## Abstract

An ideal optical backpropagation (OBP) scheme to compensate for dispersion and nonlinear effects of the transmission fibers is proposed. The scheme consists of an optical phase conjugator (OPC), N spans of dispersion-decreasing fibers (DDFs) and amplifiers, placed at the end of the fiber optic link. In order to compensate for the nonlinear effects of the transmission fibers exactly, the nonlinear coefficient of the backpropagation fiber has to increase exponentially with distance or equivalently the power in the backpropagation fiber should increase exponentially with distance if the nonlinear coefficient is constant. In this paper, it is shown that a combination of DDFs and amplifiers can compensate for the nonlinear effects exactly. An analytical expression for the dispersion profile of the DDF is derived. Numerical simulation of a long haul wavelength division multiplexing (WDM) fiber optic system with the proposed OBP scheme shows that the system reach can be enhanced by 54% as compared to digital backpropagation (DBP).

© 2013 Optical Society of America

## 1. Introduction

The maximum reach of a long haul fiber optic system with advanced modulation formats is mainly limited by fiber nonlinear impairments. The backpropagation techniques can be used to compensate for dispersion and nonlinear effects of the transmission fiber (TF). The compensation schemes can be divided into three types: digital [1–11], optical [12–19], and the combination of both [20]. The optical backpropagation (OBP) has the following advantages/disadvantages over digital backpropagation (DBP). (i) A very large bandwidth (~4 THz) is available for OBP while the bandwidth of the DBP is limited by the bandwidth of the coherent receiver. (ii) DBP requires significant computational resources, especially for wavelength division multiplexing (WDM) system and hence it is currently limited to off-line signal processing. In contrast, OBP provides compensation in real time. (iii) Number of samples per symbol available for DBP is limited by the sampling rate of the analog-to-digital converter (ADC). Although it is possible to do upsampling on the digital signal processor (DSP), it leads to additional computational complexity. However, for OBP, the signal processing is done on the analog optical waveform. (iv) OBP requires a real fiber which has loss. So, amplifiers are needed to compensate for fiber loss in the OBP section which enhances the noise in the system.

In [16,18], an OBP scheme consisting of optical phase conjugation (OPC), dispersion compensation fiber (DCF)/fiber Bragg grating (FBG), and highly nonlinear fiber (HNLF) is investigated. DCF/FBG is used to compensate for dispersion, and HNLF is used to compensate for nonlinearity. The dispersion and nonlinear effects are compensated in a split-step fashion analogous to split-step Fourier scheme (SSFS) used to solve the nonlinear Schrödinger equation (NLSE). Although this technique is quite effective for a single channel, for a WDM system, small step size is required and hence the insertion losses due to DCF/FBG and HNLF increase which limit the transmission performance. In this paper, we investigate the possibility of introducing a single optical device which can exactly compensate for dispersion and nonlinearity. A dispersion-decreasing fiber (DDF) with a specific dispersion profile is found to meet our requirements.

In the proposed scheme, an OPC is placed at the end of the transmission link which is followed by N spans of DDFs where N is the number of TF spans. The DDFs introduce a small amount of losses which are compensated by amplifiers placed in the OBP section. Numerical simulation results show that the OBP with DDF outperforms DBP and midpoint-OPC schemes. The transmission reach of a WDM system can be significantly enhanced using the proposed scheme as compared to linear compensation in the receiver or DBP. We found that the DBP is limited mainly by the sampling rate of the ADC. In this paper, the study is limited to scalar NLSE and it does not fully address the general polarization-multiplexed WDM transmission system.

## 2. OBP theory

The evolution of the optical field envelope in a fiber optic link is described by the NLSE:

*β*,

_{2}*γ*and

*α*are dispersion, nonlinear and loss coefficients of TF, respectively. The formal solution of Eq. (1) for a single span of TF is

*L*is the fiber length. Let the output signal field of the fiber pass through an OPC, as shown in Fig. 1(a). The output of the OPC is

_{a}*N*(t,z)*. Using Eq. (4), the output of the OBPF is

Thus, the input field envelope can be recovered by performing a phase conjugation in the electrical domain at the receiver. Equation (5) is equivalent to

*P*is the power launched to the TF. Equation (10) describes the field propagation in an ideal fiber with a constant

_{in}*β*and a negative loss coefficient (or equivalently the power increasing with distance) that exactly compensates for dispersion and nonlinearity of the TF. However, it is hard to realize such a fiber in practice. For an ideal OBP, we like to have a short length of a fiber (so that its insertion loss is small) which provides the same response as that of the ideal OBPF given by Eq. (10). Here, we derive an equivalent way of realizing Eq. (10) by using amplifiers and a DDF with positive loss coefficient

_{2}*α*and a dispersion profile

_{d}*β*[see Fig. 1(b)]. The optical field envelope in the DDF is described by

_{2,d}(z_{d})*G'*is the gain of the amplifier preceding DDF. Using transformationsandEquation (11) can be rewritten as

*L*is found as follows. Total accumulated dispersion of the ideal OBPF [Fig. 1(a)] should be the same as that of the DDF, i.e.,

_{d}So far we considered the compensation of dispersion and nonlinearity of a single-span fiber optic link. For a multiple-span transmission system, Fig. 3 shows the schematic of a WDM fiber optic transmission system consisting of M transmitters, N spans of TFs, the OBP module, and M coherent receivers. The OBP is applied at the end of the transmission link. A pre-amplifier with gain *G* is introduced so that the required dispersion profile and length of the DDF can be adjusted according to Eqs. (23) and (25), respectively. A band pass filter (BPF) is introduced to remove the out of band amplified spontaneous emission (ASE) noise. During backpropagation, amplifiers with gain *G _{d}* are used to fully compensate for the loss of each span of DDF.

In DBP, the compensation of fiber dispersion and nonlinearity is implemented in a step-wise manner and the performance is usually limited by the step size which has to be traded off against computational cost or system complexity. In WDM systems, the required computational load may prevent DBP from real time implementation. In the OBP with DDF, the compensation of dispersion and nonlinearity is realized by a gradually decreasing dispersion profile, which inherently has a very small step size. The DDF with exponentially dispersion decreasing fibers have been fabricated before [21,22]. The step size of the order of a few meters in DDF can be realized and hence, nearly ideal OBP can be realized using DDF. The DDF can be fabricated by tapering the fiber during drawing process which alters the waveguide contribution to the dispersion [21]. The maximum dispersion required for OBP fiber is of the same order as the commercially available dispersion compensation fiber and of the same sign.

## 3. Simulation results

We simulate a WDM fiber optic transmission system with OBP at the receiver with the following parameters: number of WDM channels = 5, channel spacing = 100 GHz, symbol rate per channel = 25 Gsymbols/s, modulation = 32 quadrature amplitude modulation (QAM), number of symbols simulated = 32768 per channel. The linewidths of the transmitter and local oscillator lasers are 100 kHz each. The dispersion, loss, and nonlinear coefficients of the TF are *β _{2}* = 5 ps

^{2}/km,

*α*= 0.2 dB/km, and

*γ*= 2.2 W

^{−1}km

^{−1}, respectively. This type of fiber has been fabricated before and it is known as negative dispersion fiber (NDF) [23,24]. The amplifier spacing is 60 km, and the spontaneous emission noise factor is

*n*= 1.5. The BPF shown in Fig. 3 is a second order Gaussian filter with full bandwidth of 450 GHz. For the DDF,

_{sp}*α*= 0.4 dB/km,

_{d}*γ*= 4.86 W

_{d}^{−1}km

^{−1}, and

*L*= 12.1 km [see Fig. 2(b)]. The corresponding amplifier gain for compensating the DDF loss is 4.84 dB. In all the simulations, 32 samples per symbol are used in the transmission link so as to obtain a frequency window covering all the WDM channels. In DBP simulations, 2 samples per symbol are used after the ADC unless otherwise specified, while in OBP simulations, backpropagation is in the optical domain and 32 samples per symbol are used. Using the method of [2], the coupled NLSE is used to compensate for the inter-channel nonlinear impairments ignoring four-wave mixing (FWM). However, the OBP scheme compensates for both cross-phase modulation (XPM) and FWM simultaneously. The central channel is demultiplexed using a second order Gaussian filter with full bandwidth of 50 GHz. In the coherent receiver, for OBP, two samples per symbol are used after the ADC and phase noise compensation is done using the approach of [25]. A low pass filter (LPF) of bandwidth 25 GHz is used prior to phase noise compensation. For the DBP scheme, coupled NLSE is solved in digital domain prior to phase noise compensation. The optical and electrical filter bandwidths are optimized in both OBP and DBP schemes.

_{d}Figure 4 shows the bit error ratio (BER) as a function of the launch power per WDM channel when the transmission distance is 1200 km. The solid curve represents the BER of OBP using DDFs, and the dashed and dotted curves represent the BER of DBP with 3 km and 10 km step sizes, respectively. The DBP step size of the simulated WDM system is limited by the walk-off length [2], which is 3.2 km. We found that there is no obvious performance improvement when a step size smaller than 3 km is chosen for DBP, consistent with the results of [2]. Also, Fig. 4 shows the simulation results of DBP with 4 samples/symbol ADC sampling rate and DBP with DSP upsampling [1] from 2 to 4 samples/symbol. The DBP performance can be improved by increasing ADC sampling rate or DSP upsampling, at the cost of increased system complexity or computational cost. The OBP outperforms DBP (2 samples/symbol, step size = 3 km) by 2.0 dBQ. The relatively poor performance of DBP as compared to OBP is mainly due to the down sampling penalty and the lack of FWM compensation. The performance of midpoint OPC is worse than DBP, because the power profile is unsymmetrical with respect to the location of OPC. The performance of OBP is worse than that of DBP (with step size = 3 km) when the launch power is less than −2 dBm which is due to the optical signal to noise ratio (OSNR) penalty resulting from OBP amplifiers. The OSNR penalty due to OBP amplifiers is found to be 0.56 dB. From Fig. 4, it can also be seen that the DBP with a step size of 10 km performs worse than the DBP with a step size of 3 km even at lower launch powers (−10 dBm to −6 dBm) due to residual nonlinearity. The curve with ‘ + ’ shows the case where no OBP (or DBP) is applied and fiber dispersion and laser phase noise are compensated in the receiver. As can be seen, the performance of this system is much worse than the system with DBP or OBP.

Figure 5 shows the minimum BER as a function of transmission distance. The BER_{min} is obtained by optimizing the launch power for each distance. At the BER of 2.1 × 10^{−3}, the transmission reaches of linear compensation only and midpoint OPC are 300 km and 360 km, respectively. For DBP with a 10 km step size and 2 samples/symbol sampling rate, the reach is 760 km, which can be increased to 1600 km by using a 3 km step size at the cost of more than tripling the computational effort. The transmission reach of OBP with DDF is 2460 km. Although the OBP fully compensates for signal-signal nonlinear interactions, it neither compensates for signal-ASE nonlinear interactions [26,27] nor mitigates nonlinear polarization mode dispersion (PMD) [28], which are the limiting factors to enhance the reach in systems based on OBP.

## 4. Conclusions

We have investigated the performance of an OBP scheme consisting of an OPC and N spans of DDFs followed by amplifiers to compensate for dispersion and nonlinear effects of an N-span fiber optic WDM system. We have identified the conditions under which the nonlinear effects (both intra- and inter-channel nonlinearities) can be fully compensated and obtained an analytical expression for the novel dispersion profile of the DDF which provides the exact compensation of intra- and inter-channel signal-signal nonlinear impairments. The performance of the proposed OBP scheme is compared with DBP and midpoint OPC and simulation results show that the transmission reach can be significantly enhanced using the OBP with DDF.

## References and links

**1. **X. Li, X. Chen, G. Goldfarb, E. Mateo, I. Kim, F. Yaman, and G. Li, “Electronic post-compensation of WDM transmission impairments using coherent detection and digital signal processing,” Opt. Express **16**(2), 880–888 (2008). [CrossRef] [PubMed]

**2. **E. Mateo, L. Zhu, and G. Li, “Impact of XPM and FWM on the digital implementation of impairment compensation for WDM transmission using backward propagation,” Opt. Express **16**(20), 16124–16137 (2008). [CrossRef] [PubMed]

**3. **E. Ip and J. M. Kahn, “Compensation of dispersion and nonlinear impairments using digital backpropagation,” J. Lightwave Technol. **26**(20), 3416–3425 (2008). [CrossRef]

**4. **E. Ip, “Nonlinear compensation using backpropagation for polarization-multiplexed transmission,” J. Lightwave Technol. **28**(6), 939–951 (2010). [CrossRef]

**5. **E. F. Mateo, F. Yaman, and G. Li, “Efficient compensation of inter-channel nonlinear effects via digital backward propagation in WDM optical transmission,” Opt. Express **18**(14), 15144–15154 (2010). [CrossRef] [PubMed]

**6. **D. S. Millar, S. Makovejs, C. Behrens, S. Hellerbrand, R. I. Killey, P. Bayvel, and S. J. Savory, “Mitigation of fiber nonlinearity using a digital coherent receiver,” IEEE J. Sel. Top. Quantum Electron. **16**(5), 1217–1226 (2010). [CrossRef]

**7. **L. B. Du and A. J. Lowery, “Improved single channel backpropagation for intra-channel fiber nonlinearity compensation in long-haul optical communication systems,” Opt. Express **18**(16), 17075–17088 (2010). [CrossRef] [PubMed]

**8. **R. Asif, C. Y. Lin, M. Holtmannspoetter, and B. Schmauss, “Optimized digital backward propagation for phase modulated signals in mixed-optical fiber transmission link,” Opt. Express **18**(22), 22796–22807 (2010). [CrossRef] [PubMed]

**9. **D. Rafique and A. D. Ellis, “Impact of signal-ASE four-wave mixing on the effectiveness of digital back-propagation in 112 Gb/s PM-QPSK systems,” Opt. Express **19**(4), 3449–3454 (2011). [CrossRef] [PubMed]

**10. **G. Gao, X. Chen, and W. Shieh, “Influence of PMD on fiber nonlinearity compensation using digital back propagation,” Opt. Express **20**(13), 14406–14418 (2012). [CrossRef] [PubMed]

**11. **Y. Gao, J. H. Ke, K. P. Zhong, J. C. Cartledge, and S. S.-H. Yam, “Assessment of intra-channel nonlinear compensation for 112 Gb/s dual polarization 16QAM systems,” J. Lightwave Technol. **30**(24), 3902–3910 (2012). [CrossRef]

**12. **A. Yariv, D. Fekete, and D. M. Pepper, “Compensation for channel dispersion by nonlinear optical phase conjugation,” Opt. Lett. **4**(2), 52–54 (1979). [CrossRef] [PubMed]

**13. **S. Watanabe and M. Shirasaki, “Exact compensation for both chromatic dispersion and Kerr effect in a transmission fiber using optical phase conjugation,” J. Lightwave Technol. **14**(3), 243–248 (1996). [CrossRef]

**14. **P. Minzioni, I. Cristiani, V. Degiorgio, L. Marazzi, M. Martinelli, C. Langrock, and M. M. Fejer, “Experimental demonstration of nonlinearity and dispersion compensation in an embedded link by optical phase conjugation,” IEEE Photonics Technol. Lett. **18**(9), 995–997 (2006). [CrossRef]

**15. **S. Kumar and D. Yang, “Optical backpropagation for fiber-optic communications using highly nonlinear fibers,” Opt. Lett. **36**(7), 1038–1040 (2011). [CrossRef] [PubMed]

**16. **J. Shao and S. Kumar, “Optical backpropagation for fiber-optic communications using optical phase conjugation at the receiver,” Opt. Lett. **37**(15), 3012–3014 (2012). [CrossRef] [PubMed]

**17. **M. D. Pelusi, “WDM signal all-optical pre-compensation of the fiber nonlinearity in dispersion-managed links,” IEEE Photonics Technol. Lett. **25**(1), 71–74 (2013). [CrossRef]

**18. **S. Kumar and J. Shao, “Optical back propagation with optimal step size for fiber optic transmission systems,” IEEE Photonics Technol. Lett. **25**(5), 523–526 (2013). [CrossRef]

**19. **M. Morshed, L. B. Du, B. Foo, M. D. Pelusi, and A. J. Lowery, “Optical phase conjugation for nonlinearity compensation of 1.21-Tb/s Pol-Mux coherent optical OFDM,” in *18th OptoElectronics and Communications Conference* (2013), paper PD3–4.

**20. **D. Rafique and A. D. Ellis, “Nonlinearity compensation via spectral inversion and digital back-propagation: A practical approach,” in *Optical Fiber Communication Conference*, OSA Technical Digest (Optical Society of America, 2012), paper OM3A.1. [CrossRef]

**21. **S. V. Chernikov, E. M. Dianov, D. J. Richardson, and D. N. Payne, “Soliton pulse compression in dispersion-decreasing fiber,” Opt. Lett. **18**(7), 476–478 (1993). [CrossRef] [PubMed]

**22. **A. J. Stentz, R. W. Boyd, and A. F. Evans, “Dramatically improved transmission of ultrashort solitons through 40 km of dispersion-decreasing fiber,” Opt. Lett. **20**(17), 1770–1772 (1995). [CrossRef] [PubMed]

**23. **I. Tomkos, D. Chowdhury, J. Conradi, D. Culverhouse, K. Ennser, C. Giroux, B. Hallock, T. Kennedy, A. Kruse, S. Kumar, N. Lascar, I. Roudas, M. Sharma, R. S. Vodhanel, and C.-C. Wang, “Demonstration of negative dispersion fibers for DWDM metropolitan area networks,” IEEE J. Sel. Top. Quantum Electron. **7**(3), 439–460 (2001). [CrossRef]

**24. **J. Conradi, S. Kumar, and S. S. Rosenblum, “Negative dispersion single mode waveguide fiber,” US patent 6430346 B1, (2002).

**25. **T. Pfau, S. Hoffmann, and R. Noé, “Hardware-efficient coherent digital receiver concept with feedforward carrier recovery for M-QAM constellations,” J. Lightwave Technol. **27**(8), 989–999 (2009). [CrossRef]

**26. **J. P. Gordon and L. F. Mollenauer, “Phase noise in photonic communications systems using linear amplifiers,” Opt. Lett. **15**(23), 1351–1353 (1990). [CrossRef] [PubMed]

**27. **S. Kumar, “Effect of dispersion on nonlinear phase noise in optical transmission systems,” Opt. Lett. **30**(24), 3278–3280 (2005). [CrossRef] [PubMed]

**28. **C. Xie, “Inter-channel nonlinearities in coherent polarization-division-multiplexed quadrature-phase-shift-keying systems,” IEEE Photonics Technol. Lett. **21**(5), 274–276 (2009). [CrossRef]