Pilot-added carrier-phase recovery scheme for Nyquist M-ary quadrature amplitude modulation optical fiber communication system

Transcription

1 Pilot-added carrier-hase recovery scheme for Nyquist M-ary quadrature amlitude modulation otical fiber communication system Wenbo Zhang ( 张文博 ) 1, *, Dongwei Pan ( 潘董威 ) 2, Xiaofei Su ( 宿晓飞 ) 2, Xiaoguang Zhang ( 张晓光 ) 2, **, Lixia Xi ( 席丽霞 ) 2, and Xianfeng Tang ( 唐先锋 ) 2 1 School of Science, Beijing University of Posts and Telecommunications, Beijing , China 2 State Key Laboratory of Information Photonics and Otical Communications, Beijing University of Posts and Telecommunications, Beijing , China *Corresonding author: zhangwb@but.edu.cn; **corresonding author: xgzhang@but.edu.cn ceived July 22, 2015; acceted November 26, 2015; osted online January 19, 2016 A new two-stage carrier-hase recovery scheme using a combination of an otical ilot-aided algorithm with the crossed constellation transformation algorithm for either square-framed or non-square-framed M-levelquadrature amlitude modulation (QAM) Nyquist systems is roosed. It is verified in 32- and 128-QAM systems that it can rovide high linewidth tolerance with little comlexity. OCIS codes: , , doi: /COL To meet the ever-increasing demands of high-seed, longhaul otical fiber communication systems, high-order modulation formats and suerchannel signaling are two issues eole are ursuing to achieve higher sectral efficiency. Comared to quadrature hase-shift keying, which has been adoted in the current commercial 100G coherent otical fiber communication system, M-level-quadrature amlitude modulation (M-QAM), in which M is more than 16, is the best candidate for next-generation 400G or 1T systems. Nyquist signaling and otical orthogonal frequency-domain multilexing might also be used in 400G or 1T systems. In the former, a combination with M-QAM is referred, because a Nyquist-M-QAM-wavelengthdivision multilexing (WDM) system is more backward comatible with the current WDM system. Carrier-hase recovery (CPR) is a significant and integral art of the digital signal rocessing (DSP) rocedure to comensate for the hase noise (PN) induced by both the transmitter and the local oscillator (LO) lasers. Many blind CPR algorithms have been roosed and demonstrated to mitigate the PN [1 10]. Among these methods, blind hase search (BPS) has the best linewidth tolerance for arbitrary M-QAM formats, but at the cost of high comutation comlexity [1]. Moreover, the algorithm s comlexity increases as the QAM order (M) increases [11]. Many modified CPR algorithms have been introduced, such as maximum likelihood (ML) and constellation transformation (CT), to reduce the comutation comlexity, but with reduced linewidth tolerance [4,10]. In this work, a two-stage CPR scheme is roosed, in which the otical ilot-aided (PA) algorithm [7] works as the first stage and a so-called crossed constellation transformation (CCT) algorithm works as the second algorithm. Using the otical ilots just alongside the Nyquist- WDM signal bands, we can mitigate or comensate for imairments such as frequency offset, carrier PN, etc. because the ilots exerience the same imairments from which the signals suffer. The analyses show that the roosed two-stage CPR scheme ossesses a good linewidth tolerance for arbitrary M-QAM formats with much lesser comlexity. Numerical simulations for dual-carriers Nyquist 32-QAM and 128-QAM otical communication systems using the roosed and other CPR methods are imlemented. The erformance of the PA + CCT CPR method is studied and comared with other CPR methods, such as only PA, PA + ML, and BPS. Figure 1(a) shows the configuration of roosed otical PA dual-carriers Nyquist system. The laser light works at the frequency f 0, which is emitted by an external cavity PC ODL BPS PBC Fig. 1. (a) Proosed PA dual-carriers Nyquist transmission system. (b) Electrical Nyquist signal generation /2016/020601(5) Chinese Otics Letters

2 laser (ECL) slit into two beams. One is used as ilot tone, while the other is modulated with the data through a Mach Zenhder modulator driven by a sinusoidal RF signal with frequency f s ¼ 14.5 GHz to generate otical dual-carriers whose frequencies are f 0 f s and f 0 f s : E out ðtþ ¼A in cosð2πf 0 tþ sin½γ sinð2πf s tþš ¼ A in cosð2πf 0 tþ½2j 1 ðγþ sinð2πf s tþ 2J 3 ðγþ sinð2πf s tþ Š ¼ A in fj 1 ðγþ sin½2πðf 0 f s ÞtŠ J 1 ðγþ sin½2πðf 0 f s Þt hight order termsg: (1) The two generated subcarriers are then searated by a delay interferometer (DI). Each subcarrier is modulated with a Nyquist M-QAM (e.g., 32- or 128-QAM) signal at the symbol rate of 28 Gbaud, which will lead to a ilot ga width of 1 GHz between two sub-channels and will sacrifice only 1.8% bandwidth overhead. A variable otical attenuator is used to adjust the ilot-to-signal ratio (PSR), and an otical delay line (ODL) is used to ensure hase synchronization between the ilot tone and two carrier signals. The Nyquist signal is then olarization multilexed through a 50/50 slitter (BPS), an ODL, and a olarization beam combiner (PBC). At the receiver end, a coherent front end with an LO laser rovides the electric signals corresonding to the I and Q comonents. After analog-to-digital conversion (ADC), these signals are rocessed by the DSP unit. In the simulation, we mainly focus on the laser hase estimation and carrierhase recovery techniques in the Nyquist 32- or 128-QAM system, and assume that all the other imairments, such as chromatic disersion and olarization mode disersion, etc., are comensated for by the work in fs. [12 14]. In the CPR rocess, the PA-CPR acts as the first stage to comensate for the PN, and the CCT algorithm following the PA-CPR acts as the second stage. At the transmitter (Tx) end, the DSP block diagram for generating the Nyquist electrical 32-/128-QAM signal is shown in Fig. 1(b). The transmitted symbols are first maed to a 32-/128-QAM constellation with I and Q tributaries. After being u-samled to 2 samles/symbol, the 32-/128-QAM signal is fed into a 32-ta digital Nyquist root-raised cosine ulse shae filter with a 0.02 roll-off factor. Finally, digital-to-analog conversions (DAC) are carried out. The basic rincile of the roosed PA-CPR is to insert a ilot tone in the middle of the two sub-channels at the Tx end. As the ilot tone and the data signal are generated from same source and are exactly aligned in the time domain, they will exerience the same PN caused by both the Tx and Rx lasers. Thus, we can use the information extracted from the ilot tone to comensate for the PN jointly in the both the sub-channels at the receiver end. The transmitted signal can be exressed as E Tx ðtþ ¼ 1 ðtþe j½2πðf 0 f s Þtθ Tx ðtþš 2 ðtþe j½2πðf 0f s Þtθ Tx ðtþš e j½2πf 0tθ Tx ðtþš ; (2) E ilot where M 1 ðtþ and M 2 ðtþ are the signals of the two Nyquist data sub-channels modulated with 32- or 128-QAM, θ TX is the PN caused by the Tx laser, and E s and E ilot are the amlitudes of the data signal and the ilot tone, resectively. Through transmission in the otical fiber channel, for simlicity, only amlified sontaneous emission (ASE) noise induced by the erbium-doed fiber amlifiers is considered. The central frequency of the LO is f 0.Ifwe consider the Rx laser to have a PN of θ RX, the received signal can be written as E Rx ðtþ ¼μ 1 ðtþe j½2πð f sþtθ Tx ðtþθ Rx ðtþš 2 ðtþe j½2πf stθ Tx ðtþθ Rx ðtþš e j½θ TxðtÞθ Rx ðtþš nðtþ: (3) E ilot Figure 2 shows the details of the imlementation of the roosed two-stage carrier hase estimation scheme in the receiver s DSP module. In the first stage, PA-CPE is used. Two sub-channels and ilot tones are searated by using a digital frequency shifter and a low-ass filter (LPF) with a suitable order and bandwidth. After the ilot tone is filtered by the LPF, the instantaneous hase of the extracted ilot tone can be written as follows: θ ilot ðtþ ¼θ Tx ðtþθ Rx ðtþθ n ðtþ: (4) We can now easily extract the hase θ ilot ðtþ as an estimated value for the comensation for the combined laser PN θðtþ induced by the Tx and Rx lasers, where the estimation error deends on θ n ðtþ. 32QAM 128QAM Signal received Filtered ilot constellation Frequency Shifting LPF Frequency Shifting Pilot Phase tracking Match Filter Angle Match Filter ceived signal ex(-jθ) Signal PA-CPE CCT CCT Stage 1 Stage 2 Decision BER Est Decision BER Est PA-CCT Sub-ch1 Sub-ch2 Fig. 2. DSP imlementation of the multi-stage carrier-hase estimation scheme. The evolutions of the 32-QAM and 128-QAM constellations are shown in two stages

3 In order to comensate for the residual hase shift, we emloy the CCT hase estimation as the second stage of our roosed two-stage CPR algorithm. The key idea of the CCT algorithm is to transform the comlex constellations down to the crossed-signal constellation, just like the transformations from Fig. 3(b) to 3(c). After the original constellations are transformed down to the crossed-signal constellation, we can easily use the conventional fourthower method for the rest of the hase estimation. In addition, the CCT algorithm is not only suitable for square-frame QAMs like 16-QAM and 64-QAM, but also for non-square-frame QAMs like 32-QAM and 128-QAM, and hence it is more suitable to be the second stage after the PA-CPR algorithm because the PA-CPR is also modulation- format transarent. The transformation equation of the CCT is as follows: C ¼ I A I sign½i B I signði ÞŠ ifq A Q sign½q B Q signðqþšg; (5) where I and Q are the real and imaginary arts of the coordinates, A I and A Q are the absolute values of the dislacement comonents between the old oints and the new oints in the real and imaginary arts, B I and B Q are the absolute values of the coordinate oints at the target location, and signð Þ is the sign function. The values of A I, A Q, B I, and B Q are different, and rely on the modulation formats. Figure 3 shows the transformation rocedure of a 32-QAM (Figs. 3(b) and 3(c)) and a 128-QAM (Fig. 3(a) 3(c)) into the crossed constellation. The simulation latform of a Nyquist 32-QAM or 128- QAM transmission system was constructed to investigate the erformance of the roosed PA-CCT-CPR scheme. A total number of 2 17 symbols at the symbol rate of 28 Gbaud was used in this simulation. The PN is modeled Fig. 4. BER surface versus PSR and the bandwidth of the ilot LPF at two different laser linewidths for 32-QAM. as a Wiener rocess with a variance of σ 2 ¼ 2πΔf T s, where Δf denotes the combined linewidth of the transmitter and the local lasers, and T s is the symbol eriod. Different amounts of ASE noise are loaded to obtain different values of the otical signal-to noise ratio (OSNR), which is defined as the ratio of the otical ower of the signal to the noise ower with a 0.1 nm reference BW. There are two imortant arameters that have great influence on the erformance of the PA-CPR. One is the PSR and the other is the bandwidth of the ilotextracted LPF, which is denoted by B LPF. In order to investigate the erformance of the CPR scheme under (a) (b) (c) Fig. 3. Transformation rocedures of 128-QAM and 32-QAM into the crossed-signal constellation. Fig. 5. Actual and estimated PN fluctuations by PA: (a) 32-QAM, and (b) 128-QAM

4 Fig. 6. (a) OSNR enalty versus linewidth symbol rate at BER ¼ for 32-QAM. (b) OSNR enalty versus linewidth symbol rate at BER ¼ for 128-QAM. the different linewidths, the PSR and the B LPF must be otimized simultaneously. We set the PSR to range from 20 to 3 db and the B LPF to range from 50 to 250 MHz. Figure 4 illustrates how the values of the PSR and B LPF affect the erformance of the 32-QAM system, where the lower and uer surfaces corresond to Δf T s ¼ and Δf T s ¼ , resectively. It shows that as Δf T s increases, the otimum values of the PSR and B LPF also increase. The otimal PSR and B BPF values can be found based on the minimum value of the bit error rate (BER) under the different values of Δf T s. In Fig. 4, the otimal combinations of the PSR and B BPF (14 db, 70 MHz) and (11 db, 230 MHz) are found for Δf T s ¼ and Δf T s ¼ , resectively. In the following calculations, the otimum values of the PSR and B LPF are always considered for different Δf T s. In order to assess the erformance of our roosed twostage CPR algorithm in the Nyquist 32/128-QAM system, we first comare the actual PN induced jointly by the Tx and Rx lasers with the hase estimated by the extracted ilot, as shown in Fig. 5. The Δf T s is set to be , and the OSNR is 23 db in the Nyquist 32-QAM system and 30 db in the Nyquist 128-QAM system. The figure confirms that the PA-CPR can correctly track the hase fluctuations caused by the laser PN under the different modulation formats. After comensating for the PN with the PA algorithm, a second-stage CCT method is imlemented as the fine PN comensation to coe with the residual PN. The erformance of the PA + CCT CPR is assessed and comared to other CPR schemes in terms of the evaluation of linewidth tolerance for 32- and 128-QAM formats. Both the OSNR and linewidth are swet to get the curves of the OSNR enalty versus Δf T s at a BER of When the equivalent linewidth is 0 Hz, the OSNR required at a BER of is regarded as the reference for the initial enalty. Figures 6(a) and 6(b) show the OSNR enalty as a function of Δf T s for 32-QAM and 128-QAM, resectively, and also give the color constellation diagrams of each QAM format on the right. The CPR schemes used in the comarison are PA only, PA + ML, PA + CCT, and BPS. We can find that all the CPR algorithms show low OSNR enalty requirements when the Δf T s takes small values. As the value of Δf T s increases, a larger OSNR is required. As shown in Fig. 6, for a large value of Δf T s, it turns out that the two-stage PA + CCT CPR algorithm even shows a better erformance than the BPS algorithm, which is known to have best linewidth tolerance, according to the literature. Moreover, the results show that PA + ML and PA + CCT have a similar linewidth tolerance for 32-QAM, while for 128-QAM, the erformance of PA + CCT is a little better than that of the PA + ML. Table 1 lists ðδf T s Þ max of the CPR schemes mentioned above and the corresonding tolerable linewidth values Δf for the 28 Gbaud Nyquist 32/128-QAM system. Here, we define the linewidth symbol rate (Δf T s )ata 1 db OSNR enalty as the maximum tolerable linewidth value ðδf T s Þ max. We can conclude that the roosed two-stage PA + CCT CPR scheme shows the best linewidth tolerance erformance in the 32/128-QAM Nyquist system. Table 1. Linewidth Tolerance of Various CPR Schemes for 32- and 128-QAM 32-QAM 128-QAM CPR scheme ðδf T s Þ max Δf max at 28 Gbaud (MHz) ðδf T s Þ max Δf max at 28 Gbaud (khz) PA PA + ML PA + CCT BPS

5 In comarison with the BPS method, the first-stage PA-CPR only needs to searate the ilot from the carrier signals with a BPF and extract the instantaneous hase to comensate for the PN directly. The BPF can be achieved by a digital finite imulse resonse filter, which is easy to imlement in the hardware. Therefore, the roosed PA-CPR scheme has fewer oerations and is less comlex. The comutation comlexities of BPS, ML, and CT have been discussed in f. [3], and the conclusion is that the comlexity of the CT algorithm is similar to that of the ML algorithm, but much lower than that of the BPS algorithm. In the resent work, the mentioned CCT algorithm is another form of the CT algorithm. The difference is that CCT is suitable for non-squared-frame QAMs, and CT is not. Therefore, they have similar comutation comlexity. In conclusion, the roosed PA-CCT CPR has the best erformance of linewidth tolerance, and is less comlex than the BPS. We roose an otical domain PA CPR scheme by insetting the otical ilot between two subcarriers, which can realize hase recoveries of both two-sided subcarriers at the same time and reduced the bandwidth overhead. The PSR and the bandwidth of the LPF are otimized. Using the otimal arameters, the erformances of different CPR schemes are comared. PA+CCT shows good linewidth tolerance of Δf T s ¼ (for 32-QAM) and Δf T s ¼ (for 128-QAM) at a 1 db OSNR enalty, which is better than the BPS and is much less comlex. This work was artly suorted by the National Natural Science Foundation of China (Nos and ), the Oen Fund of IPOC (BUPT) (No. IPOC2013B005), the Fund of the State Key Laboratory of Information Photonics and Otical Communications (Beijing University of Posts and Telecommunications), and the Fundamental search Funds for the Central Universities (No. 2014RC1201). ferences 1. T. Pfau, S. Hoffmann, and R. Noé, J. Lightwave. Technol. 27, 989 (2009). 2. X. Zhou, IEEE Photon. Technol. Lett. 22, 1051 (2010). 3. J. H. Ke, K. P. Zhong, Y. Gao, J. C. Cartledge, A. Karar, S. Abdullah, and M. A. Mohammad, J. Lightwave Technol. 30, 3987 (2012). 4. K. P. Zhong, J. H. Ke, Y. Gao, and J. C. Cartledge, J. Lightwave Technol. 31, 50 (2013). 5. S. M. Bilal, C. R. S. Fludger, V. Curri, and G. Bosco, J. Lightwave Technol. 32, 2973 (2014). 6. M. Morsy-Osman, Q. Zhuge, L. R. Chen, and D. V. Plant, Ot. Exress 19, (2011). 7. Z. Zhang, J. Lin, X. Su, L. Xi, Y. Qiao, and X. Zhang, Ot. Eng. 53, (2014). 8. M. Morsy-Osman, Q. Zhuge, M. Chagnon, X. Xu, and D. V. Plant, in Proceedings of OFC 2013 OTu31.6 (2013). 9. K. Kikuchi, in Proceedings of ECOC 2008 Th.2.A.1 (2008). 10. Y. Gao, A. P. T. Lau, S. Y. Yan, and C. Lu, Ot. Exress 19, (2011). 11. S. M. Bilal, G. Bosco, P. Poggiolini, and C. R. S. Fludger, in Proceedings of ECOC (2013). 12. W. Huang, D. Liu, X. Zhang, and Y. Zhang, Chin. Ot. Lett. 9, (2011). 13. Y. Sun, L. Xi, X. Tang, D. Zhao, Y. Qiao, X. Zhang, and X. G. Zhang, Chin. Ot. Lett. 12, (2014). 14. J. Zhang, Y. Gu, X. Yuan, F. Tian, X. Zhang, and J. Tao, Chin. Ot. Lett. 10, (2012)