “Parametric solitons in two-dimensional lattices of purely nonlinear origin,” Phys. Rev. Lett. 100(5), 053901 (2008). [CrossRef] [PubMed]
19. K. T. Vu, A. Malinowski, M. A. F. Roelens, M. Ibsen, P. Petropoulos, and D. J. Richardson, “Full characterization of low-power picosecond pulses from a gain-switched diode laser using electro-optic modulation-based linear FROG,” IEEE Photon. Technol. Lett. 20(7), 505–507 (2008). [CrossRef]
