We demonstrate a novel kind of tunable optical delays based on dynamic grating generated by Brillouin scattering in an optical fiber. An axial strain gradient is applied to a 15 m section of a polarization-maintaining fiber, and the Brillouin reflection grating is generated position-selectively by controlling the optical frequencies of Brillouin pump waves. Tunable time delays of up to 132 ns are achieved with an 82 ns Gaussian pulse.
©2009 Optical Society of America
Optically-controlled time delays can play important roles in optical communication and sensor applications including optical coherence tomography, all-optical buffers and phased-array antennas [1–3]. Various types of delay schemes using optical fibers have been proposed based on chirped fiber Bragg gratings , temporal gratings , wavelength conversion with dispersive media [6, 7], and slow light propagation [8–11]. Each of these approaches shows advantages and drawbacks in terms of the amount of delay, the accuracy of control, the speed of operation, the distortion of pulse, and the structural complexity . For example, the method based on the wavelength conversion can offer a large amount of fractional delays of more than 1,000 while the operation speed is limited owing to the use of tunable optical filters [6, 7]. On the contrary, the slow light-based scheme, which has been among the hot issues in recent studies, can provide a simple and high speed operation, while suffering from a practical limitation in the maximum amount of fractional delays [8–11].
In this paper, we present a novel kind of tunable optical delays based on Brillouin dynamic grating in a polarization-maintaining fiber, where the location of the reflection grating can be controlled within a 15 m strain-applied section of the fiber by tuning the optical frequencies of the pump waves. Tunable time delays of up to 132 ns are achieved using 82 ns Gaussian pulses without significant broadening or distortion. Since the amount of delay is extendable simply by increasing the length of the fiber, our scheme is expected to provide a powerful delay mechanism with the advantages of high-speed operation and large amount of delay.
When stimulated Brillouin scattering (SBS) takes place in a birefringent medium like a polarization maintaining fiber (PMF), the acoustic waves generated by optical waves in one polarization (‘writing beams’) can be used as a reflection grating for orthogonally-polarized optical waves at a different wavelength [12, 13]. This grating is called Brillouin dynamic grating (BDG) and originated from the polarization-independent nature of the longitudinal acoustic waves.
Figure 1 shows the operation principle of BDG. Brillouin pump1 and pump2 are counter-propagated through the slow axis of a PMF with their frequency difference set to the Brillouin frequency (νB) for the generation of acoustic waves. When a signal pulse with the optical frequency properly detuned from pump1 (by Δν) is propagated through the fast axis in the direction of pump1, it is reflected by the acoustic waves with the frequency down-shifted by νB. When the birefringence (Δn≡nx-ny) of the PMF is small, the frequency offset Δν is determined by a simple equation as follows :
where n and ν are the average refractive index in the fiber and the average optical frequency of the pump waves, respectively. In conventional PMF’s with Δn~0.004, Δν is around 50 GHz at the wavelength of 1550 nm. Since Δn is a function of external parameters like pressure, temperature or axial strain , it is noticeable that one can change the local value of Δν, and consequently, the position of the BDG by manipulating those external parameters. In our experiment, we apply a gradual variation of the axial strain to a PMF for inducing a similar gradient of Δν, and control the location of the BDG by tuning the optical frequencies of the pump waves.
3. Experiments and results
We constructed an experimental setup as shown in Fig. 2. A 21 m PANDA PMF (manufactured by Nufern) with a mode-field diameter of 10.5 µm and a nominal Δn~3.1×10-4 at the wavelength of 1550 nm was used as a BDG medium. The fiber was bonded to the surface of a flexible PVC rod using an epoxy, and the rod was coiled around an acryl reel with an increasing diameter from 40 cm to 50 cm as shown in the inset. This structure was intended to generate a gradual increase of the axial strain along the PMF.
A 1550 nm DFB laser diode (Pump LD) with a high-accuracy current controller was used as a light source for the Brillouin pump waves, and the output was divided by a 50/50 coupler. In one of the arms, pump1 was amplified to 23 dBm by an Er-doped fiber amplifier (EDFA) and launched through the slow axis of the PMF via a polarizer and a polarization beam splitter (PBS). In the other arm, pump2 was prepared by a sideband-generation method using a single-sideband modulator (SSBM) and a microwave generator, so that the optical frequency down-shift (Δf) from the carrier is generated by around the νB (~11.04 GHz) of the PMF. The output from the SSBM was amplified to 13 dBm by another EDFA and launched to the fiber in the opposite direction to pump1 through the slow axis via a polarizer. At first, we measured the distribution of the νB in the PMF based on BOTDA method with 2 m spatial resolution by inserting an electro-optic modulator (EOM1) for 20 ns pulse generation and a photo detector (PD1) as depicted in the dashed box. The result is shown in Fig. 3(a), where a gradual increase of the νB is clearly seen within 15 m region (the position from 4 m to 19 m) with a slope of about 2.4 MHz/m which corresponds to about 48 µε/m . Fluctuations of the νB are also observed, which seem to have come from the residual stress induced when the fiber was bonded to the rod.
For signal pulses, another 1550 nm DFB laser diode (Probe LD) and EOM2 were used to generate Gaussian pulses with the width (FWHM) of 82 ns at the repetition rate of 600 kHz, which were amplified by an EDFA with the output power of 10 dBm and launched to the PMF through the fast axis in the direction of pump1 via a polarizer and the PBS. The reflected signals are monitored by PD2 and an oscilloscope, and also by an optical spectrum analyzer (OSA). In order to roughly check the spectral region of the BDG, we slowly tuned the optical frequency of Pump LD and searched for the condition of the maximum reflection from the BDG while monitoring the OSA. The result is depicted in Fig. 3(b), where large reflection from the BDG is distinctively observed with the frequency offset Δν from the pump wave (BDG-Pump2 or Signal-Pump1) of around 52.6 GHz.
The control of the time delay of the signal was achieved by controlling the optical frequency of Pump LD with the current controller, and the time waveforms were recorded with the oscilloscope while tuning Δν from 52.4 GHz to 53.2 GHz. We performed the screen average of the traces 128 times to reduce the noise. Figure 4(a) shows examples of the measured time waveforms of the signal pulse at different Δν’s. One can clearly see the time delay of the signal pulse as Δν changes, while the pulse shape is decently maintained. As shown in Fig. 4(b) where the time delay is plotted with respect to Δν, more time delay was achieved as Δν increased. Since both νB and Δν are linearly depend on the applied strain with a positive slope , it can be concluded that the BDG was moved to a distant position as Δν was increased. The maximum delay was about 132 ns corresponding to the position-shift of 13.6 m within the PMF. According to former studies [12, 13], the bandwidth of the BDG was similar to that of Brillouin gain (30~50 MHz) and the strain dependence of Δν-shift was order of 1 MHz/µε. Assuming the gradient of the strain is linear in our case, one can estimate that the effective length of the BDG within the PMF is about 1 m considering the strain slope (48 µε/m) within the PMF. Therefore, it might be reasonable to understand the position-shift of 13.6 m as the total length of the PMF excluding the effective length of the BDG.
Figure 5(a) shows the relation between Δν and the frequency offset (Δf) between pump1 and pump2 (i.e. local νB) for maximizing the reflectance of the BDG with a result of linear fit (red line). This result explains the fact that Δf needs to be tuned to local νB since not only Δν but also νB varies within the PMF as shown in Fig. 3(a). It is also remarkable that the slope of 16.1 corresponds to the ratio of the strain-dependencies between Δν and νB in this PMF. The reflectance of the signal pulse under the optimum tuning of Δf is plotted as a function of time delay in Fig. 5(b), where the reflectance around 0.1% is maintained except 0.3% in the case of near 50 ns. This peak position corresponds to the location of a uniform strain appearing due to the irregularity near the position of 10 m in Fig. 3(a). The reflectance of 0.1% matches well with the former result  where the peak reflectance of ~4% was reported using a 30 m fiber with similar pump powers, considering the effective length (~1 m) of the BDG in our experiment. Since the BDG reflectance shows linear dependence on the power of pump2 , we think it will be possible to enhance the reflectance up to several % with current setup if an EDFA with higher output power (> 23 dBm) is used for pump2. We believe the use of specialty fibers [15, 16] will be also helpful in improving the reflectance of the BDG.
Figure 6 shows the width of the signal pulse normalized to the original one (82 ns) as a function of time delay. The maximum broadening was about 30%, which seems to originate from the irregularity of the strain gradient within the effective length of the BDG. Additionally, we think further research is needed on the relation between the spectral bandwidth and the effective length of the BDG for optimizing the gradient slope of the applied strain.
Considering the inherent bandwidth (30~50 MHz) of Brillouin scattering, the minimum duration of the available signal pulse is approximately 20~30 ns. The adoption of 82 ns pulses in our experiment was to suppress significant distortion we observed in narrower pulses, which seems to originate from the frequency chirp induced by the intrinsic irregularity of the local birefringence of the PMF. We believe the use of broadband Brillouin pumps can be a possible solution to the bandwidth issue .
We have proposed and demonstrated a novel kind of tunable optical delays based on Brillouin dynamic grating in a polarization maintaining fiber. Time delays of up to 132 ns were achieved using 82 ns Gaussian pulses, and the amount of the delay was tunable by controlling the optical frequencies of Brillouin pump waves. Currently, the maximum achievable delay is limited simply by the length of the strain-applied fiber without other fundamental limiting factors such as optical powers or pulse broadening. Therefore, we believe our new scheme has a potential to provide a practical solution to the applications that require both the high speed operation and the large delays.
References and links
1. D. Huang, E. A. Swanson, C. P. Lin, J. S. Schuman, W. G. Stinson, W. Chang, M. R. Hee, T. Flotte, K. Gregory, C. A. Puliafito, and J. G. Fujimoto, “Optical coherence tomography,” Science 254, 1178–1181 (1991). [CrossRef] [PubMed]
2. R. Ramaswami and K. N. Sivarajan, Optical networks: a practical perspective (Morgan Kaufmann, San Francisco, CA, 2002) 2nd Ed., Chap. 12.
3. 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]
4. 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]
5. K. L. Hall, D. T. Moriarty, H. Hakami, F. Hakami, B. S. Robinson, and K. A. Rauschenbach, “An ultrafast variable optical delay technique,” IEEE Photon. Technol. Lett. 12, 208–210 (2000). [CrossRef]
6. 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]
8. R. W. Boyd and D. J. Gauthier “‘Slow’ and ‘Fast’ Light,” Ch. 6 in Progress in Optics 43, E. Wolf, Ed. (Elsevier, Amsterdam, 2002), 497–530. [CrossRef]
9. K. Y. Song, M. G. Herráez, and L. Thévenaz, “Observation of pulse delaying and advancement in optical fibers using stimulated Brillouin scattering,” Opt. Express 13, 82–88 (2005). [CrossRef] [PubMed]
11. D. Dahan and G. Eisenstein, “Tunable all optical delay via slow and fast light propagation in a Raman assisted fiber optical parametric amplifier: a route to all optical buffering,” Opt. Express 13, 6234–6249 (2005). [CrossRef] [PubMed]
12. K. Y. Song, W. Zou, Z. He, and K. Hotate, “All-optical dynamic grating generation based on Brillouin scattering in polarization-maintaining fiber,” Opt. Lett. 33, 926–929 (2008). [CrossRef] [PubMed]
13. W. Zou, Z. He, and K. Hotate, “Complete discrimination of strain and temperature using Brillouin frequency shift and birefringence in a polarization-maintaining fiber,” Opt. Express 17, 1248–1255 (2009. [CrossRef] [PubMed]
14. M. Nikles, L. Thevenaz, and P. A. Robert, “Brillouin gain spectrum characterization in single mode optical fibers,” J. Lightwave Technol. 15, 1842–1851 (1997). [CrossRef]
17. M. G. Herráez, K. Y. Song, and L. Thévenaz, “Arbitrary-bandwidth Brillouin slow light in optical fibers,” Opt. Express 14, 1395–1400 (2006). [CrossRef]