Open Access
Add to BibSonomy
Abstract
Brillouin optical time-domain analysis (BOTDA) using distributed Brillouin amplification (DBA) only requires a milliwatt-level pump to achieve a sensing range beyond 100 km, which provides a powerful tool for temperature/strain sensing. However, similar to the majority of other long-range BOTDAs, the state-of-the-art reports require > 1000 times average, severely restricting the sensing speed. The blind area over tens of kilometers caused by the nonuniform Brillouin response and parasitic amplitude modulation (AM) are crucial factors affecting the signal-to-noise ratio (SNR). Here, a comprehensive performance optimization and substantial enhancement for BOTDA sensors was presented by the direct demodulation of an injection-locked dual-bandwidth probe wave. Injection locking (IL) can completely eliminate the impact of AM noise; dual-bandwidth probe enables self-adaptive pulse loss compensation, thereby intensifying the SNR flatness along the ultralong fiber, and direct probe demodulation can overcome nonlocal effects and allows ∼19.7 dB enhancement of probe input power. Therefore, using only 100 times average, ∼148.3 km sensing, and ∼5 m spatial resolution were achieved with < ∼0.8 MHz standard deviation of Brillouin frequency shift (BFS) over a broad range (∼131.7 km). The reduction in averages was more than 10 times that of the reported majority of long-range BOTDAs. Such performances were achieved without using time-consuming or post-processing techniques, such as optical pulse coding and image denoising. Because this approach is compatible with optical chirp chain technique without frequency sweeping, fast acquisition (0.3 s) was also realized, which has the potential for fast sensing at 3.3 Hz along a ∼150 km fiber.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
In various sensor types, fiber-optic sensing based on Brillouin optical time domain analysis (BOTDA) can achieve a fully distributed measurement of temperature and strain along >100 km fibers and has wide application prospects in fire alarms and structural health monitoring for oil pipelines, civil buildings, facilities, etc. Overcoming the degradation of the signal-to-noise ratio (SNR) caused by the intrinsic loss of optical fibers is crucial for simultaneous long-range sensing and fast acquisition. However, additional depletion is generated by modulation instability [1,2] and spectral broadening induced by self-phase modulation (SPM) [3], and the maximal peak of the pulse must be less than ∼23 dBm.
Several methods have been proposed to solve this problem, including optical pulse coding (OPC) [4–15], digital denoising (wavelet analysis [16], and 2D/3D image denoising [17]), first- and higher-order distributed Raman amplification (DRA) [18–28], and distributed Brillouin amplification (DBA) [29–38]. By injecting multiple pulses simultaneously within one round-trip time, the signal-to-noise ratio (SNR) can be effectively improved using OPC [4–15]. Digital denoising is also an advanced method for enhancing SNR [16,17]. However, both methods require time-consuming postprocessing (i.e., decoding or denoising procedures). DRA can compensate for the impact of nonlinear effects but suffers from pump-to-probe relative intensity noise (RIN) transfer [21,23,24]. Furthermore, due to the much lower gain coefficient of stimulated Raman scattering (SRS) (∼3.8 × 10−4 W−1m−1) [39], a watt-level pump is required. In recent years, based on third-order random-laser-pumped DRA combined with 127 bits OPC (eight times pre-average) and 2D image denoising, a repeaterless BOTDA as long as 175 km has been achieved. However, this system requires two pumps at 1280/1365 nm, up to 2.51/1.27 W [28]. The drawbacks of DRA can be overcome by DBA: it is free of RIN transfer and features a much higher gain coefficient (∼0.3 W−1m−1) of stimulated Brillouin scattering (SBS) [36]; hence, sensing distances beyond 100 km can be achieved using only milliwatt-level pumps. Recently, a single-shot optical chirp chain (OCC) BOTDA with a 150 km sensing distance and 3 MHz standard deviation (STD) of the Brillouin frequency shift (BFS) has been achieved using DBA, at the expense of 2000 times average [37].
Nevertheless, similar to the majority of other long-range BOTDAs, all reported DBA-BOTDAs require averages of > 1000 times [29–38], which seriously hinders the increase in the acquisition rate. This is mainly attributed to the finite effective length Leff (∼21.7 km) of Brillouin pump [[1-exp(-αL)]/α, α is ∼0.2 dB/km, L is the fiber length]. Consequently, the generated nonuniform distribution of pulse power considerably deteriorates the SNR and forms a wide blind area over tens of kilometers [29–38]. In 2017, Kim et al. used the current modulation of a semiconductor laser to form a frequency modulation (FM) pump with an exponential bandwidth distribution to compensate for the pump loss in a self-adaptive manner [33]. A constant SNR distribution of ∼50 km was demonstrated at the expense of a much lower pumping efficiency (∼17 dBm pump is required for ∼100 km sensing in theory [36]; in practice, it is quite difficult to achieve a constant response for such long fibers considering the power restriction due to the onset of SBS noise [33]).
Recently, to achieve both gain flatness enhancement and high-efficiency pumping, a tailored DBA-BOTDA with dual-bandwidth pump was proposed by the authors, and a sensing distance of ∼98.9 km was demonstrated [36]. However, it is difficult to acquire the Brillouin gain spectra (BGS) with an additional probe light. Moreover, the probe input power is limited to ∼−10 dBm considering the non-local effect [40,41]. Correspondingly, average times as high as 12000 must be used [36]. Here, we give a comprehensive performance optimization and substantial enhancement for this method by combining it with the direct demodulation of a dual-bandwidth probe to suppress the nonlocal effect. In particular, the injection-locking (IL) concept [42–47] was utilized for the complete elimination of parasitic amplitude modulation (AM) noise while maintaining the ideal FM for dual-bandwidth probe. Such designs enable ∼9.7 dBm probe power injection (∼19.7 dB improvement compared with previous configuration [36]) without significant AM noise. As a result, ∼148.3 km sensing distance with ∼5 m spatial resolution was achieved by only 100 times average, without relying on intricate and time-consuming OPC/image denoising. This study is an extension of the work the authors presented at the Asia Communications and Photonics Conference (ACP) [48]. Notably, the utilized method is substantially different from that in [38], in which a differential pulse pair (DPP) was used to reduce the impact of AM noise, resulting in a two-fold increase in acquisition time.
2. Operation principle
Figure 1(a) shows a schematic of BOTDA with direct demodulation of an injection-locked dual-bandwidth probe wave. The Brillouin gain coefficient gSBS(z) can be controlled by altering the probe bandwidth Δν(z) according to the gain spectrum with a Lorentz shape: gSBS[Δν(z), δν(z)] = g0/{[1+Δν(z)/ΔνB][1+δν(z)2/(ΔνB/2)2]}, where g0 is the peak gain coefficient of the SBS with a limit of Δν(z) close to zero; ΔνB(∼30 MHz) is the intrinsic gain bandwidth; and δν (z)=νprobe(z)–νpulse–νB(z) [33,36], where νB is the BFS. For the FM probe, the frequency of the probe wave (νprobe) was also z-dependent. When the pulse arrives at z∈[0, L], the interacting probe wave is transmitted at the instant t = (2z–L)/vg (vg is the group velocity), as shown in Fig. 1(a). For the purpose of elegantly flattening the gain response along the fiber under test (FUT), the FM bandwidths of the probe wave need to be optimized. The dots in Fig. 1(a) show the positions at which the probe interacts with the pulse. When the interactions are located at the far end of the FUT, owing to the higher probe power in this regime (below Leff), the FM bandwidth of the probe wave is designed to be larger, giving rise to a much smaller gSBS (z). Thus, the excessive amplification and nonlinear impacts can be effectively weakened. However, the FM bandwidth of the probe in the left regime shown in Fig. 1(a) has a narrower Δν(z) to ensure sufficient pulse amplification by compensating for the loss of the probe wave. In the experiment, the optimized bandwidth of the first 131.7 km was a constant of 110 MHz, whereas the rest of the regime (16.6 km) held another constant of 250 MHz. Since the probe wave power is concentrated at the far end, the larger bandwidth in this regime should be used to obtain the lower Brillouin gain, and avoid the excessive pulse amplification. Here, a discontinuous bandwidth distribution with a jump (other than a continuous, exponentially increased bandwidth distribution [33]) was used to obtain the higher-efficiency amplification through a constant bandwidth of 110 MHz for the first 131.7 km fiber. According to our experimental results, the optimized length of the second subsection was 15∼20 km for a sensing distance of > 100 km. The FM period was 4 µs.
Fig. 1. Operation principle of BOTDA with the direct demodulation of an injected-locked dual-bandwidth probe wave. (a) Schematic of probe (red) and pulse (green) transmissions. Due to the smaller loss of Brillouin probe at the far end of FUT, probe bandwidth in this subsection is larger to avoid the excessive amplification and nonlinear accumulation. The dots show the positions and instants when the probe wave interacts with the pulse. (b) Spectral structure of probe wave and pulse and master laser (black). ΔνAOM is the frequency shift of the used AOM, νB is the BFS of FUT.
Direct demodulation of the probe wave [31,34] was employed. The normalized Brillouin response is given by ΔPprobe(z)/Pprobe(0)=gSBS[Δν(z), δν(z)]Ppulse(z)Δz [Pprobe(0) is the received probe power, Ppulse(z) is the pulse power distribution, z = vgt/2, and Δz is the spatial resolution] [4]. Considering the z-dependent δν(z), the order of frequency scanning differs from that of conventional BOTDAs [31]. Here, reconstruction of the complete BGS is performed via a circular FM shift in one period by 40 times, corresponding to the scanning steps of 2.75 MHz (110/40) and 6.25 MHz (250/40). Unlike conventional BOTDA, in terms of the order of the scanned frequency, the ascending order can be readily obtained with a simple and fast frequency rearrangement according to the known frequency of νprobe(z)–νpulse [31].
Owing to the wideband FM probe wave, the pulse power variation and distortion can be suppressed substantially during the BGS reconstruction, leading to a negligible nonlocal effect, which allows for a much higher probe injection [31,34,49,50]. This can be explained as follows: Theoretically, the evolution of pulse power can be described as [31]
For conventional BOTDAs, the nonlocal effect arises from the stepwise pulse power variation during the frequency scanning owing to frequency-shift-dependent pulse depletion or gain from a probe light [51]. For the direct demodulation of the FM probe, as previously mentioned, BGS reconstruction is performed by the circular FM shift in a very narrow range [maximum of 4 µs (equivalent to 400 m-long fiber), which is far less than the roundtrip time of FUT (∼1500 µs)]; therefore, the pulse interacts with the wideband probe wave in a similar pattern. Thus, the pulse power variation during BGS reconstruction is substantially suppressed. In other words, this much weaker frequency-shift dependence of pulse power amplification leads to a significantly suppressed nonlocal effect. For more details, see Refs. [31,49,50]. A detailed experimental validation of the suppressed nonlocal effect is shown in Fig. 8 (see Section 4).
Parasitical AM noise generated by imperfect optical and electronic modulations is an obstacle to direct probe demodulation [38]. Under our experimental conditions (Δz and the peak power of the injected pulse are ∼5 m and 20 dBm, respectively), the power fluctuation due to AM noise should be << the maximum of ΔPprobe/Pprobe(0) (∼8.2%). This is a very stringent threshold for AM noise that is difficult to be satisfied for ultralong sensing. We overcome this entirely by using the sideband IL of a slave laser (SL) [42–47]. The physical mechanism behind noise elimination originates from the band-pass filtering characteristics around the relaxation oscillation (RO) peak for the optical intensity fluctuation of the master laser (ML) [43–46,52]. An experimental demonstration of AM noise elimination for a dual-bandwidth probe wave is presented in section 3.1.
The spectral structure of the light used is shown in Fig. 1(b). The central frequencies of the optical source, pulse, and SL (probe wave) are denoted by νLD1, νpulse, and νprobe, respectively. Because of the frequency shift (ΔνAOM, 200 MHz) of the acousto-optic modulator (AOM), the frequency separation between νprobe and νLD1 should be ΔνAOM +νB (νB is ∼10.840 GHz for the FUT used). The spatial resolution is determined by
where Tp is the pulse-width. Experimentally, spatial resolution can be obtained from the rising distance for a hot-spot; if the length of hot-spot is the same with spatial resolution, the real spatial resolution can also be estimated by the full with half maximum (FWHM) of hot-spot. In addition, if frequency sweeping is used, the effective acquisition time T can be estimated as where Nf is the number of frequency sweeping, Nave is the used average times for a single frequency.3. Experimental setup and procedures
3.1 Experimental setup
The experimental setup of the ultralong BOTDA using injection-locked dual-bandwidth probe demodulation is shown in Fig. 2(a). A distributed feedback (DFB) laser diode (LD1) emitted at 1549.47 nm was split into two branches. An optical pulse with a width of ∼50 ns was formed by an AOM driven by the output of an arbitrary waveform generator (AWG). The extinction ratio and frequency shift of the AOM were ∼52 dB and 200 MHz, respectively. The dual-bandwidth probe was generated by an electro-optic modulator (EOM) with optical carrier suppression (via a modulator bias controller, not shown for clarity). Microwave modulation with dual bandwidths was realized using an in-phase and quadrature (IQ) mixer module [36]. The sampling rate of the AWGs for the generation of IQ outputs was 250 MSa/s. Another DFB LD2 (whose wavelength is close to the high-frequency side after EOM) was used as the slave laser for the side-band IL with an injection ratio of Rinj=−27 dB. As mentioned earlier, the central frequency of microwave source (MWS) should be νB+200 MHz. A polarization scrambler (PS) in the probe branch was employed to suppress polarization-dependent SBS noise. The pulse and dual-bandwidth probe were boosted by two Erbium-doped fiber amplifiers (EDFAs) before being injected into the FUT.
Fig. 2. Experimental setup. (a) Block diagram of experimental implementation. (b) Measured optical spectrum of SL without and with IL. The paths of dual-bandwidth probe wave with IL and pulse are expressed by red and green lines, respectively. (c) Measured waveform outputs by a detector with and without IL. No average was applied when measuring the waveform of SL with IL. Evidently, AM noise can be perfectly eliminated by IL. LD: laser diode; EOM: electro-optic modulator; MWS: microwave source; MWA: microwave amplifier; EDFA: Erbium-doped fiber amplifier; CIR: circulator; PS: polarization scrambler; VOA: variable optical attenuator; AOM: acousto-optic modulator; TFBG: tunable fiber Bragg grating; FUT: fiber under test; AWG: arbitrary waveform generator; OSA: optical spectrum analyzers; OSC: oscilloscope; DAQ: data acquisition card; PD: photo detector.
As for the receiver end, the dual-bandwidth probe passes through a tunable fiber Bragg grating (TFBG) with < 0.1 nm bandwidth to reject the noise of Rayleigh scattering and end refection of pulse. Using such a transmission design instead of reflection is helpful for avoiding excessive SNR degradation for the FM probe owing to frequency-dependent reflection, as previously mentioned. Finally, the received probe was detected by a photodetector (PD2) with a 400 MHz bandwidth and 8000 V/A trans-impedance gain. The sampling rate and vertical resolution of data acquisition card (DAQ) is 100 MSa/s and 14 bits, respectively. The probe optical spectrum was monitored using an optical spectrum analyzer (OSA). Another PD1 and an oscilloscope (OSC) were used to observe the pulse waveform after propagation in real time.
Figure 2(b) reveals the observed optical spectrum of SL with and without IL. The optical signal-to-noise ratio (OSNR) before IL is ∼19.2 dB (0.03 nm resolution); it is significantly enhanced (∼47 dB) by IL due to the higher power output of LD2 (∼4 dBm). Meanwhile, the side-mode suppression ratio (SMSR) before IL is ∼12 dB due to the imperfect carrier suppression; it is improved by ∼16 dB after IL. Figure 2(c) displays the measured waveform outputs by a detector with and without IL. AM noise as large as ∼23% ripple can be seen before IL (different amplitudes of AM noise are observed because of the use of dual-bandwidth modulation); after IL, this noise is almost completely eliminated, confirming the prominent advantage of using IL for our dual-bandwidth probe wave.
3.2 Experimental procedures
The basic experimental procedures include the following steps: 1) generate a circularly shifted dual-bandwidth FM probe (40 times in one period) with an AWG; 2) acquire the BGS using averages, rearrange the order of response according to the known frequency [31], and subtract the corresponding direct current component to remove the impact of SBS-induced noise; 3) perform Lorentz fitting to obtain the BFS distribution; and 4) repeat the above measurements eight times to estimate the STD distribution of BFS.
4. Experimental results for ∼148.3 km sensing
The length of FUT used (SMF28) is ∼148.3 km, which consists of four spools with lengths of 49.3, 24.9, 48.9, and 25.2 km. For comparison, sensing performance of DBA-BOTDA with conventional uniform modulation at 110 MHz bandwidth and ∼8.1 dBm probe (transparent pulse transmission) is also evaluated.
The raw data of the Brillouin response at 10.840 GHz for various Nave are shown in Figs. 3(a) and (b). Because of the linear FM in each period, the acquired response manifests as a Lorentz-shape in every period. For the proposed BOTDA with a dual-bandwidth probe [see (a)], a much higher SNR is observed along the entire fiber owing to the perfect compensation of fiber loss. By comparison, for conventional uniform bandwidth modulation, the smaller noise induced by the SBS of the probe (represented by dotted lines) dominates the entire valley of the Brillouin response [see (b)]. This noise is confirmed by noting that it can be observed even when the pulse leaves the FUT. Although this effect is negligible for the proposed configuration, it is non-negligible for the conventional uniform modulation. However, it can be easily excluded by subtracting the corresponding direct current component after frequency rearrangement because it is identical for the same frequency. The reconstructed BGSs are shown in Figs. 3(c) and (d), where only 100 times average was utilized. The substantially enhanced flatness is clearly visible [see (c)] when using the proposed method. The reconstructed Brillouin responses at 10.840 GHz are shown in Figs. 3(e) and (f). Again, a much wider measurement blind area with a low SNR (60∼130 km) can be observed for the conventional uniform modulation [see (f)]. For the case of 100 times averaging, the calculated minimal SNRs [defined as the ratio of response mean and STD [17]] for the proposed scheme is ∼14.7 dB, whereas it is ∼−0.2 dB for the conventional uniform modulation; thus, SNR enhancement of ∼14.9 dB over a quite broad regime was realized.
Fig. 3. Experimental results. (a) Raw data of Brillouin response at 10.840 GHz for the proposed configuration. (b) Raw data of Brillouin response at 10.840 GHz for DBA-BOTDA with conventional uniform bandwidth modulation. (c) Reconstructed BGS of proposed configuration. (d) Reconstructed BGS of DBA-BOTDA with conventional uniform bandwidth modulation. (e) Reconstructed response after rearrangement for the proposed configuration at 10.840 GHz. (f) Reconstructed response after rearrangement for the conventional uniform bandwidth modulation at 10.840 GHz. Nave is the average times. The dotted lines in a and b represent the noise peak due to SBS. In (c) and (d), 100 times average was used.
The BFS distributions extracted by Lorentz fitting for various average times are shown in Figs. 4(a) and (b). The corresponding STDs of the BFS are shown in Figs. 4(c) and (d) for eight repeated measurements. Evidently, a very large BFS fluctuation is observed for conventional uniform bandwidth modulation, which is generated by incorrect fitting owing to the extremely low SNR over a very wide regime [see (b), (d)]. This tendency cannot be altered with 500 times average [∼39 MHz maximal STD over 44∼131 km, see (d)]. With increased average times (100, 300, and 500), the corresponding STD over 0∼131.7 km for the proposed scheme decreases (less than ∼0.8, 0.6, and 0.3 MHz respectively), whereas the residual regime is correspondingly less than ∼1.8, 1.3, and 0.7 MHz [see (c)].
Fig. 4. Results of Lorentz fitting. (a) BFS distribution of the proposed configuration. (b) BFS of DBA-BOTDA with conventional uniform bandwidth modulation. (c) STD distribution of proposed configuration. (d) STD of DBA- BOTDA with conventional uniform bandwidth modulation. Nave is the average times. The much larger STD for conventional scheme shown in (b), (d) is generated by the incorrect fitting due to extremely lower SNR over a very wide regime.
To demonstrate the advantage of the proposed scheme with IL, a comparison experiment without IL was also conducted for the same probe power. In this case, the block diagram referred to as “side-band IL” (see Fig. 2) is replaced by a TFBG operating in transmission to select the high-frequency probe (such a design can avoid further deterioration of AM noise owing to the frequency-dependent reflection). The recovered Brillouin response is shown in Fig. 5. Because the noise level is comparable to that of the Brillouin response, a noisy trace is observed, which fully impedes further fitting for BFS extraction. This comparison clearly indicates the advantages of using a dual-bandwidth probe with an IL. Additionally, due to the abrupt decrease of Brillouin gain coefficient for the last 16.6 km fiber (caused by larger probe bandwidth of 250 MHz), the response profile of this regime has significant variations. Specifically, a first falling beginning at 131.7 km is attributed to the net loss for pulse, and a smaller pulse power rising is observed at the far end, due to lower loss of probe power in this subsection. A jump observed at 131.7 km is a consequence of the discontinuous bandwidth variation.
Previous studies on the effects of pulse peak powers have demonstrated that the positioning precision is affected by slow light owing to the finite phonon lifetime (∼10 ns) of SBS [30,31], as shown in Fig. 6(a). This degradation is ∼1.3 m due to a ∼13 ns delay of the rising edge. A pulse-width compression instead of broadening in this work was observed, which is consistent with the observation in Ref. [31]. This phenomenon may be explained by the phonon intensity reduction at the falling edge of pulse, which is determined by ∼10 ns lifetime, smaller than the used pulse-width (∼50 ns). In addition, an error within ∼5 MHz for the BFS at the far end can be observed owing to the SPM of the pulse [see Fig. 6(b)]. However, such errors are limited only within a shorter region and can be reduced in practice by discarding this subsection as a sensing fiber or measuring the BFS difference with/without temperature/strain change. Therefore, the DPP method [38] (two-fold increase in measure-ment time, as previously mentioned) was not used in this work.
Fig. 6. Effects of slow-light delay and SPM. (a) Pulse waveforms after transmission with probe on and off. (b) BFS distributions for different on-off gains of pulse by slightly adjusting the probe input power (∼−3 dB on-off gain by decreasing the probe power of ∼0.3 dB). In (a), the top and bottom correspond to the proposed scheme and conventional uniform modulation, respectively, and the peak power of injected pulse was ∼20 dBm. Average of 100 times was used.
We then implemented the temperature sensing with 100 times average by introducing a ∼5 m long hotspot at the far end of the FUT in a water bath. The room temperature was maintained at ∼25 °C using an air conditioner. The measured BFS as a function of temperature increase is shown in Fig. 7(a). The fitted sensitivity of ∼1 MHz/°C agreed well with previously reported results. From the extracted temperature distribution around the hot spot shown in Fig. 7(b), a spatial resolution of ∼5 m is achieved by noting that the observed FWHM of the hot spot is consistent with its actual length.
Fig. 7. Results of temperature sensing. (a) Measured BFS of hot spot as a function of temperature. (b) Extracted temperature distribution around ∼5 m hot spot at ∼60 °C. Average of 100 times was used.
Fig. 8. (a) Validation in suppressed nonlocal effect during frequency sweeping, by measuring the power variation of pulse after FUT (∼6.2%). (b) BFS distributions along the FUT before and after swapping the two ends of sensing fiber. Except for the BFS deviation (∼4 MHz maximal value) within the last (first) ∼4.5 km regime before (after) swapping due to SPM, the two BFS distributions were consistent with each other, implying that the nonlocal effect was controlled well. Measurement condition is the same with Figs. 3(a), (c) and (e). In (b), average of 300 times was used.
The experimental validation of the suppressed nonlocal effect is shown in Fig. 8, where the pulse power variation after the FUT during frequency sweeping [by circularly shifting the frequency of the FM probe in one period (4 µs)] was measured [see (a)]. A smaller power variation of ∼6.2% was obtained, confirming the suppression capability of the nonlocal effect. We further performed the comparison of BFS distribution before and after swapping the two ends of sensing fiber, as shown in Fig. 8(b). Except for the BFS deviation (∼4 MHz maximal value) within the last (first) ∼4.5 km regime before (after) swapping due to SPM, the two BFS distributions were consistent with each other, implying that the nonlocal effect was controlled well.
5. Sensing of ∼148.3 km without frequency sweeping
It should be stressed that the proposed scheme is fully compatible with the reported single-shot OCC technology [37,38,53] because the BGS can also be directly extracted even without the need for additional frequency sweeping owing to the linear FM [see Figs. 3(a) and (b)]. In this section, we attempt to realize ultralong sensing without frequency sweeping. The fiber length of the second subsection with a 250 MHz bandwidth was set to 20 km. We fixed the pulse width at ∼50 ns and reduced the FM period to 80 ns, corresponding to a spatial resolution of ∼8 m. Note that, in this case, the spatial resolution is determined by the maximum pulse width and FM period [37]. The sampling rate of the AWG was set to 500 MSa/s. Because the vertical resolution of the DAQ used with 4 GSa/s is lower (8 bits), we acquire the sensing signal by moving average of every eight sampling points, corresponding to 11 bits vertical resolution and a 500 MSa/s equivalent sampling rate (or 2 ns sampling period). Thus, the number of frequency reconstruction points is 40(80 ns/2 ns), and the frequency step is the same as that in Section 4. Note that the moving average used in this study is unnecessary if a DAQ with a higher vertical resolution and sampling rate is available.
Figure 9(a) shows the reconstructed BGS with average of 200 times. Unlike in Figs. 3(c) and (d), where the BGSs are fully symmetrical along frequency shift axis, here, we found that the BGS has a nonsymmetrical feature (the peak is shifted to higher frequency). This has been theoretically verified in Ref. [53] and is caused by the transient SBS because the frequency change is too fast. During drawing Fig. 9(a), the “ghost lines” caused by transient process in the adjacent subsection [53] have been removed. A higher SNR is observed in this diagram. The STD of fitted BGS frequency peak is shown in Fig. 9(b) by eight repeated measurements. The maximal STD along FUT is ∼3 MHz. This STD degradation compared with Fig. 4(c) is mainly attributed to the smaller vertical resolution of the DAQ used and needs to be investigated further in future work. Figure 9(c) shows the extracted temperature distribution near the ∼8 m hot spot heated at ∼50 °C. The spatial reconstruction step (8 m) is determined by the FM period. The diagram clearly confirms the ∼8 m spatial resolution.
Fig. 9. Sensing at ∼148.3 km without frequency sweeping. (a) Reconstructed 3D BGS. (b) STD of fitted frequency peak for 8 repeated measurements. (c) Temperature distribution near ∼8 m hot spot. Average of 200 times was used.
6. Conclusions and discussions
In summary, we proposed a highly energy-efficient, and simplified configuration for ultralong BOTDA sensors by directly demodulation of an injection-locked dual-bandwidth probe. These designs can simultaneously compensate for the loss along the ultralong fiber by dual-bandwidth modulation, avoid the effects of AM noise by the IL, and eliminate the nonlocal effects that enables ∼19.7 dB probe power enhancement. With only 100 times average and 9.7 dBm probe, ∼148.3 km sensing with ∼5 m spatial resolution has been demonstrated without any time-consuming OPC and image denoising.
For the case of frequency sweeping, according to Eq. (3), for L = 150 km, Nave = 100, and Nf =40, the acquisition time T corresponds to 6 s (150 ms for a single-frequency point). In our experiment in Section 4, a nonreal-time software average was used, leading to a total acquisition time of ∼29 s. If a real-time hardware average based on a digital signal processor or field-programmable gate array is available, a total acquisition time of ∼6 s can be readily utilized.
Furthermore, because this approach is compatible with the OCC technique, we also performed sensing along a ∼148.3 km fiber without frequency sweeping under only 200 times average. In this case, the response time for single measurement is further reduced to 0.3 s (1.5 ms×200). To the best of our knowledge, this is the fastest acquisition speed for such long BOTDAs, paving the way for fast dynamic sensing along ultralong fibers. Further work should focus on improving the sensing performance by optimizing the system parameters, exploring the influence of nonsymmetrical BGS on sensing accuracy, and possible compensation configurations, such as pattern recognition [38] and artificial neural network-based machine learning.
The measurement range can be further broadened by simultaneously widening the dual bandwidth of the probe, provided that their ratio remains unchanged to maintain the flatness of the Brillouin response. Moreover, a more cost-effective structure is feasible by replacing the AWG and IQ modules (see Fig. 2) with a voltage-controlled oscillator [31], whereby the complexity of the proposed scheme is comparable to that of the standard BOTDA. In addition, this configuration is applicable for the case where the BFSs along different spools have a larger difference, and it is achieved by introducing additional frequency shifts for the corresponding subsections during the generation of the dual-bandwidth probe wave.
Finally, Table 1 gives a summary of performance list of long-range BOTDAs developed in recent years, in which the figure of merit (FoM) is evaluated according to [54]
Table 1. Performance list for long-range BOTDAs developed in recent years.
Funding
Sichuan Science and Technology Program (2019YJ0530); Innovative Training Program for College Student of Sichuan Normal University (X202110636178); National Natural Science Foundation of China (61205079).
Acknowledgments
The authors greatly thank Prof. Zinan Wang from University of Electronic Science and Technology of China for helpful discussions.
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. M. Alem, M. A. Soto, and L. Thévenaz, “Analytical model and experimental verification of the critical power for modulation instability in optical fibers,” Opt. Express 23(23), 29514–29532 (2015). [CrossRef]
2. 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]
3. S. M. Foaleng, F. Rodríguez-Barrios, S. Martin-Lopez, M. González-Herráez, and L. Thévenaz, “Detrimental effect of self-phase modulation on the performance of Brillouin distributed fiber sensors,” Opt. Lett. 36(2), 97–99 (2011). [CrossRef]
4. M. A. Soto, G. Bolognini, and F. Di Pasquale, “Analysis of pulse modulation format in coded BOTDA sensors,” Opt. Express 18(14), 14878–14892 (2010). [CrossRef]
5. M. A. Soto, G. Bolognini, F. Di Pasquale, and L. Thévenaz, “Simplex-coded BOTDA fiber sensor with 1 m spatial resolution over a 50 km range,” Opt. Express 35(2), 259–261 (2010). [CrossRef]
6. H. Liang, W. H. Li, N. Linze, L. Chen, and X. Y. Bao, “High-resolution DPP-BOTDA over 50 km LEAF using return-to-zero coded pulses,” Opt. Lett. 35(10), 1503–1505 (2010). [CrossRef]
7. M. A. Soto, S. L. Floch, and L. Thévenaz, “Bipolar optical pulse coding for performance enhancement in BOTDA sensors,” Opt. Express 21(14), 16390–16397 (2013). [CrossRef]
8. F. Baronti, A. Lazzeri, R. Romcella, R. Saletti, A. Signorini, M. A. Soto, G. Bolognini, and F. Di Pasquale, “SNR enhancement of Raman-based long-range distributed temperature sensors using cyclic Simplex codes,” Electron. Lett. 46(17), 1221–1222 (2010). [CrossRef]
9. M. Taki, A. Signorini, C. J. Oton, T. Nannipieri, and F. Di Pasquale, “Hybrid Raman/Brillouin-optical-time-domain-analysis-distributed optical fiber sensors based on cyclic pulse coding,” Opt. Lett. 38(20), 4162–4165 (2013). [CrossRef]
10. M. Taki, Y. Muanenda, C. J. Oton, T. Nannipieri, A. Signorini, and F. Di Pasquale, “Cyclic pulse coding for fast BOTDA fiber sensors,” Opt. Lett. 38(15), 2877–2880 (2013). [CrossRef]
11. Y. Muanenda, M. Taki, and F. Di Pasquale, “Long-range accelerated BOTDA sensor using adaptive linear prediction and cyclic coding,” Opt. Lett. 39(18), 5411–5414 (2014). [CrossRef]
12. H. Iribas, A. Loayssa, F. Slorian, M. Llera, and S. Le Floch, “Cyclic coding for Brillouin optical time-domain analyzers using probe dithering,” Opt. Express 25(8), 8787–8800 (2017). [CrossRef]
13. S. Le Floch, F. Sauser, M. Llera, and E. Rochat, “Novel Brillouin optical time-domain analyzer for extreme sensing range using high-power flat frequency-coded pump pulses,” J. Lightwave Technol. 33(12), 2623–2627 (2015). [CrossRef]
14. X. Sun, Z. Yang, X. Hong, S. Zaslawski, S. Wang, M. A. Soto, X. Gao, J. Wu, and L. Thévenaz, “Genetic optimised aperiodic code for distributed optical fibre sensors,” Nat. Commun. 11(1), 5574–5584 (2020). [CrossRef]
15. Y. Zhou, L. Yan, C. Liu, H. He, Z. Li, H. Qian, J. Ye, W. Pan, and B. Luo, “Hybrid aperiodic coding for SNR improvement in a BOTDA fiber sensor,” Opt. Express 29(21), 33926–33936 (2021). [CrossRef]
16. M. A. Farahani, M. T. V. Wylie, E. Castillo-Guerra, and B. G. Colpitts, “Reduction in the number of averages required in BOTDA sensors using wavelet denoising techniques,” J. Lightwave Technol. 30(8), 1134–1142 (2012). [CrossRef]
17. M. A. Soto, J. A. Ramírez, and L. Thévenaz, “Intensifying the response of distributed optical fibre sensors using 2D and 3D image restoration,” Nat. Commun. 7(1), 10870 (2016). [CrossRef]
18. F. Rodríguez-Barrios, S. Martín-López, A. Carrasco-Sanz, P. Corredera, J. D. Ania-Castañón, L. Thévenaz, and M. González-Herráez, “Distributed Brillouin fiber sensor assisted by first-order Raman amplification,” J. Lightwave Technol. 28(15), 2162–2172 (2010). [CrossRef]
19. X. H. Jia, Y. J. Rao, L. Chang, C. Zhang, and Z. L. Ran, “Enhanced sensing performance in long distance Brillouin optical time-domain analyzer based on Raman amplification: theoretical and experimental investigation,” J. Lightwave Technol. 28(11), 1624–1630 (2010). [CrossRef]
20. X. Angulo-Vinuesa, S. Martin-Lopez, P. Corredera, and M. Gonzalez-Herraez, “Raman-assisted Brillouin optical time-domain analysis with sub-meter resolution over 100 km,” Opt. Express 20(11), 12147–12154 (2012). [CrossRef]
21. M. A. Soto, G. Bolognini, and F. Di Pasquale, “Optimization of long-range BOTDA sensors with high resolution using first-order bi-directional Raman amplification,” Opt. Express 19(5), 4444–4457 (2011). [CrossRef]
22. S. Martin-Lopez, M. Alcon-Camas, F. Rodriguez, P. Corredera, J. D. Ania-Castañon, L. Thévenaz, and M. Gonzalez-Herraez, “Brillouin optical time-domain analysis assisted by second-order Raman amplification,” Opt. Express 18(18), 18769–18778 (2010). [CrossRef]
23. X. H. Jia, Y. J. Rao, Z. N. Wang, W. L. Zhang, C. X. Yuan, X. D. Yan, J. Li, H. Wu, Y. Y. Zhu, and F. Peng, “Distributed Raman amplification using ultra-long fiber laser with a ring cavity: characteristics and sensing application,” Opt. Express 21(18), 21208–21217 (2013). [CrossRef]
24. X. H. Jia, Y. J. Rao, C. X. Yuan, J. Li, X. D. Yan, Z. N. Wang, W. L. Zhang, H. Wu, Y. Y. Zhu, and F. Peng, “Hybrid distributed Raman amplification combining random fiber laser based 2nd-order and low-noise LD based 1st-order pumping,” Opt. Express 21(21), 24611–24619 (2013). [CrossRef]
25. M. A. Soto, X. Angulo-Vinuesa, S. Martin-Lopez, S. H. Chin, J. D. Ania-Castañon, P. Corredera, E. Rochat, M. Gonzalez-Herraez, and L. Thévenaz, “Extending the real remoteness of long-range Brillouin optical time-domain fiber analyzers,” J. Lightwave Technol. 32(1), 152–162 (2014). [CrossRef]
26. X. H. Jia, Y. J. Rao, K. Deng, Z. X. Yang, L. Chang, C. Zhang, and Z. L. Ran, “Experimental demonstration on 2.5-m spatial resolution and 1°C temperature uncertainty over long-distance BOTDA with combined Raman amplification and optical pulse coding,” IEEE Photonics Technol. Lett. 23(7), 435–437 (2011). [CrossRef]
27. Y. Fu, Z. N. Wang, R. C. Zhu, N. T. Xue, J. L. Jiang, C. Y. Lu, B. Zhang, L. Yang, D. Atubga, and Y. J. Rao, “Ultra-long-distance hybrid BOTDA/Φ-OTDR,” Sensors 18(4), 976 (2018). [CrossRef]
28. Y. Fu, R. C. Zhu, B. Han, H. Wu, Y. J. Rao, C. Y. Lu, and Z. N. Wang, “175 km repeaterless BOTDA with hybrid high-order random fiber laser amplification,” J. Lightwave Technol. 37(18), 4680–4686 (2019). [CrossRef]
29. H. Q. Chang, X. H. Jia, X. L. Ji, C. Xu, L. Ao, H. Wu, Z. N. Wang, and W. L. Zhang, “DBA-based BOTDA using optical-comb pump and pulse coding with a single laser,” IEEE Photonics Technol. Lett. 28(10), 1142–1145 (2016). [CrossRef]
30. J. Urricelqui, M. Sagues, and A. Loayssa, “Brillouin optical time-domain analysis sensor assisted by Brillouin distributed amplification of pump pulses,” Opt. Express 23(23), 30448–30458 (2015). [CrossRef]
31. J. J. Mompó, J. Urricelqui, and A. Loayssa, “Brillouin optical time-domain analysis sensor with pump pulse amplification,” Opt. Express 24(12), 12672–12681 (2016). [CrossRef]
32. X. H. Jia, H. Q. Chang, L. Ao, X. L. Ji, C. Xu, and W. L. Zhang, “BOTDA sensors enhanced using high-efficiency second-order distributed Brillouin amplification,” Opt. Express 24(13), 14079–14085 (2016). [CrossRef]
33. Y. H. Kim and K. Y. Song, “Tailored pump compensation for Brillouin optical time-domain analysis with distributed Brillouin amplification,” Opt. Express 25(13), 14098–14105 (2017). [CrossRef]
34. X. H. Jia, H. Q. Chang, K. Lin, C. Xu, and J. G. Wu, “Frequency-comb-based BOTDA sensors for high-spatial-resolution/long-distance sensing,” Opt. Express 25(6), 6997–7007 (2017). [CrossRef]
35. K. Lin, X. H. Jia, H. L. Ma, C. Xu, X. Zhang, and L. Ao, “Comparative study on reduced impacts of Brillouin pump depletion and nonlinear amplification in coded DBA-BOTDA,” Chin. Opt. Lett. 16(9), 090604 (2018). [CrossRef]
36. H. L. Ma, X. H. Jia, K. Lin, and C. Xu, “High-efficiency DBA-BOTDA with optimized SNR by multiple bandwidths pump modulation,” Chin. Opt. Lett. 17(1), 010603 (2019). [CrossRef]
37. Y. Dong, B. Wang, C. Pang, D. Zhou, D. Ba, H. Zhang, and X. Bao, “150 km fast BOTDA based on optical chirp chain probe wave and Brillouin loss scheme,” Opt. Lett. 43(19), 4679–4682 (2018). [CrossRef]
38. B. Z. Wang, B. H. Fan, D. W. Zhou, C. Pang, Y. Li, D. X. Ba, and Y. K. Dong, “High-performance optical chirp chain BOTDA by using a pattern recognition algorithm and the differential pulse-width pair technique,” Photonics Res. 7(6), 652–658 (2019). [CrossRef]
39. C. Headley and G. P. Agrawal, “Raman amplification in fiber optical communication systems,” San Diego: Academic Press (2005).
40. A. Dominguez-Lopez, X. Angulo-Vinuesa, A. Lopez-Gil, S. Martin-Lopez, and M. Gonzalez-Herraez, “Non-local effects in dual-probe-sideband Brillouin optical time domain analysis,” Opt. Express 23(8), 10341–10352 (2015). [CrossRef]
41. A. Dominguez-Lopez, Z. Yang, M. A. Soto, X. Angulo-Vinuesa, S. Martin-Lopez, L. Thévenaz, and M. Gonzalez-Herraez, “Novel scanning method for distortion-free BOTDA measurements,” Opt. Express 24(10), 10188–10204 (2016). [CrossRef]
42. B. Wang, X. Y. Fan Y, X. Fu, and Z. Y. He, “Dynamic strain measurement with kHz-level repetition rate and centimeter-level spatial resolution based on Brillouin optical correlation domain analysis,” Opt. Express 26(6), 6916–6928 (2018). [CrossRef]
43. A. Murakami, “Synchronization of chaos due to linear response in optically driven semiconductor lasers,” Phys. Rev. E 65(5), 056617 (2002). [CrossRef]
44. A. Murakami and K. A. Shore, “Chaos-pass filtering in injection-locked semiconductor lasers,” Phys. Rev. A 72(5), 053810 (2005). [CrossRef]
45. A. Murakami, “Phase locking and chaos synchronization in injection-locked semiconductor lasers,” IEEE J. Quantum Electron. 39(3), 438–447 (2003). [CrossRef]
46. A. Uchida, Y. Liu, and P. Davis, “Characteristics of chaotic masking in synchronized semiconductor lasers,” IEEE J. Quantum Electron. 39(8), 963–970 (2003). [CrossRef]
47. L. Thévenaz, S. L. Floch, D. Alasia, and J. Troger, “Novel schemes for optical signal generation using laser injection locking with application to Brillouin sensing,” Meas. Sci. Technol. 15(8), 1519–1524 (2004). [CrossRef]
48. X. Jia, H. Ma, J. Lin, S. Xu, and X. Wei, “Giant enhancement of gain flatness for ultra-long BOTDA by injection-locked dual-bandwidth pump demodulation,” in Asia Communications and Photonics Conference (ACPC) 2019, OSA Technical Digest (Optica Publishing Group, 2019), paper M3A.5.
49. R. Ruiz-Lombera, J. Urricelqui, M. Sagues, J. Mirapeix, J. M. Lòpez-Higuera, and A. Loayssa, “Overcoming nonlocal effects and Brillouin threshold limitations in Brillouin optical time-domain sensors,” IEEE Photonics J. 7(6), 1–9 (2015). [CrossRef]
50. J. Urricelqui, M. Sagues, and A. Loayssa, “Synthesis of Brillouin frequency shift profiles to compensate non-local effects and Brillouin induced noise in BOTDA sensors,” Opt. Express 22(15), 18195–18202 (2014). [CrossRef]
51. L. Thévenaz, S. F. Mafang, and J. Lin, “Effect of pulse depletion in a Brillouin optical time-domain analysis system,” Opt. Express 21(12), 14017–14035 (2013). [CrossRef]
52. A. Murakami, K. Kawashima, and K. Atsuki, “Cavity resonance shift and bandwidth enhancement in semiconductor lasers with strong light injection,” IEEE J. Quantum Electron. 39(10), 1196–1204 (2003). [CrossRef]
53. D. Zhou, Y. Dong, B. Wang, C. Pang, D. Ba, H. Zhang, Z. Lu, H. Li, and X. Bao, “Single-shot BOTDA based on optical chirp chain probe wave for distributed ultra-fast measurement,” Light: Sci. Appl. 7(1), 32–43 (2018). [CrossRef]
54. M. A. Soto and L. Thévenaz, “Modeling and evaluating the performance of Brillouin distributed optical fiber sensors,” Opt. Express 21(25), 31347–31366 (2013). [CrossRef]
