Large increase of effective sensing points in Brillouin optical correlation domain analysis (BOCDA) is achieved by simultaneously applying double modulation and optical time gate based on differential measurement scheme. The noise substructure of Brillouin gain spectrum induced by the double modulation is effectively suppressed by the differential measurement, leading to 2,000 times enlargement of the measurement range. Distributed strain and temperature sensing along a 10.5 km fiber with spatial resolution of less than 1 cm is experimentally demonstrated which corresponds to over 1 million effective sensing points.
© 2015 Optical Society of America
Distributed fiber sensors based on stimulated Brillouin scattering (SBS) can provide the map of temperature or strain distribution along large civil structures or materials, so have attracted considerable attention as a powerful tool for structural health monitoring. Various schemes of distributed Brillouin sensors have been successfully developed in the domain of time, frequency, and correlation in the form of analysis and reflectometry [1–6]. Among them Brillouin optical correlation domain analysis (BOCDA) can be regarded as a quasi-distributed scheme which performs point-by-point measurement with random accessibility of sensing position . The spatial resolution of BOCDA systems can easily reach sub-cm at the cost of limited measurement range, which makes the BOCDA system particularly useful for applications requiring high spatial resolution within short distance. For instance a 2 mm spatial resolution was demonstrated with a 5 m sensing range by applying beat lock-in detection , and a 7 cm resolution with a 1 km range was realized by applying optical time gating . While the sensing distance of pulse-based time-domain Brillouin sensors is limited by the reduction of signal to noise ratio at distant positions, the limited sensing range of the BOCDA system originates from the periodic nature of the sensing point called correlation peak .
The number of effective sensing points (Neff) is defined as the ratio of sensing range to spatial resolution which can be used as a figure of merit in comparing the performance of distributed sensors. The Neff of a basic BOCDA system is determined only by the modulation amplitude (Δf) of a light source, which is generally limited to a few thousand . Optical time gate  and double modulation [9,10] schemes were introduced to enlarge the measurement range of correlation-domain Brillouin sensors, however the maximum Neff so far demonstrated was only about 15,000  limited by the depletion of the pump in the time gate and the rise of background noise in the double modulation. This figure is pretty smaller than those of other types of Brillouin sensors such as Brillouin optical time-domain analysis (~240,000)  or PRBS-based phase-correlation technique (~1,250,000) . Recently a differential measurement scheme was successfully introduced to the BOCDA where on-off switching of the phase modulation of the pump wave is additionally applied to the lock-in detection for the enhancement of the spatial resolution or the implementation of an inline structure [13,14]. In this paper we newly apply the differential measurement to the BOCDA with the double modulation and optical time-gate to improve and maximize the Neff of the system. The noise substructure in the Brillouin gain spectrum (BGS) induced by the double modulation is effectively suppressed, which consequently leads to large extension of the measurement range. The spatial resolution less than 1 cm with the sensing range longer than 10 km is experimentally demonstrated which corresponds to the Neff of more than 1 million, around 67 times improvement compared to former BOCDA results.
The measurement range (R) of a BOCDA system is determined by the distance between two adjacent correlation peaks, which are generated periodically by the sinusoidal frequency modulation of light source, given by the following equation :5]:Eq. (3), the Neff is proportional to Δf, the maximum of which is about 20 GHz in commonly used distributed feedback laser diode (DFB LD) under typical operation conditions (i.e. fm of 100 kHz to 10 MHz). Thus the Neff of a basic BOCDA system is limited to ~2,000 with ΔνB = 30 MHz for a single mode fiber.
An optical time gate scheme has been applied to enlarge the measurement range of the BOCDA system, where the pump and probe waves are synchronously gated every N period of the frequency modulation to obtain N times of range enlargement at the cost of as much longer measurement time. Although the time gate is a well-established method, the maximum Neff of the time-gated BOCDA system so far demonstrated is about 15,000 . The main difficulty in applying the time gate to high resolution and long range measurement is the depletion of the gated pump wave due to the onset of the modulation instability .
Another approach is the simultaneous modulation at two different frequencies which is called double modulation scheme [9,10]. In this method slow and fast modulations are applied together and a sharp BGS is obtained only at particular positions where correlation peaks of two modulations are overlapped. The measurement range is multiplied by the ratio of two modulation frequencies while one can keep the measurement time. However, the maximum range improvement by the double modulation so fat demonstrated has been only around 10 times . The main problem of this scheme is the accumulation of the noise substructure coming from the nonlocal SBS interaction through the fiber except the sensing position which considerably deteriorates the signal to noise ratio (SNR) under a large frequency ratio. This feature can be confirmed by the simulation results shown in Figs. 1(a)–1(c), where the BOCDA signals with single frequency modulation (a), double modulation with the frequency ratio of 10 (b) and 100 (c) are compared. For the simulation a time-domain approach is adopted  where we assume that the length of the fiber equals to the theoretical measurement range, and the Brillouin frequency (νB) is uniform along the fiber with the location of the sensing position in the middle of the fiber. Red curves correspond to the case when the νB of the sensing position is shifted + 200 MHz by strain while black curves to the case without strain. The dashed lines indicate the level of noise substructure (i.e. the maximum level of un-shifted part under strain). One can see that the rise of the noise level from 0.830 to 0.994 as the frequency ratio increases. Considering that additional intensity noises possibly appear in real measurement it may be difficult to avoid large measurement errors in the cases of Figs. 1(b) and 1(c).
In our work, we apply the differential measurement to the BOCDA system with the double modulation and optical time gate to suppress the noise substructure and maximize the Neff. The effect of the differential measurement is schematically shown in Fig. 2 where the pump wave is additionally phase modulated and this modulation is periodically turned on and off . Since the applied phase modulation effectively broadens only the sharp signal peak coming from the sensing position, one can obtain the BOCDA signal with the noise substructure suppressed by taking difference between the phase modulation on and off through a lock-in amplifier.
The modulation waveform for the BOCDA system with the double-modulation and time gate is depicted in Fig. 3. In the operation of double modulation two modulation frequencies (fm, fm / N1) are applied to a light source with modulation amplitude of Δf1 and Δf2 for each. It is notable that the spatial resolution is determined only by the parameters of the higher frequency (or primary) modulation, i.e. fm and Δf1, and the role of the secondary modulation (fm / N1 and Δf2) is to extend the measurement range by N1 times. One of key constraints in determining the modulation parameters is that the spatial resolution by the slower secondary modulation needs to be shorter than the distance between nearby correlation peaks in the faster primary modulation to prevent the interference of the signals from adjacent correlation peaks of the primary modulation and obtain the sharp BGS signal from only a single correlation peak. The sensing range is further extended by N2 times using the optical time gate with the width set to the wavelength of the secondary modulation where the gating period is an integer (i.e. N2) multiple of that of the secondary modulation. Therefore one can achieve the overall enlargement of sensing range by N1 x N2 times.
Figures 4(a) and 4(b) are the results of test measurement showing the BOCDA signal with the double modulation under ordinary lock-in detection (a) and differential measurement (b), respectively. One can observe great suppression of the noise substructure by the effect of the differential measurement. The main difficulty solved in this work is the suppression of the noise substructure in the double modulation scheme, which leads to the effective sensing points over 1 million by extending the sensing range, about 67 times larger than former BOCDA results.
The experimental setup of the BOCDA system is depicted in Fig. 5. A DFB LD with the center frequency of 1550 nm was used as a light source, to which double modulation was applied with parameters of f1 ~18 MHz, Δf1 ~4.5 GHz, f2 ~90 kHz, and Δf2 ~14 GHz, respectively. The peak to peak spectral width of the light source with the double modulation was about 37 GHz as shown in the inset ‘A’. The measurement range and spatial resolution by the primary modulation are about 5.7 m and 1.2 cm by Eqs. (1) and (2). By applying the secondary modulation f2 and the optical time gating with a ratio of 10, the measurement range was extended by 2000 times to ~11.4 km. The spatial resolution calculated from the slow modulation parameters is ~0.79 m, which is much smaller than 5.7 m, the distance between nearby correlation peak of the fast modulation. It is notable that the double modulation does not induce additional complexity to the system in the measurement. The double modulation is an amplitude sum of two sinusoidal waves which was generated as an arbitrary waveform by single arbitrary function generator. Once the double modulation waveform is generated, no additional generation process is needed. The operation of the double modulation is identical to that of ordinary single modulation considering that the sensing position is controlled by changing the sampling rate of the modulation waveform. The output was divided into two arms by a 50/50 coupler, and the optical frequency of the probe wave was ramp-swept by a single-sideband modulator (SSBM) and a microwave generator from 10.3 to 11.3 GHz. For the differential measurement 30 MHz phase modulation was applied to the pump wave at the on-off frequency of 93 kHz which was used as the reference frequency for a lock-in amplifier. It is also notable that the 30 MHz phase modulation provides an additional advantage of improved spatial resolution to ~0.6 cm, the half of the theoretical value . Therefore, the expected Neff of this system is about 1.9 million (11.4 km/0.6 cm). Both pump and probe waves were synchronously time-gated by electro-optic modulators (EOM’s) and a two-channel pulse generator at the repetition rate (fG) of ~9 kHz which was kept 10 times smaller than f2 throughout the distributed measurement. A polarization switch (PSW) was used for the probe wave to compensate for the polarization dependent gain change, and the pump and probe waves were amplified by Er-doped fiber amplifiers (EDFA’s) to ~23 dBm (peak power), respectively. For controlling the order of correlation peak, a 52 km delay fiber was used for the probe wave. The pump and probe waves were counter-propagated in a 10.5 km fiber under test (FUT), and the out probe wave was detected by a photo detector (PD) through a variable optical attenuator (VOA). Finally, the BOCDA signal was obtained by a data acquisition (DAQ) system through a LIA. The interrogation time of single position was about 20 s. Inset ‘B’ shows the structure of FUT, a 10.5 km conventional single mode fiber with an average νB of ~10.9 GHz. Two pieces of 1 cm dispersion shifted fiber (DSF) with νB of 10.5 GHz are inserted near both ends of the FUT for the test of spatial resolution, and a 1.5 cm strain test section (‘Δε – test’) and a 2 cm temperature test section (‘ΔΤ – test’) are also positioned near the rear end of the FUT.
The results of the distributed measurement are depicted in Figs. 6(a)–6(f), where the full distribution map of the BGS is plotted in Fig. 6(b) and local BGS at the DSF (red) and the SMF (black) near the front and rear ends of the FUT are shown in Figs. 6(a) and 6(c), respectively. One can clearly see that the BGS at 1 cm DSF sections are shifted by ~400 MHz compared to the BGS at SMF sections, and the SNR is large enough for detection. It is also observed that the signal amplitude at the rear end of the FUT decreases by about 35% compared to the front end, which is attributed to the propagation loss of the pump wave. The peak search result of Fig. 6(b) is plotted in Fig. 6(e) where almost uniform distribution of νB ~10.9 GHz is observed throughout the fiber. Zoomed views of the νB distribution in the vicinity of the DSF sections are depicted with 0.5 cm step in Figs. 6(d) and 6(f), respectively. The DSF pieces are detected as two or three points of νB ~10.5 GHz which confirms that the spatial resolution of our measurement is less than 1 cm under more than 10.5 km sensing range and the Neff of this system is more than 1 million.
As the second experiment we performed distributed temperature and strain measurement in the vicinity of test sections located near the end of the FUT (inset ‘B’ in Fig. 5). The Δε test section was acrylate-coated SMF bonded to a translational stage with the full length of the strain applied region of 5 cm (1.5 cm acrylate coating + 3.5 cm epoxy bonding). The ΔΤ test section was also acrylate coated SMF bonded to a Peltier element with 2 cm width. The results are depicted in Figs. 7(a)–7(c) and Figs. 7(d)–7(f) for strain and temperature, respectively. The shift of local BGS according to stain and temperature change is observed in Figs. 7(a) and 7(d), and the distribution map of νB is plotted in Figs. 7(b) and 7(e) for strain and temperature variation, respectively, showing the change of νB near the test sections. The strain and temperature sensitivity was measured to be 0.021 MHz/με and 1.06 MHz/°C as plotted in Fig. 7(c) and 7(f), respectively. The measured strain sensitivity is lower than the reported value (~0.05 MHz/με), which might be attributed to the mitigation effect of the acrylate coating. It should be noted that the νB was averaged over the test section for the plot of Fig. 7(c) and 7(f) due to large fluctuations of νB up to 10 MHz as shown in Fig. 7(b) and 7(e). The standard deviation of the measurement error was about 3 MHz. Such a large measurement error mainly comes from the position-drift of the correlation peak by the ambient temperature change due to the use of long delay fiber (~52 km) which may cause the position error of ~cm by the temperature change of 0.1 degree.
We have demonstrated the BOCDA system with huge enlargement of the measurement range using double modulation and optical time gate based on differential measurement scheme. The number of effective sensing points is confirmed to be more than 1,050,000 (10500 m / 1 cm) and is theoretically expected to reach 1,900,000 (11400 m / 0.6 cm), which, however, needs experimental confirmation. We think further improvement would be possible with the proposed scheme by proper optimization of the modulation parameters, and further research is necessary on the suppression of the position-drift due to the ambient temperature change to make this system more practical. Another weak point of this system is the huge amount of sensing time in the distributed measurement. Considering the single point interrogation time of 20 s it takes more than 231 days to access all of the 1 million sensing points, which seriously deteriorate the practicality of the system. As a possible solution to reduce the measurement time we believe the time-multiplexing technique recently proposed for the PRBS-based correlation method  may be applicable to this system. In such a case the time-multiplexing will be used as the replacement of the time-gating for the access of multiple correlation peaks, and the modulation parameters and the signal processing of the system need to be also modified to maximize the sensing points per unit time.
Finally it may be meaningful to compare the PRBS-based correlation system and the BOCDA system in terms of the Neff. The PRBS-based method has the advantage to intrinsically allow an arbitrary number of Neff, however needs several GHz-order phase modulation to apply PRBS pattern for cm-order spatial resolution. Meanwhile, the BOCDA needs only MHz-order sinusoidal modulation for realizing cm-order spatial resolution, however provides a limited Neff in a basic form. Our results have shown that the BOCDA system can provide Neff comparable to that of the PRBS-based system by proper modification.
This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIP) (No. NRF-2015R1A2A2A01007078). Kwanil Lee was supported by the Civil-Military Integration Project (CMP-13-01-KRISS) funded by National Research Council of Science & Technology (NST).
References and links
2. D. Garus, K. Krebber, F. Schliep, and T. Gogolla, “Distributed sensing technique based on Brillouin optical-fiber frequency-domain analysis,” Opt. Lett. 21(17), 1402–1404 (1996). [CrossRef] [PubMed]
3. X. Bao, M. DeMerchant, A. Brown, and T. Bremner, “Tensile and compressive strain measurement in the lab and field with the distributed Brillouin scattering sensor,” J. Lightwave Technol. 19(11), 1698–1704 (2001). [CrossRef]
4. M. N. Alahbabi, Y. T. Cho, and T. P. Newson, “150-km-range distributed temperature sensor based on coherent detection of spontaneous Brillouin backscatter and in-line Raman amplification,” J. Opt. Soc. Am. B 22(6), 1321–1324 (2005). [CrossRef]
5. K. Hotate and T. Hasegawa, “Measurement of Brillouin gain spectrum distribution along an optical fiber using a correlation based technique—proposal, experiment and simulation,” IEICE Trans. Electron. E83-C, 405–412 (2000).
7. K. Y. Song, Z. He, and K. Hotate, “Distributed strain measurement with millimeter-order spatial resolution based on Brillouin optical correlation domain analysis,” Opt. Lett. 31(17), 2526–2528 (2006). [CrossRef] [PubMed]
8. K. Hotate, H. Arai, and K. Y. Song, “Range-enlargement of simplified Brillouin optical correlation domain analysis based on a temporal gating scheme,” SICE J. Control Meas. Syst. Int. 1(4), 271–274 (2008). [CrossRef]
9. K. Hotate and M. Tanaka, “Correlation-based continuous-wave technique for optical fiber distributed strain measurement using Brillouin scattering with cm-order spatial resolution – applications to smart materials,” IEICE Trans. Electron. E84-C, 1823–1828 (2001).
10. Y. Mizuno, Z. He, and K. Hotate, “Measurement range enlargement in Brillouin optical correlation-domain reflectometry based on double-modulation scheme,” Opt. Express 18(6), 5926–5933 (2010). [CrossRef] [PubMed]
11. M. A. Soto, M. Taki, G. Bolognini, and F. Di Pasquale, “Optimization of a DPP-BOTDA sensor with 25 cm spatial resolution over 60 km standard single-mode fiber using Simplex codes and optical pre-amplification,” Opt. Express 20(7), 6860–6869 (2012). [CrossRef] [PubMed]
12. A. Denisov, M. A. Soto, and L. Thevenaz, “1,000,000 resolved points along a Brillouin distributed fiber sensor,” Proc. SPIE 9157, 9157D2 (2014).
13. J. H. Jeong, K. Lee, K. Y. Song, J. M. Jeong, and S. B. Lee, “Differential measurement scheme for Brillouin optical correlation domain analysis,” Opt. Express 20(24), 27094–27101 (2012). [CrossRef] [PubMed]
14. J. H. Jeong, K. H. Chung, S. B. Lee, K. Y. Song, J.-M. Jeong, and K. Lee, “Linearly configured BOCDA system using a differential measurement scheme,” Opt. Express 22(2), 1467–1473 (2014). [CrossRef] [PubMed]
15. M. N. Alahbabi, Y. T. Cho, T. P. Newson, P. C. Wait, and A. H. Hartog, “Influence of modulation instability on distributed optical fiber sensors based on spontaneous Brillouin scattering,” J. Opt. Soc. Am. B 21(6), 1156–1160 (2004). [CrossRef]
16. K. Y. Song, Z. He, and K. Hotate, “Effects of intensity modulation of light source on Brillouin optical correlation domain analysis,” J. Lightwave Technol. 25(5), 1238–1246 (2007). [CrossRef]
17. Y. London, Y. Antman, R. Cohen, N. Kimelfeld, N. Levanon, and A. Zadok, “High-resolution long-range distributed Brillouin analysis using dual-layer phase and amplitude coding,” Opt. Express 22(22), 27144–27158 (2014). [CrossRef] [PubMed]