Experimental Demonstration of Postnonlinearity ... - IEEE Xplore

1 downloads 0 Views 630KB Size Report
Abstract—We demonstrate significant performance improve- ment in 10-Gb/s multispan differential phase-shift keying trans- missions by employing ...
IEEE PHOTONICS TECHNOLOGY LETTERS, VOL. 18, NO. 9, MAY 1, 2006

1007

Experimental Demonstration of Postnonlinearity Compensation in a Multispan DPSK Transmission Guanghao Zhu, Linn Mollenauer, and Chris Xu

Abstract—We demonstrate significant performance improvement in 10-Gb/s multispan differential phase-shift keying transmissions by employing postnonlinearity compensation (PNC) for both dispersion-managed soliton and quasi-linear transmission. Measurements of timing misalignment tolerance as well as bandwidth requirement of the PNC unit show that the PNC technique is robust and practical, and can be easily extended to higher bit-rate transmissions. Index Terms—Differential phase-shift keying (DPSK), optical fiber communication.

I. INTRODUCTION

I

N long-haul differential phase-shift keying (DPSK) transmission, the dominant nonlinear effect is the nonlinear phase fluctuation caused by the combination of amplified spontaneous emission (ASE) noise and self-phase modulation (SPM) [1], [2]. During the transmission process, the intensity profile of the signal is first perturbed by the distributed ASE noise. As a result of the intensity-dependent refractive index, the optical phase of the signal pulse is then randomly shifted. At the receiver end, since the phase shifts due to SPM of two nearby pulses are statistically independent, the DPSK eye gets blurred and bit errors occur. To mitigate this problem, it has been proposed that a properly adjusted nonlinear crystal [3] or an electrooptic phase modulator [4], with a phase sign opposite to that of the fiber Kerr nonlinearity, can be employed to enhance the system performance in the presence of the nonlinear phase perturbations. Using such a PNC technique [4], significant improvement in the -factor was demonstrated in a simulated transmission experiment using a single span of highly nonlinear fiber [5]. Here we report the results of our study on a single-channel DPSK multispan transmission experiment in a recirculating fiber loop employing the PNC technique. Our measurements show that with PNC, at a transmission distance of 4800 km, the -factor can be improved by 7.5 dB for dispersion-managed soliton and 1.9 dB for the quasi-linear transmission, solidly confirming the effectiveness of the PNC technique in a practical multispan DPSK transmission system. II. EXPERIMENTS AND RESULTS The experimental setup is shown in Fig. 1(a). The DPSK transmitter consists of a 1570-nm continuous-wave distributed feedback laser, a LiNbO Mach–Zehnder intensity modulator Manuscript received November 8, 2005; revised February 1, 2006. The authors are with the School of Applied and Engineering Physics, Cornell University, Ithaca, NY 14850 USA (e-mail: [email protected]). Digital Object Identifier 10.1109/LPT.2006.873522

Fig. 1. (a) Experimental setup for demonstrating performance improvement of a single-channel multispan DPSK transmission using PNC technique. DFB: distributed feedback laser. AO: acoustooptic. EDFA: erbium-doped fiber amplifier. DI: delayed interferometer. BERT: BER detector. (b) Polarization diversity scheme of the PNC unit. PBS: polarization beam splitter. ATT: variable optical attenuator. PD: photodetector.

as the pulse carver, and a second LiNbO Mach–Zehnder intensity modulator as the DPSK encoder. The generated DPSK pulse train had a duty cycle of 33% and was launched into an all-Raman amplified dispersion-managed recirculating fiber loop [6]. The recirculating loop is comprised of six spans of TrueWave Reduced Slope (TWRS) fiber. Each span consists of 100 km of TWRS fiber followed by nearly slope-matched dispersion-compensating fiber (DCF). The resulting path-averaged dispersion for the loop is measured to be 0.13 ps/km/nm at 1570-nm wavelength. Two discrete Raman amplifiers [not shown in Fig. 1(a)] were employed in the loop to compensate the losses from an intraloop acoustooptic switch as well as a 3-dB coupler. Tunable bandpass filters were also used both inside and outside the loop to select the signal and reject the out-band ASE.

1041-1135/$20.00 © 2006 IEEE

1008

IEEE PHOTONICS TECHNOLOGY LETTERS, VOL. 18, NO. 9, MAY 1, 2006

Q

Fig. 3. Measured -factor as a function of the transmission distance for the dispersion-managed soliton transmission. Solid line: without PNC. Dashed line: with PNC. The arrow indicates that the -factor is greater than 16.5 (BER 10 ), which is the largest value measured in our experiment.

Q

Fig. 2. Measured eye diagrams for dispersion-managed soliton transmission at a distance of 3600 km. (a) Without PNC; (b) with PNC.

The received signal was sent through an optional postdispersion compensation coil (used only for the case of quasi-linear transmission), followed by a DPSK receiver to evaluate the transmission performance. The DPSK receiver consists of a PNC unit, a 100-ps delayed-interferometer, and a balanced detector. To remove the polarization sensitivity, a polarization diversity scheme [7] was used to build the PNC unit. The detailed construction of the PNC unit is shown in Fig. 1(b). The received DPSK signal with different polarization state is first separated by a polarization beam splitter, sent to two different electrooptic phase modulators, and then recombined together by a second polarization beam splitter. The driving signal of the phase modulators is derived from the received optical signal via a combination of a photodetector and a broadband radio-frequency (RF) amplifier [not shown in Fig. 1(b)]. The bandwidth of the PNC unit can be changed by inserting low-pass RF filters. The sign of the driving signal is chosen so that the electrooptic phase modulation is opposite to the phase shift caused by the Kerr nonlinearity. The strength of the driving signal can be tuned by adjusting the optical attenuator placed in front of the photodetector. The total insertion loss of the PNC unit is 9 dB. The optical and electrical path length of the PNC unit were carefully matched to achieve the best compensation effect. We first study the dispersion-managed soliton transmission by launching the DPSK pulse train at a path averaged power of 10.8 dBm. The effect of the phase noise reduction was readily observed from the eye diagrams without and with the PNC unit (Fig. 2). The bit-error rate (BER) was measured as a function of the transmission distance. The -factor calculated from the

Q

Fig. 4. Measured -factor as a function of the transmission distance for the quasi-linear transmission case (4 dB below the soliton power level). Solid line: without PNC. Dashed line: with PNC. The arrow indicates that the ), which is the largest value measured -factor is greater than 16.5 (BER 10 in our experiment.

Q

BER is plotted in Fig. 3 both with and without PNC. The measured results show that when PNC is employed to partially remove the nonlinear phase noise, a significant -factor improvement (7.5 dB), can be achieved at a transmission distance of 4800 km (corresponding to a total Kerr phase shift of 2.5 rad). We note that our measured -factor improvement exceeds previous theoretical prediction (6 dB) [3], [4], [8], [9]. This extra improvement is caused by the fact that, before the DPSK signal was launched into the loop, the intensity profile of the pulse train was already uneven as a result of the pattern dependence of the data modulator. The assumption on the initial intensity fluctuation caused by transmitter imperfection is supported by our theoretical analysis, following the method of [1]. We have also measured the system BER for quasi-linear transmission by decreasing the average launch power 4 dB below the soliton level and compensating the residual dispersion outside the recirculating loop with an optimized DCF spool. The -factor calculated from the measured BER is plotted in Fig. 4 both with and without PNC. Comparing Figs. 4 and 3, one sees that by lowering the optical launch power and switching to quasi-linear transmission, there is less performance gained by

ZHU et al.: EXPERIMENTAL DEMONSTRATION OF PNC IN A MULTISPAN DPSK TRANSMISSION

Fig. 5. Measured average PNC phase (in radian) as a function of distance for the dispersion-managed soliton transmission.

1009

, the maximum certain lower launch power. At a BER of transmission distance for our system is obtained at a launch power of 4 dB below the soliton level. The applied average PNC phase (for achieving the best BER) as a function of the transmission distance was also measured in this study and the result is plotted in Fig. 5 for soliton transmission. We observed that the average PNC phase increases nearly linearly as a function of the transmission distance. The magnitude of the measured average PNC phase is slightly larger than the value predicted by the previous theoretical analysis, i.e., half of the soliton Kerr phase shift [3], [4], [8], [9]. Once again, such a deviation can be explained by transmitter imperfections before the loop. To characterize the practicality of the PNC unit, we measured the optical–electrical path mismatch tolerance and the phase modulation bandwidth requirement for our PNC unit. The measured results for dispersion-managed soliton transmission are plotted in Fig. 6. It can be observed that at a transmission distance 3000 km and with a PNC bandwidth 10 GHz, a mismatch of 10 ps can be tolerated for 1-dB -factor degradation [see Fig. 6(a)]. It is also found that a decrease in PNC bandwidth from 10 to 5 GHz will introduce a 0.5-dB reduction in the -factor [see Fig. 6(b)]. These results clearly indicate that the PNC technique is robust against small timing misalignment and has a bandwidth requirement no greater than the transmission bit rate. III. CONCLUSION We have demonstrated the PNC technique in a multispan DPSK transmission experiment. We show that significant performance improvement can be achieved for both dispersionmanaged soliton and quasi-linear transmission. Our measurements on the timing misalignment tolerance and bandwidth requirement also show that a robust and practical PNC unit can be readily achieved using line-rate optoelectronics. REFERENCES

Q

Fig. 6. Measured characteristics of the PNC unit. (a) -factors as a function of optical–electrical path mismatch at distance 3000 km with PNC bandwidth 10 GHz. (b) -factors with PNC bandwidths of 10 and 5 GHz. Both (a) and (b) are obtained for dispersion-managed soliton transmission. The arrow indicates ), which is the largest value that the -factor is greater than 16.5 (BER 10 measured in our experiment.

Q

Q

the PNC technique, but the overall transmission performance improves. This performance improvement can be readily explained by the reduction of the nonlinear phase noise, which is proportional to the SPM accumulated during the transmission. The maximum transmission distance at a given BER is eventually limited by the reduced optical signal-to-noise ratio at

[1] J. P. Gordon and L. F. Mollenauer, “Phase noise in photonic communications system using linear amplifiers,” Opt. Lett., vol. 15, pp. 1351–1353, Dec. 1990. [2] H. Kim and A. Gnauck, “Experimental investigation of the performance limitation of DPSK systems due to nonlinear phase noise,” IEEE Photon. Technol. Lett., vol. 15, no. 2, pp. 320–322, Feb. 2003. [3] X. Liu, X. Wei, R. Slusher, and C. J. McKinstrie, “Improving transmission performance in PSK systems by lumped nonlinear phase shift compensation,” Opt. Lett., vol. 27, pp. 1616–1618, 2002. [4] C. Xu and X. Liu, “Post-nonlinearity compensation with data driven phase modulators in phase shift keying transmission,” Opt. Lett., vol. 27, pp. 1629–1631, 2002. [5] J. Hansryd, J. van Howe, and C. Xu, “Experimental demonstration of nonlinear phase jitter compensation in DPSK modulated fiber links,” IEEE Photon. Technol. Lett., vol. 17, no. 1, pp. 232–234, Jan. 2005. [6] C. Xu, X. Liu, L. F. Mollenauer, and X. Wei, “Comparison of return-to zero differential phase-shift keying and ON–OFF keying in long-haul dispersion managed transmission,” IEEE Photon. Technol. Lett., vol. 15, no. 4, pp. 617–619, Apr. 2003. [7] C. Xu, L. F. Mollenauer, and X. Liu, “Compensation of nonlinear selfphase modulation with phase modulators,” Electron. Lett., vol. 38, pp. 1578–1579, Nov. 2002. [8] A. Mecozzi, “Probability density functions of the nonlinear phase noise,” Opt. Lett., vol. 29, pp. 673–675, Apr. 2004. [9] K. P. Ho, “Probability density of nonlinear phase noise,” J. Opt. Soc. Amer. B., vol. 20, pp. 1875–1879, Sep. 2003.