We demonstrate all-optically tunable delays in optical fiber via a dispersive stage and two stages of nonlinear spectral broadening and filtering. With this scheme, we achieve continuously tunable delays of 3.5-ps pulses and advancements over a total range of more than 1200 pulsewidths. Our technique is applicable to a wide range of pulse durations and delays.
© 2006 Optical Society of America
Communication networks require components that have the capability of buffering or delaying information. In ultra-high speed communications, where information is encoded in pulses, optical/electronic conversion of information is a bottleneck for increasing the data transmission rate. Thus, it is desirable to have all-optical components for buffering and delaying signal pulses. Development of tunable all-optical delays is important for applications other than telecommunication, including optical coherence tomography , optical control of phased array antennas for radio frequency communication , and optical sampling, and pattern correlation .
One approach for demonstrating tunable all-optical delays, which has attracted significant interest, is the use of slow light based on laser-induced resonances to reduce the group velocity . There have been many demonstrations of slow light using different techniques such as electromagnetically-induced transparencies [5, 6], coherent population oscillations [7–9], and stimulated scattering [10–13]. However, the maximum relative delay, which is the total delay divided by the input pulse duration, that has been generated using a slow-light scheme has been limited to approximately a few pulsewidths.
An alternative approach involves wavelength conversion, where the central wavelength of the pulse is shifted, and then sent into a medium with large group-velocity dispersion (GVD) [14–17], and all-optical delays based on this technique were recently demonstrated [18–20]. Sharping, et al.,  used four-wave mixing (FWM) for wavelength conversion and achieved 80 pulsewidths of delay. In this scheme, the wavelength and bandwidth of the signal pulse were preserved after the delay system, which allows for a much larger range of delays while maintaining the signal pulse information.
In this paper, we present a novel wavelength conversion and dispersion technique for all-optical delays which offers a significant reduction in complexity as compared to the previous technique involving FWM. For wavelength conversion, we use the Mamyshev regenerator [21–23], which involves spectral broadening via self-phase modulation (SPM), and wavelength filtering. Using this delay scheme, we demonstrate tunable delays of more than 4.2 ns for a 3.5-ps input pulse, which corresponds to 1200 pulsewidths of delay. Our system drastically increases the range of achievable delays, while maintaining the original wavelength and bandwidth of the pulses and represents a significant step in the development of optical signal processing devices.
The pulse-delay generator consists of three stages: wavelength conversion, dispersive delay, and wavelength reconversion (Fig. 1). The wavelength conversion is done in two steps. First, the pulse is sent through a length of highly nonlinear fiber (HNLF) where the pulse undergoes SPM. The broadened spectrum is then sent through an optical bandpass filter, where the desired wavelength window is selected. After the wavelength conversion stage, the pulse is sent through a dispersive fiber, where the delay is generated. The amount of delay (advancement) induced is proportional to the product of the GVD of the dispersive fiber and the wavelength shift. After the dispersive fiber, the pulse is sent through another length of HNLF where the spectrum is again broadened and filtered at the original wavelength to achieve wavelength reconversion so that the output pulse wavelength is the same as the input pulse wavelength.
In the first and third stages of the pulse-delay generator, we take advantage of the spectral broadening that accompanies SPM. For an unchirped Gaussian pulse, the spectral pulse bandwidth after broadening can be estimated by ΔωSPM~Δω0γP0Leff, where Δω 0 is the initial spectral width of the signal pulse, γ is the nonlinear coefficient, P0 is the peak power of the signal pulse and L eff is the effective length of the fiber [21, 24]. By changing the signal power, it is possible to control the amount of spectral broadening.
In Fig. 2, we illustrate an SPM-broadened spectrum for a 3.5-ps transform-limited signal pulse. The signal wavelength was tuned to the regime of the fiber with small normal dispersion in order to achieve symmetric spectral broadening and to avoid Raman self-frequency shifting.
The experimental setup for the pulse-delay generator is shown in Fig. 3. The output of an optical parametric oscillator is filtered using a 1-nm bandpass filter to generate 3.5-ps signal pulses with a repetition rate of 75 MHz and a center wavelength of 1535 nm. The average signal power is approximately 5 mW (18 W peak power). The signal pulses are sent through 1-km of highly nonlinear fiber (HNLF), where the pulse spectrum is broadened by SPM. We use a low-power Erbium-doped fiber amplifier (EDFA) to amplify the broadened spectrum. A tunable bandpass filter (FWHM=0.5 nm) is then used to select the desired wavelength (Fig. 4). The wavelength shifted signal propagates through a dispersion compensating module (DCM) with a group-velocity dispersion of -342 ps/nm (4 km length). After the dispersive delay, the signal is amplified using a 1-W EDFA and sent into a 2-km spool of HNLF, where the pulse spectrum is once again broadened through SPM. The broadened spectrum is then filtered using a 1-nm bandpass filter centered at 1535 nm to return to the original signal wavelength. The center wavelength of the tunable bandpass filter is varied from 1533 nm to 1546 nm. The temporal data was taken using a 12.5 GHz detector and a sampling oscilloscope. The zero-dispersion wavelength of the first HNLF is 1551±2 nm with a slope of 0.04 ps/(nm2·km), and the nonlinear coefficient for the HNLF is γ=11 (W·km)-1. The second HNLF has a zero-GVD wavelength 1555 nm with a slope of 0.018 ps/(nm2·km) and a nonlinear coefficient of γ=10 (W·km)-1.
In Fig. 5(a), we show the measured signal delays/advancements through the system. Each of the different traces shown represents a 1-nm change in the tunable bandpass filter center wavelength. The total tuning range of the delay line is 4.2 ns. Figure 5(b) shows temporal pulse delay/advancement as a function of the center wavelength of the tunable bandpass filter, and as expected we observe a linear relationship.
Pulse broadening does occur due to the dispersion of the DCM, and the output pulse duration from the delay generator is approximately 350 ps. In addition, the temporal broadening is more significant as the tunable bandpass filter is tuned closer to the center wavelength of the signal, which is due to the spectral modulations that arise due to SPM. Therefore, post-compensation using a fixed amount of dispersion of the opposite sign will not be sufficient to compensate for this broadening. However, this broadening can be minimized by optimizing the bandwidth of the tunable bandpass filter before the DCM. The DCM used in the experiment has a GVD of -342 ps/nm. Thus, assuming a transform-limited Gaussian pulse, a pulse with a bandwidth of 0.1 nm, which corresponds to a 34-ps pulse, will have minimal broadening through the DCM. Due to the minimum broadening that occurs, this selection of the bandwidth allows for data rates of greater than 10 Gb/s. Any additional broadening can be post-compensated by adding an appropriate span of standard communication fiber after the pulse-delay generator. The length of the DCM and the bandwidth of the tunable bandpass filter must be carefully optimized for a desired data rate.
Figure 6 shows the pulse spectrum of the input and output of the pulse-delay generator. The central wavelength and the bandwidth of the output pulse are the same as those of the input pulse because of the choice of the fixed bandpass filter centered at 1535 nm after the second HNLF. By choosing the correct bandwidth for the bandpass filter, the output pulse bandwidth can be made to match the input pulse bandwidth. In this particular experiment, the fixed bandpass filter at the output was chosen to exactly match the fixed bandpass filter used to generate the input. The accuracy of the center wavelength and bandwidth of the output depend on that of the bandpass filter used at the output. Background extinction depends on the extinction ratio of the bandpass filter.
The reconfiguration time of this delay generator is limited by how fast the center wavelength of the tunable bandpass filter can be changed. In our current scheme in which the wavelength was tuned manually, the reconfiguration time is long. Recently however, an all-fiber tunable bandpass filter has been demonstrated which allows for a reconfiguration time of 90 µs . Utilizing such a scheme will allow for rapid tuning of the delay.
This delay scheme has the ability to generate even larger delays by increasing the signal power to induce larger spectral broadening due to SPM or simply by using a DCM that has a larger group-velocity dispersion. In addition, longer or shorter signal pulses can be used by scaling the input powers accordingly to generate the SPM-induced broadening. The scheme can be made more compact by using different media for the large nonlinearity and dispersion. Lastly, since our technique relies on the Mamyshev regenerator, it can support RZ data formats.
We have demonstrated a novel tunable all-optical delay scheme based on wavelength conversion, dispersion, and reconversion. We show that the temporal position of a 3.5 ps pulse can be tuned continuously over 4 ns, which corresponds to a fractional delay of 1200. The advantage of using the Mamyshev regenerator configuration for the wavelength conversion is that the scheme helps reduce noise that arises in data transmission. In principle, the scheme also makes it possible to choose the appropriate bandpass filter for the output such that the center wavelength and the pulse bandwidth match those of the input signal. The simplicity of the setup and the use of “off-the-shelf” components allow for easy integration into communication systems and other applications.
We gratefully acknowledge D. J. Gauthier for providing the 2-km length of HNLF used in the experiment, I. H. Agha, M. A. Foster, and J. van Howe for useful discussions, and financial support from the Center for Nanoscale Systems, supported by the NSF under grant No. EEC-0117770, and the DARPA DSO Slow-Light Program.
References and links
1. E. Choi, J. Na, S. Y. Ryu, G. Mudhana, and B. H. Lee, “All-fiber variable optical delay line for applications in optical coherence tomography: feasibility study for a novel delay line,” Opt. Express 13, 1334–1345 (2005). [CrossRef] [PubMed]
2. J. L. Corral, J. Marti, J. M. Fuster, and R. I. Laming, “True time-delay scheme for feeding optically controlled phased-array antennas using chirped-fiber gratings,” IEEE Photon. Technol. Lett. 9, 1529–1531 (1997). [CrossRef]
3. R. Ramaswami and K. N. Sivarajan, “Optical networks: a practical perspective,” (Morgan Kaufmann, San Francisco, CA, 2002) 2nd Ed., Chap. 12.
4. R. W. Boyd and D. J. Gauthier, ‘Slow’ and ‘fast’ light, Progress in Optics43, E. Wolf, ed., (Elsevier, Amsterdam, 2002), Chap. 6, p. 497–530. [CrossRef]
5. L. V. Hau, S. E. Harris, Z. Dutton, and C. H. Behroozi, “Light speed reduction to 17 metres per second in an ultracold atomic gas,” Nature 397, 594–598 (1999). [CrossRef]
6. M. M. Kash, V. A. Sautenkov, A. S. Zibrov, L. Hollberg, G. R. Welch, M. D. Lukin, Y. Rostovtsev, E. S. Fry, and M. O. Scully, “Ultraslow group velocity and enhanced nonlinear optical effects in a coherently driven hot atomic gas,” Phys. Rev. Lett. 82, 5229–5232 (1999). [CrossRef]
8. P. C. Ku, F. Sedgwick, C. J. Chang-Hasnain, P. Palinginis, T. Li, H. Wang, S. W. Chang, and S. L. Chuang, “Slow light in semiconductor quantum wells,” Opt. Lett. 29, 2291–2293 (2004). [CrossRef] [PubMed]
10. Y. Okawachi, M. S. Bigelow, J. E. Sharping, Z. Zhu, A. Schweinsberg, D. J. Gauthier, R. W. Boyd, and A. L. Gaeta, “Tunable all-optical delays via Brillouin slow light in an optical fiber,” Phys. Rev. Lett. 94, 153902 (2005). [CrossRef] [PubMed]
11. K. Y. Song, M. G. Herráez, and L. Thvenaz, “Observation of pulse delaying and advancement in optical fibers using stimulated Brillouin scattering,” Opt. Express 13, 82–88 (2005). [CrossRef] [PubMed]
15. L. Zucchelli, M. Burzio, and P. Gambini, “New solutions for optical packet delineation and synchronization in optical packet switched networks,” in Proc. ECOC ’96, Oslo, Norway, 3, 301–304 (1996).
16. K. Shimizu, G. Kalogerakis, K. Wong, M. Marhic, and L. Kazovsky, “Timing jitter and amplitude noise reduction by a chirped pulsed-pump fiber OPA,” in Proc. OFC ’03, Anaheim, USA , 1, 197–198 (2003).
18. J. E. Sharping, Y. Okawachi, J. van Howe, C. Xu, Y. Wang, A. E. Willner, and A. L. Gaeta, “All-optical, wavelength and bandwidth preserving, pulse delay based on parametric wavelength conversion and dispersion,” Opt. Express 13, 7872–7877 (2005). [CrossRef] [PubMed]
19. Y. Wang, C. Yu, L. Yan, A. E. Willner, R. Roussev, C. Langrock, and M. M. Fejer, “Continuously-tunable dispersionless 44-ns optical delay element using a two-pump PPLN, DCF, and a dispersion compensator,” in Proc. ECOC ’05, Glasgow, Scotland , 4, 793–794 (2005).
20. S. Oda and A. Maruta, “All-optical tunable delay line based on soliton self-frequency shift and filtering broadened spectrum due to self-phase modulation,” Optics Express 14, 7895–7902 (2006). [CrossRef] [PubMed]
21. P. V. Mamyshev, “All-optical data regeneration based on self-phase modulation effect,” in Proc. ECOC ‘98, Madrid, Spain , 1, 475–476 (1998).
22. V. G. Ta’eed, M. Shokooh-Saremi, L. Fu, D. J. Moss, M. Rochette, I. C. M. Littler, B. J. Eggleton, Y. Ruan, and B. Luther-Davies, “Integrated all-optical pulse regenerator in chalcogenide waveguides,” Opt. Lett. 30, 2900–2902 (2005). [CrossRef] [PubMed]
23. J. T. Mok, J. L. Blows, and B. J. Eggleton, “Investigation of group delay ripple distorted signals transmitted through all-optical 2R regenerators,” Opt. Express 12, 4411–4422 (2004). [CrossRef] [PubMed]
24. G. P. Agrawal, Nonlinear fiber optics, 3rd Ed. (Academic Press, Inc., 2001).
25. P. Z. Dashti, C. S. Kim, Q. Li, and H. P. Lee, “Demonstration of a novel all-fiber bandpass acousto-optic tunable filter,” in Proc. OFC ’05, Anaheim, USA, 6, OFC2 (2005).