Long-range and high-precision correlation optical time-domain reflectometry utilizing an all-fiber chaotic source

Z. N. Wang; M. Q. Fan; L. Zhang; H. Wu; D. V. Churkin; Y. Li; X. Y. Qian; Y. J. Rao

doi:10.1364/OE.23.015514

1. Introduction

Chaotic sources have attracted much attention due to its unique characteristics and the huge potential for various applications including physical random bit generation [1], secure communication [2], chaotic lidar [3], and so on. As an example of the applications of electrical driven chaotic source, in 2007, a cross-correlation OTDR [4] was proposed utilizing a pseudo-random pulse sequence as the signal. However, its measurement accuracy is still limited by the bottleneck of the pseudo-random modulation bandwidth, and it is difficult and costly to generate broadband electrical random codes. For the above reasons, electronic chaotic sources are hard to compete with optical chaotic sources which have advantages of wider bandwidth and faster chaotic dynamics.

Followed this work, Y. Wang et al. demonstrated a 6cm spatial correlation OTDR based on a multi-GHz optical chaotic source utilizing LD laser [5]. This is the first proof-of-concept experiment of correlation OTDR by utilizing the optical chaotic source. However, that work only presented a 140m detection ranging. There is always tradeoff between the spatial resolution and detection range.

Since then, some modified correlation OTDR techniques have been proposed to further improve the spatial resolution and operation range, and the applications are shown in the wavelength/time division multiplexing passive optical network (W/TDM-PON) for the fiber fault detection [6–9]. All these works have the ability to locate the fiber fault accurately with several centimeter spatial resolution and several tens of kilometers of operation range. Most of the previous optical chaotic sources are based on semiconductor lasers with optical feedback [10], optical injection [11] or optoelectronic feedback [12] methods. However, limited by the bandwidth of the source, the spatial resolution cannot be further improved; what is more important, the underlying physics limiting the performance of such system has not been adequately addressed. Shentu et. al. demonstrate a 217km OTDR with 10cm spatial resolution based on single photon detection, while the two-point resolution is as large as 100m [13]. Moreover, the system is complex and costly, making it difficult for field application.

In this work, we propose a long-range, high-precision correlation OTDR based on an all fiber supercontinuum (SC) source. The time-domain properties of the proposed source are investigated for the first time. It is demonstrated that the SC source is a powerful tool to generate ultra-wideband fluctuation in time domain, which is an ideal source for correlation OTDR, and its ability to achieve millimeter-scale spatial resolution is shown. On the other hand, the limiting factors for realizing long-range correlation OTDR are analyzed, such as integral effect of weak Rayleigh backscattering and the fiber loss. We analyze the potential methods to overcome the bottleneck of operation range. Specifically, the effects of distributed amplification and acquisition duration on the enhancement of signal quality are analyzed. As a result, we experimentally demonstrate a correlation OTDR which can achieve 100 kilometers fiber fault location range with 8.2cm spatial resolution (1.2 million resolved points), with only 2ms data acquisition time. To the best of our knowledge, this is the longest fiber fault location system based on the correlation OTDR technique.

2. Characteristics of the supercontinuum source

The schematic setup of the proposed system is shown in Fig. 1(a) . The proposed correlation OTDR system mainly consists of two parts (i.e., source and OTDR). As the box diagram in Fig. 1(a) shows, the source simply consists of a 1455nm quasi-CW Raman fiber laser and 16km TrueWave (TW) fiber. The TW fiber has a zero dispersion wavelength at 1440nm with a dispersion slope of 0.045 ps/nm²/km. The pump operates in anomalous-dispersion regime of TW fiber thus modulation instability (MI) could happen and the SC generation can be realized with the combination of modulation instability, stimulated Raman scattering (SRS) and four-wave mixing (FWM) [14–16].

Fig. 1 (a) Schematic setup (ISO: isolator; TW Fiber: TrueWave Fiber; SC source: supercontinuum source; Cir: circulator; FBG: fiber Bragg grating; OC: optical coupler; PD: photo-detector; RA: Raman amplifier; FUT: fiber under test; OSC: oscilloscope); (b) spectrum evolution with pump power (the vertical dotted line corresponds to the zero-dispersion wavelength of the TW fiber).

Download Full Size | PDF

The optical spectra recorded at the end of the TW fiber are shown in Fig. 1(b). It should be noted that, when pump power reaches 0.95W, spectral components extended to 1550nm region are generated. If the pump power is increased to 1.48W, the generated spectrum has the widest bandwidth, i.e., 141nm within 10dB range. By further increasing pump power, the bandwidth of generated spectrum would be reduced. The reason is that more powerful pump stimulates higher-order Stokes waves and enhances total pump-Stokes conversion efficiency, thus shorter wavelength photons are significantly depleted. However, if the pump power reaches 3.49W, the source output spectrum covers from 1588nm to 1650nm. Therefore, it is possible to perform data transmission and optical cable maintenance at the same time without interruption of data transmission.

We investigate the source in time domain (filtered with 0.26nm FBG). A 45GHz PD and an oscilloscope with 25GHz bandwidth and 50Gs/s sampling rate are used. The time series are shown in Fig. 2(a) . It shows that the SC has time dynamics with high-contrast fluctuations which is a favorable condition for correlation analysis. The auto-correlation function (ACF) with DC removed is calculated as 20ps (corresponding to ~2mm spatial resolution for OTDR). The narrow-width ACF peak could result in the ultra-fine resolution for correlation OTDR. To verify this point, we demonstrate a fault-locating experiment with ultra-fine spatial resolution while detection range is 117m. The experiment setup is similar to Fig. 1, but the RA is bypassed, two PDs respectively with 45GHz and 20GHz bandwidth are used, and the sampling rate of the oscilloscope is set to be 100GS/s. As shown in the correlation trace in Fig. 2(b), the spatial resolution is 9mm according to the full width of half maximum (FWHM) of the correlation peak. With a pair of matched PDs, the spatial resolution could even be smaller. Therefore, the demonstrated SC source could be an important member in the family of chaotic sources because of its temporal properties.

Fig. 2 (a) Time series of optical output (with 1.48W pump power); (b) correlation trace demonstrating ultra-fine spatial resolution (inset: location of fiber-fault at 117m).

Download Full Size | PDF

3. Discussions for performance enhancement of correlation OTDR

With the increase of sensing range, OSNR of the detected signal would decrease thus it will be difficult to recognize the peaks among correlation trace. To extend the detection range of correlation OTDR, the signal quality must be improved.

First of all, the total Rayleigh backscattering power detected at receiver versus fiber length should be evaluated. Rayleigh backscattering along the fiber can be considered as a linear response and evenly distributed. At the receiver side, Rayleigh scattering components are an integral result. In another word, the fiber can be seen as a point-mirror with the integral Rayleigh scattering. When the fiber length is long enough, the Rayleigh scattering power will be significant [17], though the Rayleigh coefficient is very small ( $ε$ = 4.3 × 10⁻⁸/m @ 1550nm). The integral Rayleigh-backscattered power could be expressed as $P_{i n} \cdot \int_{0}^{L} ε \cdot e^{- 2 α_{s} l} d l$ . α_s represents the loss coefficient of signal (measured as 0.18dB/km), L is the fiber length. The integral terms in above expression $ε \cdot e^{- 2 α_{s} l} d l$ represents equivalent reflectivity at position l, which includes the round-trip fiber loss. The green solid curve in Fig. 3(a) shows the accumulated reflectivity taking both Rayleigh-backscattering and round-trip loss into account, which saturates at ~20km.

Fig. 3 (a) Accumulated Raman gain over length and equivalent distribution reflectivity (inset: ratio of end-reflected signal power to Rayleigh-backscattered power along fiber without Raman amplification); (b) calculated SNR versus acquisition duration.

Download Full Size | PDF

Furthermore, we calculate the ratio of fiber-fault-reflected signal power to the Rayleigh-backscattered noise, which could be expressed as:

O S N R = \frac{P_{i n} \cdot R \cdot e^{- 2 α_{s} L}}{P_{i n} \cdot \int_{0}^{L} ε \cdot e^{- 2 α_{s} l} d l}

where R stands for the reflectivity of the fiber fault. In our model, we assume a 4% Fresnel reflection to simulate the fiber fault. As shown in the inset of Fig. 3(a), the OSNR exponentially decreases with increased fiber length. From this calculation we see that if fiber length is 100km, the components of fiber-fault-reflected signal accounts very little (−17.2dB) of the total power received by the detector.

Here we show that the distributed amplification will perform much better than a pre-amplifier in front of the detector. A pre-amplifier such as an EDFA will boost the total optical power, but the OSNR would not enhance since the signal is actually buried in the in-band noise. On the other hand, when a distributed Raman amplifier is used, the intensity evolution of the chaotic probe light launched into the fiber can be calculated by the equations [18,19]:

\frac{d I_{p}}{d z} = - G_{R} I_{p} I_{c} - α_{p} I_{p}

\frac{d I_{c}}{d z} = G_{R} I_{p} I_{c} + α_{c} I_{c}

Here I_p represents the Raman pump intensity, I_c represents the chaotic probe light, G_R is the gain coefficient of Raman amplification (0.41W⁻¹km⁻¹). By solving Eqs. (2) and (3) with shooting method, the distribution of probe intensity I_c(z) along fiber can be obtained. When assuming I_p = 0, we could obtain probe intensity distribution without Raman amplification I^’_c(z). Accumulated Raman gain distribution along fiber g(z) is calculated by I_c(z)/I^’_c(z), showing as the blue dotted curve in Fig. 3(a). It can be seen that, with co-pumping Raman amplification, the photons reflected (by either Rayleigh or fiber-fault) at a further location will experience larger accumulated gain. Here, for the convenience of next step, we assume the fiber length as 100km, i.e. Z_max = 100km. Based on these assumptions, we calculate the ratio of fiber-fault-reflected signal power to the Rayleigh-backscattered power under Raman amplification as OSNR_Raman:

O S N R_{R a m a n} = \frac{R \cdot I_{c} (Z_{\max}) \cdot g (Z_{\max}) \cdot e^{- α_{c} Z_{\max}}}{\int_{0}^{Z_{\max}} I_{c} (z) \cdot g (z) \cdot ε \cdot e^{- α_{c} z} d z}

By choosing the parameters as typical values I_p = 26dBm and I_c = −5.5dBm, we could obtain that OSNR will increase to −6.6dB, which means a 10.6dB enhancement compared with the case without Raman amplification (−17.2dB). According to the above analysis, it can be concluded that Raman amplification is able to enhance the OSNR and therefore extend the sensing range of correlation OTDR significantly.

Furthermore, increasing data acquisition duration is another way to enhance system performance. In correlation OTDR, the cross-correlation procedure is essentially matched filtering. Assuming x(t) is the signal, the maximum achievable SNR can be expressed as [20]:

S N R = \frac{{\int_{- \infty}^{+ \infty} | x (t) |}^{2} d t}{N_{0}}

where N₀ is power spectral density of white noise. On the other hand, the energy of signal x(t) is as

E = {\int_{- \infty}^{+ \infty} | x (t) |}^{2} d t

. Finally, we could obtain that the maximum achievable SNR = E/N₀. It is clear that the SNR basically varies linearly with the energy of the reflected signal ignoring changes of N₀. The possible influences could come from the variation of acquisition duration or point attenuation. If the acquisition duration is increased, SNR of correlation OTDR would be increased linearly. As the verification, we demonstrate a 1km sensing range experiment without any amplification. Figure 3(b) shows SNR varies with different acquisition duration, and the experiment result is qualitatively in accordance with the theoretical analysis.

4. Experimental demonstration of fiber fault locating with correlation OTDR

Based on above discussions, we establish an experimental system following Fig. 1. The combination of 1550nm fiber Bragg grating (FBG) and a circulator (CIR1) is act as a narrow bandwidth filter. The 3dB bandwidth of the filter is 0.26nm. After port3 of the CIR1, a 1:99 optical coupler is used to split the filtered light into two branches. One branch (1%) is used as the reference light detected by a 1GHz photo-detector, while the other branch (99%) transmission light acts as the chaotic probe light of OTDR. A Raman amplifier is used to enhance the signal quality. The fiber under test is 100km standard single mode fiber (SMF) with 0.18dB/km loss at 1550nm. Because the reflected signal is very weak, we use a 1GHz avalanche photo-detector (APD) to replace the original PD for signal detection. A 50cm FC/PC jumper attached to the end of the 100km SMF with FC/APC connector, both ends of which are used to emulate fiber-faults.

We adjust all the parameters according to above analysis. The pump power of SC is 31.6dBm, the power of 1% branch of the filtered chaotic light is −25.5dBm, while the 99% branch launched into RA is −5.5dBm. The RA power is set to 26dBm. The total reflected light power detected by the APD is −22.1dBm. The signals of the two detectors are simultaneously recorded for 2ms by a multi-channel oscilloscope with 5GHz sampling rate. The recorded data is processed with cross correlation algorithm, and the fiber fault could be located by the correlation trace. Figure 4(a) shows the normalized correlation OTDR trace. The position of the peak (100.34080km) is corresponding to the open end of the FC/PC jumper. The SNR of our experiment result is 16.34dB which is an excellent result considering the fiber length, suggesting that even longer range is achievable without any modification of the setup. As a comparison, we carried out an experiment with a pre-amplifier (EDFA) plus and an optical filter before the detector (with RA off), and the resulting SNR is 9dB lower than the RA case.

Fig. 4 (a) Experimental result of fault at 100.34080km; (b) magnified peaks of the correlation trace.

Download Full Size | PDF

Figure 4(b) shows the magnified correlation peaks. Because the FC/PC jumper has two flat-ends, the magnified correlation trace has two peaks. The two peaks are separated by 51.4cm which represents the jumper length and the result is reasonably accurate. From the magnified trace, we also identify the spatial resolution of the system as 8.2cm according to FWHM of the correlation peaks.

Although we only demonstrate point to point operation, but one can simply apply the proposed source to any other correlation OTDRs utilized in WDM PON. For example, we could replace the chaotic source reported in Ref [6]. with our proposed SC source. The WDM in Ref [6]. could slice the SC source into many channels, and for each channel the operation principle is the same as a point to point case.

5. Conclusion

In summary, a long-range and high-precision correlation OTDR based on all-fiber SC source is proposed. The temporal properties of source are investigated, showing that the chaotic radiation of the source is a promising for applications such as correlation OTDR. The key factors limiting the sensing range of correlation OTDR are analyzed, including integral effect of weak Rayleigh backscattering and the fiber loss. Based on the analysis, an instruction about how to achieve an ultra-long range correlation OTDR is presented. The key to extend sensing range is improving OSNR of correlation OTDR, while distributed amplification and increasing acquisition duration are two viable approaches. Finally, we realize a centimeter level spatial resolution and 100km range fiber-fault-locating based on distributed Raman amplification, which is a significant improvement compared to the conventional OTDR.

Acknowledgments

This work is supported by Natural Science Foundation of China (61205048, 61290312), Research Fund for the Doctoral Program of Higher Education of China (20120185120003), and PCSIRT (IRT1218), and the 111 project (B14039), and ERC Ultralaser project 267763. The authors also thank Prof. Xingwen Yi in UESTC and Chengdu Best Xingbang Technology Limited for providing the oscilloscopes.

References and links

1. A. Uchida, K. Amano, M. Inoue, K. Hirano, S. Naito, H. Someya, I. Oowada, T. Kurashige, M. Shiki, S. Yoshimori, K. Yoshimura, and P. Davis, “Fast physical random bit generation with chaotic semiconductor lasers,” Nat. Photonics 2(12), 728–732 (2008). [CrossRef]

2. A. Argyris, D. Syvridis, L. Larger, V. Annovazzi-Lodi, P. Colet, I. Fischer, J. García-Ojalvo, C. R. Mirasso, L. Pesquera, and K. A. Shore, “Chaos-based communications at high bit rates using commercial fibre-optic links,” Nature 438(7066), 343–346 (2005). [CrossRef] [PubMed]

3. F. Y. Lin and J. M. Liu, “Chaotic lidar,” IEEE J. Sel. Top. Quantum Electron. 10(5), 991–997 (2004). [CrossRef]

4. Y. Takushima and Y. C. Chung, “Optical reflectometry based on correlation detection and its application to the in-service monitoring of WDM passive optical network,” Opt. Express 15(9), 5318–5326 (2007). [CrossRef] [PubMed]

5. Y. C. Wang, B. J. Wang, and A. B. Wang, “Chaotic correlation optical time domain reflectometry utilizing laser diode,” IEEE Photon. Technol. Lett. 20(19), 1636–1638 (2008). [CrossRef]

6. L. Xia, D. Huang, J. Xu, and D. Liu, “Simultaneous and precise fault locating in WDM-PON by the generation of optical wideband chaos,” Opt. Lett. 38(19), 3762–3764 (2013). [CrossRef] [PubMed]

7. A. Wang, N. Wang, Y. Yang, B. Wang, M. Zhang, and Y. Wang, “Precise fault location in WDM-PON by uitilizing wavelength tunable chaotic laser,” J. Lightwave Technol. 30(21), 3420–3426 (2012). [CrossRef]

8. H. K. Shim, K. Y. Cho, Y. Takushima, and Y. C. Chung, “Correlation-based OTDR for in-service monitoring of 64-split TDM PON,” Opt. Express 20(5), 4921–4926 (2012). [CrossRef] [PubMed]

9. L. Sandstrom, D. Joffe, G. Bekken, J. Brooks, K. Schneider, and R. Goodson, “High performance, in-service correlation OTDR,” in Optical Fiber Communication Conference, (Optical Society of America, 2013), paper OW3G. 3. [CrossRef]

10. J. Ohtsubo, “Chaos synchronization and chaotic signal masking in semiconductor lasers with optical feedback,” IEEE J. Quantum Electron. 38(9), 1141–1154 (2002). [CrossRef]

11. T. B. Simpson, J. M. Liu, A. Gavrielides, V. Kovanis, and P. M. Alsing, “Period-doubling route to chaos in a semiconductor laser subject to optical injection,” Appl. Phys. Lett. 64(26), 3539–3541 (1994). [CrossRef]

12. K. E. Callan, L. Illing, Z. Gao, D. J. Gauthier, and E. Schöll, “Broadband chaos generated by an optoelectronic oscillator,” Phys. Rev. Lett. 104(11), 113901 (2010). [CrossRef] [PubMed]

13. G. L. Shentu, Q. C. Sun, X. Jiang, X. D. Wang, J. S. Pelc, M. M. Fejer, Q. Zhang, and J. W. Pan, “217 km long distance photon-counting optical time-domain reflectometry based on ultra-low noise up-conversion single photon detector,” Opt. Express 21(21), 24674–24679 (2013). [CrossRef] [PubMed]

14. M. G. Herráez, S. M. López, P. Corredera, P. Corredera, M. L. Hernanz, and P. R. Horche, “Supercontinuum generation using a continuous-wave Raman fiber laser,” Opt. Commun. 226(1-6), 323–328 (2003). [CrossRef]

15. A. E. El-Taher, J. D. Ania-Castañón, V. Karalekas, and P. Harper, “High efficiency supercontinuum generation using ultra-long Raman fiber cavities,” Opt. Express 17(20), 17909–17915 (2009). [CrossRef] [PubMed]

16. A. K. Abeeluck, C. Headley, and C. G. Jørgensen, “High-power supercontinuum generation in highly nonlinear, dispersion-shifted fibers by use of a continuous-wave Raman fiber laser,” Opt. Lett. 29(18), 2163–2165 (2004). [CrossRef] [PubMed]

17. H. Wu, Z. Wang, M. Fan, L. Zhang, W. Zhang, and Y. Rao, “Role of the mirror’s reflectivity in forward-pumped random fiber laser,” Opt. Express 23(2), 1421–1427 (2015). [CrossRef] [PubMed]

18. G. P. Agrawal, Nonlinear Fiber Optics (Academic, 5th edition, 2012).

19. Z. N. Wang, J. J. Zeng, J. Li, M. Q. Fan, H. Wu, F. Peng, L. Zhang, Y. Zhou, and Y. J. Rao, “Ultra-long phase-sensitive OTDR with hybrid distributed amplification,” Opt. Lett. 39(20), 5866–5869 (2014). [CrossRef] [PubMed]

20. M. A. Richards, Fundamentals of Radar Signal Processing (McGraw-Hill, 2005).

Long-range and high-precision correlation optical time-domain reflectometry utilizing an all-fiber chaotic source

Abstract

1. Introduction

2. Characteristics of the supercontinuum source

3. Discussions for performance enhancement of correlation OTDR

4. Experimental demonstration of fiber fault locating with correlation OTDR

5. Conclusion

Acknowledgments

References and links

Cited By

Figures (4)

Equations (5)

Optics Express