Open Access
Add to BibSonomy
Abstract
We propose and demonstrate a novel approach for measuring the modal attenuation of the splice loss of a purely LP11 mode group by using a Rayleigh-based OTDR with a dynamic modal crosstalk (XT) suppression technique for few-mode fibers (FMFs). With the proposed approach, the Brillouin loss interaction with a Brillouin Stokes beam co-propagating with the OTDR probe removes the modal XT caused at the modal conversion point and suppresses the accumulated modal XT that is detected. A preliminary experiment is demonstrated using spliced FMFs with a core-offset. Experiments revealed that the proposed technique can accurately measure the splice loss variations of a purely LP11 mode group.
© 2019 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Space division multiplexing transmission [1] based on multi-core fibers (MCFs) and few-mode fibers (FMFs) has attracted attention with a view to solving the problem of the capacity crunch that we face with single mode fiber. The attenuation of all linearly polarized (LP) modes are intrinsically different especially at bends and splice points in FMF. Differential modal attenuation will become a major factor [2] limiting FMF transmission, and this will arise from practical cable installation where the cable is spliced every 1-2 km. Guided LP mode groups have different polarization modes, and the intensity of the electric fields in LPmn (m ≠ 0) modes vary depending on the polarization modes. The characterization of splices in FMF links will be a practical issue as regards quantifying constructed links in terms of the differential modal attenuation since both the core-offset direction and the electric fields excited in such modes result in a range of modal attenuation values even in the same mode group [3,4]. In other words, the splice loss for such a mode group (LPmn (m ≠ 0) modes) might be a good barometer of the splicing quality as well as the modal attenuation range.
Several reflectometry-based methods that use Rayleigh backscattering have been proposed for use with FMFs [5–7]. The longitudinal distribution of the mode coupling ratio along an FMF has been measured using a synchronous multi-channel optical time domain reflectometer (OTDR) [5]. This technique has been used to evaluate the differential modal attenuation between the LP01 and LP11a modes on spooled FMF by separating the power of the LP01 and LP11 modes with a mode MUX/DEMUX at the input fiber end. The related work described in [6] focused on mode coupling at a splice point and reported that local mode coupling near the splice point could be obtained from multi-channel OTDR traces with a matrix-based analytical method. Another approach for measuring differential modal attenuation along FMF can also be found in [7], where a conventional OTDR with a mode filter placed at the input fiber end was utilized to suppress the mode coupling during propagation in the FMF.
However, these techniques were verified under ideal conditions where the distinct mode conversions were relatively uniform or few. With practical links where many splice points are placed serially along FMF links, the accumulated effect of cascaded mode conversions would be more complex. This may affect not only an OTDR probe propagating in the forward direction, but also a Rayleigh backscattered signal coupled into several LP modes during the scattering process [8–10], which propagates in the backward direction. This may result in an underestimation of the modal attenuation at a splice point for LPmn (m ≠ 0) modes. Due to the use of mode selective functions (i.e. a mode coupler or mode filter) employed at the input fiber end, it is difficult to completely handle the accumulated modal crosstalk (XT) in the measurement.
In this paper, we propose and demonstrate a novel modal attenuation measurement technique with a Rayleigh-based OTDR employing a dynamic modal XT suppression method [11] that can focus solely on the splice loss of the LP11 mode group in two-mode FMF. The dynamic modal XT suppression method, which is realized by employing the Brillouin loss interaction between modal XT and a Brillouin Stokes beam co-propagating in the FMF, removes the accumulation of the modal XT at the mode conversion point. A preliminary experiment is demonstrated with an FMF link with cascaded splice points with a core-offset. The results revealed that the splice loss variations of the LP11 mode group were accurately measured and agreed well with those obtained with the transmission method.
2. Principle
Figure 1 shows a schematic illustration of a modal attenuation measurement performed using a Rayleigh-based OTDR with a dynamic modal XT suppression method. Our aim here is to measure purely individual mode groups by suppressing the mode XT that occurs during the Rayleigh backscattering process. For simplicity, we here describe the principle of two-mode FMF and consider the characterization of modal attenuation in the LP11 mode, which is normally larger than that in the LP01 mode.
Fig. 1 Schematic illustration of modal attenuation measurement by Rayleigh-based OTDR with dynamic modal crosstalk suppression method.
A pulsed LP11 probe and a Brillouin Stokes LP01 beam modulated at f1 and f2 ( = f1 - fB) as shown in Fig. 1 are launched from a single side in an FMF. In the FMF, Rayleigh backscattered light generated by the testing probe pulse couples with several LP modes and propagates backwards to the detector side. Then, Rayleigh backscattered lights in the desired LP11 mode selected by the mode MUX are analyzed in the time domain to obtain the LP11 attenuation distributions. However, the observed LP11 attenuation is affected by the Rayleigh XT of the LP01 mode since LP01 to LP11 mode conversion occurs during backward propagation along the FMF.
When launching the Brillouin Stokes beam to suppress the dynamic XT in FMF, these Rayleigh backscattered lights interact with the Brillouin Stokes beam, and the desired Rayleigh signal power P11 of the LP11 mode and the Rayleigh XT power P01 of the LP01 mode decrease because of the Brillouin loss interaction in the collision area of the Brillouin Stokes beam along the FMF. These Rayleigh power changes are given by [12]
Here gBi-j is the modal Brillouin loss of the LPi and LPj (i, j = 01 or 11) modes, and α01 and α11 are the attenuation coefficients of the LPi and LPj modes, respectively. In Eqs. (1) and (2), the Brillouin loss intensities are different as a result of the Brillouin frequency shift (BFS) of the modal Brillouin interaction due to it being set at a frequency offset of fB. Normally, the intra-modal BFS of LP01, LP11 and the inter-modal BFS of LP01/LP11 are different [11]. The Rayleigh XT power can be suppressed with Brillouin Stokes beam by appropriately setting the intra-modal BFS of the LP01 mode. The Rayleigh XT power is converted to a Brillouin Stokes beam whose optical frequency is different from the Rayleigh optical frequency during Brillouin loss interaction and is cut off as the Brillouin Stokes frequency at the optical bandpass filter (BPF) on the detector side. Therefore, the Brillouin Stokes method can suppress Rayleigh XT as dynamic modal crosstalk suppression. Note that the Rayleigh scattering of Brillouin Stokes can also be used to suppress the crosstalk of a transmitted probe pulse by launching a Brillouin Stokes beam before the probe pulse.As we mentioned above, we developed a modal attenuation measurement method that we perform using a Rayleigh-based OTDR with a Brillouin Stokes beam as a dynamic modal crosstalk suppression method that can provide the splice loss of a purely LP11 mode group along an FMF link. The above dynamic crosstalk suppression method has a great advantage in that crosstalk can be suppressed before the position at which it occurs in a distributed FMF since normally undesired Rayleigh XT, which causes measurement error, changes to the mode to be measured at the crosstalk occurrence position.
3. Experimental setup
Figure 2(a) shows the configuration of a Rayleigh based OTDR with a dynamic crosstalk suppression method. We used a distributed feedback laser diode (DFB-LD), whose linewidth has 1 - 2 MHz. The DFB-LD is divided into two paths; a testing probe and a Brillouin Stokes beam. The testing probe is pulse modulated with an acousto-optic modulator (AOM) and amplified with an EDFA. The Brillouin Stokes beam is frequency modulated with an electro-optic modulator (EOM) to set the frequency offset at the modal BFS and is controlled in terms of injection timing by an AOM. The laser optical frequency is modulated by about 3 GHz with an injection current and suppresses Rayleigh fading as a wavelength averaging scheme [13]. A polarization scrambler (PS) was used to substantialize the depolarized Brillouin suppression along the fiber under test (FUT). Figure 2(b) shows the configuration of a mode MUX/DEMUX. This mode MUX/DEMUX with a phase plate for launching the LP11 mode was used to change the rotation angle of the excited LP11 mode. A test probe and a Brillouin Stokes beam were launched into the LP11a and LP01 ports, respectively. The LP11 Rayleigh signals from the FUT were selected from the LP01 beam with the mode MUX and from the reflected Brillouin Stokes light with an optical filter. The LP11 Rayleigh signals were detected and then digitized with a 12-bit data acquisition (DAQ) system that had a 250 MSa/s sampling speed. In the experiment, the peak powers of the launched probe and Brillouin Stokes beams were 5 and 15 dBm, respectively. The pulse width of the probe was set at 50 ns, and the average number was 2^12.
Fig. 2 Configuration of (a) Rayleigh based OTDR with dynamic crosstalk suppression method, (b) mode MUX/DEMUX.
Figure 3 shows an FUT composed of graded index (GI)-2-mode-FMF [14] with two splices with core offsets at 1000 and 1200 m from the input side, respectively.
4. Experimental results
Figure 4 shows the Brillouin loss spectra at (a) 500 m, (b) 700 m, and (c) 1000 m. In Figs. 4(a)–4(c), Rayleigh powers of LP01 were suppressed more than those of the LP11 mode at a frequency offset of 10.77 GHz. The suppressed XT values at 500, 700, and 1000 m were 5.03, 5.88, and 6.67 dB, respectively. The Rayleigh XT powers in Figs. 4(b) and 4(c) were suppressed more than before the crosstalk occurrence position in Fig. 4(a).
Fig. 4 Brillouin loss spectra at (a) 500 m, (b) 700 m, and (c) 1000 m. The black and red lines are Brillouin loss spectra of Rayleigh backscattering measured at the mode MUX/DEMUX port for LP11 and LP01, respectively.
Figure 5 shows experimentally obtained Rayleigh backscattering traces measured at the mode MUX/DEMUX port for (a) LP11 and (b) LP01 by setting a frequency offset of 10.77 GHz. At a position beyond the splice point with a core offset in Figs. 5(a) and 5(b), the Rayleigh backscattering intensities with a Brillouin Stokes beam decreased along the FMF. We found that our proposed method successfully suppressed Rayleigh XT in the vicinity of the occurrence position in the FMF.
Fig. 5 Rayleigh backscattering traces measured with mode MUX/DEMUX port for (a) LP11 and (b) LP01. The black and gray lines, respectively, show Rayleigh traces with and without a Brillouin Stokes beam. Figure 5(c) and 5(d) show enlarged views of the Rayleigh backscattering in Fig. 5(a) and 5(b).
Figure 6 shows modal attenuation measurement results for (a), (c) 1 μm and (b), (d) 2 μm core offsets obtained by changing the angle of the phase plate of the LP11 mode measured for Figs. 6(a), 6(b) with and Figs. 6(c), 6(d) without Brillouin Stokes. The broken and chain lines show the maximum and minimum splice losses, respectively, measured with a transmission method using a two-mode MUX/DEMUX. With the transmission method, we employed about 1-m of FMF cut at the splice point and including the core offset from the previous FUT. In these experiments, the angle of the phase plate was changed 36 times at every 10 °. As shown in Figs. 6(a) and 6(b), the splice loss variations measured with the proposed method overlapped more closely than in the results measured with the transmission method, while the maximum values of the actual splice losses measured with an OTDR without a Brillouin Stokes beam could not be obtained.
Fig. 6 Measured modal attenuation of (a), (c) 1 μm and (b), (d) 2 μm core offset obtained by changing the angle of the phase plate of the LP11 mode measured by (a), (b) with and (c), (d) without Brillouin Stokes. The red and black symbols, respectively, show the splice losses measured by the LP11a and LP11b port of the mode MUX/DEMUX.
Table 1 shows the minimum, maximum, average and central values of the measured splice losses. In these experiments, the splice loss values were obtained 10 times under the same condition as in Fig. 6. The maximum splice values measured with the proposed method were within 0.1 dB of those measured with the transmission method. The average values measured with each method by changing the phase plate angle tended not to match. We assumed that the rotation angle change of the LP11 electric fields at a splice point did not necessarily correspond to the rotation angle change of the phase plate at the injection point. On the other hand, the median values of the proposed and transmission methods were in good agreement since the proposed method can measure the maximum and minimum values.
Table 1. Comparison of LP11 splices loss with and without Brillouin Stokes beam
5. Summary and discussion
We proposed and demonstrated a modal attenuation measurement method using a Rayleigh-based OTDR with dynamic modal crosstalk suppression that can provide the splice loss of a purely LP11 mode group. The dynamic modal crosstalk suppression method, which was realized as Brillouin loss interaction by launching a Brillouin Stokes beam, removed the modal LP01 crosstalk in the vicinity of the generation point in FMF. A preliminary experiment was performed using spliced FMFs with core-offsets. Experiments revealed that the proposed technique can accurately measure the splice loss variations of a purely LP11 mode group.
The splice loss variation measurement allows us to evaluate an FMF transmission line. The maximum, minimum and median loss values of the LP11 mode group, not the average value, might be a good barometer for the quality of splicing for FMF link characterization because the LP11 mode group responds sensitively to the construction conditions. Additionally, the experimental results mismatch with the average value of the LP11 modal attenuation measurement shows the importance of controlling or scrambling all the conditions of the LP11 electric field in FMF.
Dynamic crosstalk suppression by using a Brillouin loss interaction is important for setting and searching for the modal BFS, but BFSs along deployed optical fiber vary because of such factors as changing temperature, residual strain. In this case, the dynamic range may be degraded since the desired Rayleigh signal is decreased due to the matching of the frequency offset and the modal BFS of Rayleigh signal. It may also not be able to effectively suppress the modal crosstalk since the Rayleigh crosstalk is not decreased due to the mismatching of the frequency offset and the modal BFS of Rayleigh crosstalk. Additionally, there is the possibility of mistakenly identifying a loss event at the location of the BFS change. Therefore, as shown in Fig. 4, it is necessary to accurately set and search the frequency difference between a test and a Brillouin Stokes beam to modal BFS at each location. As future work we will investigate the performance of dynamic modal crosstalk suppression affected by modal BFS change.
Nevertheless, we believe that the proposed dynamic modal crosstalk suppression method, which is a unique and attractive feature of a Brillouin Stokes beam, will be a powerful tool for few-mode fiber link evaluation.
References
1. T. Morioka, “New generation optical infrastructure technologies: 'EXAT initiative' towards 2020 and beyond” Proc. OECC2009, FT4 (2009).
2. A. Lobato, F. Ferreira, M. Kuschnerov, D. van den Borne, S. L. Jansen, A. Napoli, B. Spinnler, and B. Lankl, “Impact of mode coupling on the mode-dependent loss tolerance in few-mode fiber transmission,” Opt. Express 20(28), 29776–29783 (2012). [CrossRef] [PubMed]
3. S. Warm and K. Petermann, “Splice loss requirements in multi-mode fiber mode-division-multiplex transmission links,” Opt. Express 21(1), 519–532 (2013). [CrossRef] [PubMed]
4. K. Ozaki, M. Ohashi, H. Kubota, and Y. Miyoshi, “Effective mode field diameter definition and splice loss estimation of LP11 mode in few mode fibers” Asia Communications and Photonics Conference, ATh3A.98 (2014).
5. M. Nakazawa, M. Yoshida, and T. Hirooka, “Measurement of mode coupling distribution along a few-mode fiber using a synchronous multi-channel OTDR,” Opt. Express 22(25), 31299–31309 (2014). [CrossRef] [PubMed]
6. M. Yoshida, T. Hirooka, and M. Nakazawa, “Mode coupling measurement at a splice point between few-mode fibers using a synchronous multi-channel OTDR,” in Optical Fiber Communication Conference, OSA Technical Digest (online) (Optical Society of America, 2016), paper Th1J.4.
7. H. Kubota, Y. Miyoshi, M. Ohashi, T. Mori, T. Matsui, and K. Nakajima, “Mode-dependent loss measurement of a two-mode fiber using a conventional OTDR,” IEICE Commun. Express 5(11), 429–434 (2016). [CrossRef]
8. D. Yu, S. Fu, M. Tang, and D. Liu, “Mode-dependent characteristics of Rayleigh backscattering in weakly-coupled few-mode fiber,” Opt. Commun. 346, 15–20 (2015). [CrossRef]
9. D. Yu, S. Fu, Z. Cao, M. Tang, L. Deng, D. Liu, I. Giles, T. Koonen, and C. Okonkwo, “Characterization of Rayleigh backscattering arising in various two-mode fibers,” Opt. Express 24(11), 12192–12201 (2016). [CrossRef] [PubMed]
10. Z. Wang, H. Wu, X. Hu, N. Zhao, Z. Yang, F. Tan, J. Zhao, Q. Mo, and G. Li, “Rayleigh backscattering in few-mode optical fibers” in Optical Fiber Communication Conference, OSA Technical Digest (online) (Optical Society of America, 2016), paper W4F.6.
11. H. Takahashi, K. Toge, T. Oda, and T. Manabe, “Inter-modal crosstalk manipulation for Brillouin gain measurement based few-mode fiber sensors” Proc. APOS2018, Mon 4 (2018)
12. G. P. Agrawal, Nonlinear Fiber Optics 4th ed. (Springer, 2006).
13. K. Shimizu, T. Horiguchi, and Y. Koyamada, “Characteristics and reduction of coherent fading noise in Rayleigh backscattering measurement for optical fibers and components,” J. Lightwave Technol. 10(7), 982–987 (1992). [CrossRef]
14. L. Gruner-Nielsen, Y. Sun, J. W. Nicholson, D. Jakobsen, K. G. Jespersen, R. Lingle, and B. Palsdottir, “Few mode transmission fiber with low DGD, low mode coupling, and low loss,” J. Lightwave Technol. 30(23), 3693–3698 (2012). [CrossRef]
