A configuration of hybrid distributed Raman amplification (H-DRA), that is formed by incorporating a random fiber laser (RFL) based 2nd-order pump and a low-noise laser-diode (LD) based 1st-order pump, is proposed in this paper. In comparison to conventional bi-directional 1st-order DRA, the effective noise figure (ENF) is found to be lower by amount of 0 to 4dB due to the RFL-based 2nd-order pump, depending on the on-off gain, while the low-noise 1st-order Raman pump is used for compensating the worsened signal-to-noise ratio (SNR) in the vicinity towards the far end of the fiber and avoiding the potential nonlinear impact induced by excess injection of pump power and suppressing the pump-signal relative intensity noise (RIN) transfer. As a result, the gain distribution can be optimized along ultra-long fiber link, due to combination of the 2nd-order RFL and low-noise 1st-order pumping, making the transmission distance be extended significantly. We utilized such a configuration to achieve ultra-long-distance distributed sensing based on Brillouin optical time-domain analysis (BOTDA). A repeater-less sensing distance record of up to 154.4km with 5m spatial resolution and ~ ± 1.4°C temperature uncertainty is successfully demonstrated.
© 2013 Optical Society of America
Random fiber laser (RFL) [1–16], which composes of randomly distributed Rayleigh scatters and fiber-based Raman gain medium, has attracted much attention since its first experimental demonstration by S. K. Turitsyn et al. . Unlike classical laser that requires two end mirrors or loop cavity to provide feedback and generate a series of resonance modes, the lasing of RFL is based on the multiple scattering of photons in a disordered amplifying medium. As the scattered photons do not return to their initial location periodically, the emission of RFL exhibits continuous spectra without any resonant modes [1, 2]. RFL also has different lasing dynamics in contrast to classical lasers [1, 2]. A variety of RFL structures have been proposed, such as those full-open [1, 4] and half-open cavities [3, 13]. Note that compared with the full-open cavity, the half-open cavity structure can reduce the lasing threshold effectively [3, 13]. The multi-wavelength lasing [5, 6], widely tunable capability , cascaded random emission at higher-order Stokes peaks [7, 13, 15], and narrow-band random lasing generation  have also been realized. A variety of physical properties of RFL, such as spectral features , power optimization  and distribution , and relative intensity noise (RIN) transfer , have been explored.
As for the applications, the unique features of RFL make it very promising in the fields of fiber-optic communication  and sensing [18–23]. Recently, we have studied the characteristics of RFL-based 2nd-order distributed Raman amplification (DRA) over 93km standard single-mode fiber (SMF) . In particular, the RFL-based DRA with forward pumping possesses lower effective noise figure (ENF) compared to conventional bi-directional Raman amplification. Also, RFL is uniquely insensitive to ambient temperature and vibration, making it a new way to realize stable optical transmission . Very recently, the authors introduced the concept of RFL-based DRA for signal-to-noise ratio (SNR) enhancement of Brillouin optical time-domain analysis (BOTDA [24–29]), and preliminary demonstration of 122km temperature sensing was carried out . However, further transmission/sensing distance extension is strongly desired but mainly restricted by the relatively high pump power requirement (multiple-Watt level typically) and potential nonlinear impact, due to lower 2nd-order pumping efficiency , resulting from the quite weak Rayleigh back-scattering coefficient of SMF (~4.5 × 10−5 km−1) .
Here, a low-noise laser-diode (LD) based 1st-order pump is incorporated into the RFL-based DRA to form a hybrid pumping structure (H-DRA). Such an optimized pumping scheme simultaneously possesses lower ENF dominated by RFL pump and higher pumping efficiency which enables the compensation of the worsened SNR in the vicinity towards the far end of the fiber. Furthermore, by using such an improved H-DRA, a substantial sensing distance extension of up to 154.4km is successfully demonstrated for BOTDA, which is 32.4km longer than that we reported previously .
2. H-DRA incorporating RFL and LD pumps
2.1 Experimental setup
Figure 1 shows the schematic diagram of the operation principle and experimental setup for the proposed H-DRA configuration. The emission of random lasing at 1454nm is initiated by random Rayleigh distributed feedback along 154.4 km SMF, pumped by a 2nd-order pump at 1366nm (high-power fiber Raman laser (FRL)). The 2nd-order pump is coupled into the transmission span via a 1366/1550nm wavelength-division-multiplexer (WDM). Point feedback formed by a fiber Bragg grating (FBG) at 1454nm with ~95% reflectivity is used to reduce the random lasing threshold [3, 13]. Note that for this half-open cavity structure with only one FBG, the physical trait of random lasing is still ensured since there is no closed facet-end feedback [3, 13].
To alleviate the higher pumping power requirement for RFL-based DRA and extend the transmission/sensing reach, another 1st-order pump at 1455nm is incorporated to form the H-DRA via another 1455/1550nm WDM. The 1st-order pump consists of two low-noise LDs with the same output power and multiplexed by a polarization beam combiner (PBC). In this way, the polarization-dependent gain is reduced to be negligible . As a result, a higher pumping efficiency is ensured to avoid the potential nonlinear impact induced by the injection of larger pump level. This is very meaningful for long-distance optical transmission/sensing, since the required on-off gain and pump power are increased with the increased fiber length to fully compensate the fiber loss.
In the experiment, a continuous wave (CW) launched from a distributed-feedback laser-diode (DFB-LD) at 1550nm is injected into the fiber through the left WDM. The gain distribution over the fiber is acquired by an optical time-domain reflectometry (OTDR). The on-off gain and ENF arising from amplified spontaneous emission (ASE) noise [17, 30] are recorded by an optical spectrum analyzer (OSA). As shown in Fig. 1, to reduce the parasitic back-reflection and protect experimental devices, before the injection of high-power pumps at 1366 and 1455nm into the transmission system, their lasing outputs have been handled by optical isolators. An optical isolator is also used in front of the OSA to further reduce parasitic back-reflection. In this way, the random nature of 1454nm lasing is ensured. In the following, the general amplification performances of the proposed H-DRA are presented by comparison with the conventional bi-directional 1st-oder DRA.
2.2 Fundamental amplification properties
Figure 2 shows the measured on-off gain as function of input power of the primary pump for different pumping configurations. For the proposed H-DRA, the 1st-order pump power is fixed at 26dBm to provide ~10dB on-off gain, while the majority of gain is provided by adjusting the power of 1366nm pump. From this diagram, the gain of the proposed H-DRA shows a threshold feature near ~32dBm pump power. As ~10dB loss is compensated by the 1st-order pump for the hybrid pumping, the required total power of the primary pump for transparency transmission over 154.4km fiber (~30.9dB on-off gain) is ~34.4dBm, which is comparable to the case of single RFL pumping without additional 1st-order pump over 93km fiber reported in . Therefore, hybrid pumping is highly efficient compared to single RFL pumping.
Figure 3 shows the gain distribution under various on-off gains for different pumping configurations. It can be seen that, the peak gain is pushed to the position of ~40km when close to transparency transmission for the hybrid pumping; however, the peak gain is located at ~20km at the similar on-off gain for the bi-directional 1st-order pumping. This distinction is a direct result of 2nd-order random laser pumping for the H-DRA, and is beneficial for achieving higher SNR and measurement accuracy towards the far end of fiber for ultra-long-distance distributed sensing applications.
Figure 4(a) describes the variation of ENF with the on-off gains for different pumping configurations. It is noteworthy that a critical point occurs at ~23dB on-off gain. Above this point, the ENF of the proposed hybrid pumping is less than that of bi-directional 1st-order pumping by 0 to 4dB, depending on the specific on-off gain; however, below the critical point, the opposite result is drawn. As been mentioned, for hybrid pumping, the 1st-order pump component offers ~10dB gain, therefore, the gain is dominated by the 1st-order pump for lower 1366nm pump, which actually form the 1st-order backward pumping. In this case, the generation of ASE noise co-propagating with signal is mainly concentrated near the output end. Hence, the larger ENF is shown as a consequence of lower loss experienced for ASE . On the other hand, if the power of 1366nm pump is increased to boost the random laser considerably, the gain is then dominated by 2nd-order random laser pump. In this case, the generation of ASE co-propagating with signal is mainly concentrated near the input end, leading to the lower ENF, as the produced ASE experiences the full-length loss over the transmission span. From the practical point of view, the signal enhancement often operates near the transparency point (~30.9dB) that is within a regime where the on-off gain is higher than a critical value (~23 dB). As shown in Fig. 4(a), in this regime, the ENF of the proposed H-DRA is more preferable.
For the same signal input power, the nonlinear impairment of transmission channel can be characterized by the ratio of path-averaged signal power when the primary pump is on (Pave) and off (Pref), respectively [17, 30]. The result is depicted in Fig. 4(b). A similar critical feature is observed. For the on-off gain above the critical point (~21dB), larger path-averaged power is obtained for H-DRA. This is attributed to its larger gain value, as shown in Fig. 3(a). In other words, as shown in Fig. 4(b), the nonlinear impairment of H-DRA is worse than that of conventional bi-directional 1st-order pumping scheme in the regime where the on-off gain is above a critical value (~21dB). The above discussions imply that, the H-DRA is advantageous where the ASE noise accumulation dominates the system performance (for an instance, when the channel input power is enough low to make the nonlinear impairment be negligible). However, several techniques can be used for reducing the impact of nonlinearity in long-distance optical transmission, such as the use of non-zero dispersion-shifted fiber (NZ-DSF) or large effective area fiber (LEAF), periodic dispersion management by combining normal and anomalous group velocity dispersion (GVD), and advanced modulation format (e.g., differential phase shift keying (DPSK) is more tolerant to cross-phase modulation (XPM)), etc . The nonlinear effect also should not hinder its distributed sensing applications in most cases. As for instance in BOTDA sensors, the optical pulse coding (OPC) is an efficient approach to increase the SNR by launching a sequence of optical pulses under lower peak power of Brillouin pump [18, 24, 29], consequently, the system error induced by nonlinear spectral expansion could be reduced.
3. Application of H-DRA to ultra-long-distance distributed sensing
3.1 System design
In the following, H-DRA formed by RFL and 1st-order pump is employed to maximize the reach of distributed sensing using BOTDA. The experimental implementation is described in Fig. 5. The CW beam at 1550nm from a DFB-LD is split into two parts using an optical splitter. The 10% portion passes through an electro-optic modulator (EOM) to generate two side-bands, by which the carrier is suppressed down to −30dB via suitably adjusting the direct current (DC) bias. After amplification by an erbium-doped fiber amplifier (EDFA1), the probe light is directed to an acoustic-optic frequency shifter (AOFS), and then routed to sensing fiber via a variable optical attenuator (VOA1), optical isolator (ISO1) and 1550nm port of WDM2. Another 90% portion is used as the Brillouin pump. Before the operation of Simplex-coding [18, 24, 29], the amplification is performed by EDFA2 to avoid the system error caused by gain saturation due to a sequence of launched long pulses in tens of micro-second scale. The coding is then realized by an acoustic-optic modulator (AOM) with 45dB extinction ratio, driven by an arbitrary waveform generator (AWG). The coding length is 255bits, equivalent to ~9.1dB SNR enhancement [18, 24, 29]. The pulse-width is 50ns, corresponding to 5m spatial resolution. In order to overcome the impact of ~10ns phonon lifetime, return-to-zero (RZ) [18, 24, 29] pattern with 150ns period and 33.3% duty-cycle is used. After a polarization-scrambler (PS) used to reduce the polarization-dependent gain fluctuation, the Brillouin pump is then injected into the fiber through a VOA2, optical circulator (CIR1) and WDM1. The receiver consists of a pre-amplifier (EDFA3), two fiber Bragg gratings (FBG2, FBG3), two circulators (CIR2, CIR3), VOA3, 125MHz photo detector (PD), and fast data acquisition (DAQ) card with 100MSa/s sampling rate. To avoid the SNR reduction caused by saturation of PD, the anti-Stokes component with short-wavelength is used as the sensing signal (i.e., Brillouin loss spectrum), as shown in Fig. 5.The Brillouin gain spectrum (BGS) is then reconstructed by simply inverting the acquired Brillouin loss spectrum. The transient gain saturation of EDFA3 that would lead to the potential decoding error is caused by the residual lasing light at 1454nm, the Ralyleigh backscattering of Brillouin pump, and more importantly, the amplified Stokes component that experiences the Brillouin gain. Elimination of these detrimental effects is realized by a narrow-band (<0.1nm) FBG2 placed before EDFA3 and centered at anti-Stokes wavelength. The residual ASE noise is further filtered out by FBG3.The sensing fiber with 154.4km length is composed of four spools with slightly different Brillouin frequency shift (BFS) of ~10.882, 10.870, 10.867, and 10.870GHz, respectively. The splicing points of the fiber spools are positioned at ~11.965, 60.389, and 106.772km, respectively. The BFS difference is well within the full-width at half-maximum (FWHM) (~35-60MHz) of BGS, ensuring the maximized Brillouin interaction. It should be noted that, the frequency shift value (200MHz) of AOFS is exactly same with AOM, thus, symmetrical dual side-band of probe wave relative to Brillouin pump is formed. Such a symmetric spectral structure is helpful for reducing the non-local effect derived from pump depletion , due to the balanced interaction for Brillouin pump through gain and loss processes .
3.2 154.4km sensing experiment
Figure 6(a) shows the decoded BGS as function of distance and frequency shift towards the far end of sensing fiber in the range of 110-154.4km. During the measurement, a fraction of fiber with 5m long at the terminal (154.334km) is heated by a temperature controlled chamber. The temperature variation of the hot-spot is 30°C, and the room temperature is 26.2°C. The acquired Brillouin trace is averaged by 16 times before decoding process, equivalent to the total averaging times of 4080. Since the gain is pushed inside the fiber more deeply due to RFL pumping, the attenuated Brillouin pump in the vicinity of the fiber end is compensated by 1st-order pump, and the issue of RIN transfer is perfectly addressed by proper propagation direction selection (the probe beam is co-propagating with the low-noise 1st-order pump , and counter-propagating with the 1366nm pump to suppress the RIN transfer via the larger group velocity walk-off ), the higher SNR over the fully fiber is preserved. Besides, a shift of peak BFS is clearly demonstrated at the 5m hot-spot. Figure 6(b) shows the decoded BGS at various positions (10, 30, 50, 70, 90, 110, 130, and 150km). The FWHM of BGS in each case is shown in the inset of (b). Thanks to the optimized power levels of Brillouin pump (~2.5dBm) and Raman pumps (33.7dBm at 1366nm and 26dBm at 1455nm respectively for transparency transmission), the spectral broadening caused by self-phase modulation (SPM) and modulation instability (MI) is considerably suppressed within the range of ~35-60MHz. It is worth mentioning that, both of SNR and width of BGS play significant roles for the measurement uncertainty . Another feature is that, the multiple-peak structure and BGS distortion induced by non-local effect is not found along the entire fiber. This is the direct result of suppressed Brillouin pump depletion by optimized probe power (~-16dBm) and Raman pump powers, together with the use of symmetrical dual-sideband modulation  assisted by an AOFS.
Figure 7(a) depicts the temperature distribution along sensing fiber by subtracting the extracted peak BFS from Lorentz fitting before and after the heating procedure. By calculation of the standard deviation, the worst measurement uncertainty at the last section of sensing fiber (~20km) is ~ ± 1.4°C. It should be noted that, a degradation of measurement uncertainty that emerges at the position of 106.772km is mainly caused by the splicing loss of ~1.3dB, as shown in Fig. 3. The details of extracted temperature distribution around 5m hot-spot are further displayed in Fig. 7(b). As been expected, ~30°C temperature variation observed is in good agreement with prediction. From the FWHM of temperature distribution around 5m hot-spot, the real spatial resolution of ~5m is clearly revealed.
4. Discussion and conclusion
Previously reported DRA using only RFL pumping have turned out to possess lower ENF , however, this scheme suffers from very low 2nd-order pumping efficiency  due to the quite weak Rayleigh backscattering coefficient of the fiber , leading to the higher pump power requirement and potential nonlinear impact for ultra-long-distance optical transmission/ sensing. In this work, H-DRA scheme incorporating RFL and low-noise 1st-order pumps is proposed to address the above issue. The amplification properties (including gain, ENF, and nonlinear impairment) have been investigated experimentally and compared to conventional bi-directional 1st-order pumping. Results show that, the proposed H-DRA retains major benefit of lower ENF (0 to 4dB) with the growth of on-off gain (>23dB)) due to RFL pumping. Simultaneously, the worsened SNR towards the far end of fiber is compensated more efficiently, thus eliminating its potential nonlinear impact induced by excess injection of pump power. Besides, 2nd-order pumping provided by RFL makes the gain pushed inside the fiber more deeply (peak gain occurs at ~40km), greatly extending the range with higher SNR.
As a proof-of-concept, ultra-long-distance distributed sensing with BOTDA over 154.4km fiber with 5m spatial resolution and ± 1.4°C measurement uncertainty has been demonstrated by means of H-DRA, allowing for a fully resolving of a 5m hot-spot at the far end of sensing fiber. This distance, compared to our preliminary sensing result , gives a big upgrade of ~32.4km, while the similar magnitudes of spatial resolution and measurement uncertainty are preserved. Hence, the proposed H-DRA offers a new approach to realize ultra-long-distance repeater-less optical transmission/sensing with improved SNR in the regime where the on-off gain is above a critical value of ~23dB (see Fig. 4(a)).
The authors thank the reviewers for their helpful comments. This work is supported by the National Nature Science Foundation of China (NSFC) under grants No. 61290312, 61205079, 61205048, and 61106045, the Program for Changjiang Scholars and Innovative Research Team in University (PCSIRT), the Construction Plan for Scientific Research Innovation Teams of Universities in Sichuan Province under grant No. 12TD008, and the 251 Talents Program of Sichuan Normal University.
References and links
1. S. K. Turitsyn, S. A. Babin, A. E. El-Taher, P. Harper, D. V. Churkin, S. I. Kablukov, J. D. Ania-Castañón, V. Karalekas, and E. V. Podivilov, “Random distributed feedback fiber laser,” Nat. Photonics 4(4), 231–235 (2010). [CrossRef]
2. A. A. Fotiadi, “An incoherent fibre laser,” Nat. Photonics 4(4), 204–205 (2010). [CrossRef]
3. D. V. Churkin, S. A. Babin, A. E. El-Taher, P. Harper, S. I. Kablukov, V. Karalekas, J. D. Ania-Castañón, E. V. Podivilov, and S. K. Turitsyn, “Raman ﬁber lasers with a random distributed feedback based on Rayleigh scattering,” Phys. Rev. A 82(3), 033828 (2010). [CrossRef]
4. S. A. Babin, A. E. El-Taher, P. Harper, E. V. Podivilov, and S. K. Turitsyn, “Tunable random fiber laser,” Phys. Rev. A 84(2), 021805 (2011). [CrossRef]
5. A. E. El-Taher, M. Alcon-Camas, S. A. Babin, P. Harper, J. D. Ania-Castañón, and S. K. Turitsyn, “Dual-wavelength, ultralong Raman laser with Rayleigh-scattering feedback,” Opt. Lett. 35(7), 1100–1102 (2010). [CrossRef] [PubMed]
6. A. E. El-Taher, P. Harper, S. A. Babin, D. V. Churkin, E. V. Podivilov, J. D. Ania-Castanon, and S. K. Turitsyn, “Effect of Rayleigh-scattering distributed feedback on multiwavelength Raman fiber laser generation,” Opt. Lett. 36(2), 130–132 (2011). [CrossRef] [PubMed]
7. I. D. Vatnik, D. V. Churkin, S. A. Babin, and S. K. Turitsyn, “Cascaded random distributed feedback Raman fiber laser operating at 1.2 μm,” Opt. Express 19(19), 18486–18494 (2011). [CrossRef] [PubMed]
8. A. R. Sarmani, M. H. Abu Bakar, A. A. A. Bakar, F. R. M. Adikan, and M. A. Mahdi, “Spectral variations of the output spectrum in a random distributed feedback Raman fiber laser,” Opt. Express 19(15), 14152–14159 (2011). [CrossRef] [PubMed]
10. D. V. Churkin, A. E. El-Taher, I. D. Vatnik, J. D. Ania-Castañón, P. Harper, E. V. Podivilov, S. A. Babin, and S. K. Turitsyn, “Experimental and theoretical study of longitudinal power distribution in a random DFB fiber laser,” Opt. Express 20(10), 11178–11188 (2012). [CrossRef] [PubMed]
12. Y. J. Rao, “OFS research over the last 10 years at CQU & UESTC,” Photon. Sens. 2(2), 97–117 (2012). [CrossRef]
13. W. L. Zhang, Y. J. Rao, J. M. Zhu, Z. X. Yang, Z. N. Wang, and X. H. Jia, “Low threshold 2nd-order random lasing of a fiber laser with a half-opened cavity,” Opt. Express 20(13), 14400–14405 (2012). [CrossRef] [PubMed]
15. W. L. Zhang, Y. Y. Zhu, Y. J. Rao, Z. N. Wang, X. H. Jia, and H. Wu, “Random fiber laser formed by mixing dispersion compensated fiber and single mode fiber,” Opt. Express 21(7), 8544–8549 (2013). [CrossRef] [PubMed]
17. X. H. Jia, Y. J. Rao, F. Peng, Z. N. Wang, W. L. Zhang, H. J. Wu, and Y. Jiang, “Random-lasing-based distributed fiber-optic amplification,” Opt. Express 21(5), 6572–6577 (2013). [CrossRef] [PubMed]
18. X. H. Jia, Y. J. Rao, Z. N. Wang, W. L. Zhang, Y. Jiang, J. M. Zhu, and Z. X. Yang, “Towards fully distributed amplification and high-performance long-range distributed sensing based on random fiber laser,” OFS 2012, Proc. SPIE 8421(842127), 842127 (2012). [CrossRef]
19. A. M. R. Pinto, O. Frazão, J. L. Santos, and M. Lopez-Amo, “Multiwavelength fiber laser based on a photonic crystal fiber loop mirror with cooperative Rayleigh scattering,” Appl. Phys. B 99(3), 391–395 (2010). [CrossRef]
20. A. M. R. Pinto, O. Frazão, J. L. Santos, M. Lopez-Amo, J. Kobelke, and K. Schuster, “Interrogation of a suspended-core Fabry Perot temperature sensor through a dual wavelength Raman fiber laser,” J. Lightwave Technol. 28(21), 3149–3155 (2010).
21. H. F. Martins, M. B. Marques, and O. Frazão, “Temperature-insensitive strain sensor based on four-wave mixing using Raman fiber Bragg grating laser sensor with cooperative Rayleigh scattering,” Appl. Phys. B 104(4), 957–960 (2011). [CrossRef]
22. A. M. R. Pinto, M. Lopez-Amo, J. Kobelke, and K. Schuster, “Temperature fiber laser sensor based on a hybrid cavity and a random mirror,” J. Lightwave Technol. 30(8), 1168–1172 (2012). [CrossRef]
23. Z. N. Wang, Y. J. Rao, H. Wu, P. Y. Li, Y. Jiang, X. H. Jia, and W. L. Zhang, “Long-distance fiber-optic point-sensing systems based on random fiber lasers,” Opt. Express 20(16), 17695–17700 (2012). [CrossRef] [PubMed]
27. 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] [PubMed]
28. X. H. Jia, Y. J. Rao, Z. N. Wang, W. L. Zhang, Z. L. Ran, K. Deng, and Z. X. Yang, “Theoretical investigations on the non-local effect in a long-distance Brillouin optical time-domain analyzer based on bi-directional Raman amplification,” J. Opt. 14(4), 045202 (2012). [CrossRef]
29. X. Angulo-Vinuesa, M. A. Soto, S. Martin-Lopez, S. Chin, J. D. Ania-Castañon, P. Corredera, E. Rochat, M. Gonzalez-Herraez, and L. Thévenaz, “Brillouin optical time-domain analysis over a 240 km-long fiber loop with no repeater,” OFS 2012, Proc. SPIE 8421, 8421C9 (2012).
30. C. Headly and G. P. Agrawal, Raman Amplifiers in Fiber Optical Communication System (Elsevier, 2005).
31. G. P. Agrawal, Applications of Nonlinear Fiber Optics, 2nd ed. (Academic Press, 2008).