430
IEEE PHOTONICS TECHNOLOGY LETTERS, VOL. 18, NO. 2, JANUARY 15, 2006
All-Optical Chromatic Dispersion Monitoring of a 40-Gb/s RZ Signal by Measuring the XPM-Generated Optical Tone Power in a Highly Nonlinear Fiber T. Luo, C. Yu, Z. Pan, Y. Wang, J. E. McGeehan, M. Adler, and A. E. Willner
Abstract—We experimentally demonstrate an all-optical chromatic dispersion (CD) monitoring technique potential for ultrahigh-speed systems. A monochromatic continuous wave probe is coupled with the tapped-off data signal into a highly nonlinear fiber. An optical tone that is sensitive to CD is generated near the probe wavelength due to the cross-phase modulation effect and is used for CD monitoring. The monitoring window is 42 ps/nm for 40-Gb/s return-to-zero systems. This monitoring technique is simple and needs no modification at the transmitter. Index Terms—Chromatic dispersion (CD), CD monitoring, cross-phase modulation (XPM), highly nonlinear fiber, optical fiber communication.
I. INTRODUCTION
T
HERE HAS been much interest recently in optical performance monitoring due to the potential need to control and manage the physical state-of-the-network in a ubiquitous and cost-effective manner. Of course, an important characteristic of any optical signal is the amount of degradation that is caused by chromatic dispersion (CD), for which the monitoring signal may either drive an equalizer/compensator or notify the network manager to take appropriate diagnostic or routing action. We emphasize that CD effects grow as the squares of the bit rate increase, such that 40-Gb/s signals and greater are much more susceptible to problems. We note that chromatic dispersion in optical networks is not static and can change due to: 1) seasonal temperature changes; 2) repair and maintenance of the network plant; 3) chirp induced by drift in wavelength-selective components; and 4) data path changes in a dynamically reconfigurable network. A brute-force approach to monitoring the instantaneous chromatic dispersion effects on the data is to simply detect the highspeed signal and monitor the integrity of the electronic bits or the eye diagram. This approach may be costly, is bit-rate specific, and may have difficulty performing at very high bit rates. Previously published results on using more optically oriented techniques include: 1) adding a phase modulation and detecting the dispersion-induced intensity modulation [1]; 2) monitoring Manuscript received June 27, 2005; revised November 14, 2005. This work was supported in part by the National Science Foundation Initiative on HighCapacity Optical Network and in part by the Cisco URP program. T. Luo, C. Yu, Y. Wang, J. E. McGeehan, M. Adler, and A. E. Willner are with the Department of Electrical Engineering, University of Southern California, Los Angeles, CA 90089-2565 USA (e-mail:
[email protected]). Z. Pan is with the Department of Electrical and Computer Engineering, University of Louisiana, Lafayette, LA 70504-3890 USA. Digital Object Identifier 10.1109/LPT.2005.862359
the radio frequency (RF) tone [2]; 3) measuring time delay between upper and lower sidebands [3]; and 4) inserting in-band subcarriers and monitoring their RF tones [4]. Each of these techniques either requires at least one high-speed element in the receiver in order to monitor chromatic dispersion or tends to be bit-rate specific for which monitoring a higher data-rate signal would require significant modifications. Recently, two all-optical methods that do not require highspeed elements in the receiver have been reported: 1) measuring the idler signal due to four-wave mixing (FWM) effects in dispersion-shifted fiber [5] and 2) using spectral filtering of a selfphase modulation (SPM) broadened data optical spectrum [6]. These techniques measure optical power over a chosen bandwidth determined by the filter. It would be desirable, however, to have a simple optical tone as the monitoring signal. A nonlinear optics-based technique was proposed by Dorrer and Maywar to analyze the RF spectrum of optical signals from an optical spectrum analyzer [7], [8]. It was noted that the output optical spectrum was affected by the dispersion. In this letter, we experimentally demonstrate using this novel all-optical technique for monitoring the effect of chromatic dispersion on a high-speed return-to-zero (RZ) data signal by measuring the cross-phase modulation (XPM)-generated optical tone power in highly nonlinear fiber. The monitoring range and sensitivity of the technique is measured experimentally. In the monitoring module, the data signal is coupled with a monochromatic continuous wave, whose frequency is outside the spectrum of the signal. The combined signals interact with each other while traveling through a piece of highly nonlinear fiber, and an optical tone is generated near the monochromatic wave. This optical tone is sensitive to the chromatic dispersion of the data signal and therefore can be used for dispersion monitoring. The technique is demonstrated in a 40-Gb/s RZ system, the measurement window is 42 ps/nm. The monitoring sensitivity increases with the average data power. For a 6-dBm average data power system, the monitoring sensitivity is 0.5 dB/ps/nm. This monitoring technique is simple and needs no modification at the transmitter. Moreover, since no high-speed electronic components are needed for monitoring, it can potentially work for up to terabit systems by simply tuning an optical bandpass filter. II. PRINCIPLE OF MONITORING TECHNIQUE A conceptual diagram of the proposed all-optical monitoring technique is shown in Fig. 1. In the monitoring module, the is tapped out and coupled with a RZ data signal at bit rate
1041-1135/$20.00 © 2006 IEEE
LUO et al.: ALL-OPTICAL CHROMATIC DISPERSION MONITORING OF 40-GB/s RZ SIGNAL
431
Fig. 1. Concept of using XPM-induced optical tones as CD monitoring signal.
monochromatic continuous wave probe, whose frequency is outside the spectrum of the signal. The combined signals then travel through a piece of highly nonlinear fiber. At the output, an optical spectrum analyzer (OSA) is used to observe the optical spectrum around the probe frequency. Without CD distortion, a . pair of optical tones is generated at the frequency of These two optical tones, however, change with the CD value in the transmission link. As CD increases and distorts the data, these two optical tones fade with the CD value. Therefore, the are sensitive to CD optical tones at the frequency of and can be used as the CD monitoring signal. The generation of these two optical tones around the probe wavelength has previously been explained in [7] and [8]. While propagating in the highly nonlinear fiber, the intensity of the data signal modulates the electric field of the monochromatic wave via XPM effect. Therefore, the intensity information of the data signal is imprinted onto the electric field of the monochromatic probe wave. At the output of the highly nonlinear fiber, is [7], [8] the electric field around the probe frequency
Fig. 2. Simulation results of optical spectrum around probe channel in 40-Gb/s RZ system: (a) CD = 0 in transmission link and (b) CD = 42 ps/nm in transmission link.
(1) where is a constant determined by the nonlinear index coefficient of the highly nonlinear fiber for XPM effect and is the intensity of the data signal. The approximation is valid for ). The optical spectrum small modulation (i.e., measures the power of the temporal electric field. Therefore, the is expressed as [7], [8] optical power around
Fig. 3. Experimental setup of CD monitor. OBPF is optical bandpass filter.
can serve as the CD monitoring signal. Fig. 2 shows the simulation results in a 40-Gb/s RZ system. When the CD is zero, the 40 GHz is the strongest. As the CD increases optical tone at to 42 ps, the tone is completely faded. III. EXPERIMENTAL SETUP
(2) Thus, the OSA spectrum around the probe wavelength is an approximation of the RF spectrum of the data signal. As the CD distorts the data, the bit-rate RF tone of the data fluctuates [2]. The new generated optical tones, which are the approximations of the bit-rate RF tone of the data, therefore change with CD and
Fig. 3 shows the experimental setup. The nonlinear medium is a 2-km spool of highly nonlinear fiber (HNLF) with the zerodispersion wavelength at 1552 nm. The dispersion slope and nonlinear coefficient of the HNLF are 0.045 ps/nm km and 9.1/W km, respectively. The transmitter generates a 50% duty nm. The probe is a concycle 40-Gb/s RZ signal at tinuous wave signal at 1552.01 nm. Variable amounts of dispersion are provided by varying lengths and types of fibers. Some of the data power is tapped out from the transmission link to provide the monitoring signal. After boosting by an erbium-doped fiber amplifier (EDFA), the data signal is coupled with the probe
432
Fig. 4.
IEEE PHOTONICS TECHNOLOGY LETTERS, VOL. 18, NO. 2, JANUARY 15, 2006
Optical tone power versus CD for 40-Gb/s RZ system.
Fig. 5. OSA measurements of optical tone power (a) residual CD = 0 ps/nm, (b) residual CD = 21 ps/nm, and (c) residual CD = 39 ps/nm.
and enters the HNLF. At the output of the HNLF, an OSA is used to measure the optical power of the new generated tones. Alternatively, we can use a bandpass tunable filter to select the optical and measure the optical power by a power meter. tone at IV. RESULTS AND DISCUSSION In our experiment, the 40-Gb/s RZ signal is transmitted through different lengths and types of fibers, and the power of the XPM-generated optical tone is measured. The average data power is amplified to 6 dBm and the probe power is 1 dBm before entering the HNLF. During the measurement, the polarization controller (PC) is adjusted to align the probe polarization with the data signal so that the measured optical power is maximized. Fig. 4 shows the simulation and experimental results for the effect of dispersion on the optical tone in the 40-Gb/s RZ system. The optical power changes by about 20 dB for chromatic dispersion varying from 0 to 42 ps/nm. The experimentation results agree well with the simulation results. Fig. 5(a)–(c) gives the measurements of the optical tone using OSA when the accumulated dispersion in the link is 0, 21, and 39 ps/nm, respectively. The optical tone is 26.34, 34.30, and 41.78 dBm, respectively. As shown in Fig. 4, the optical power of the tone fluctuates with the chromatic dispersion. It decreases until the accumulated chromatic dispersion reaches 42 ps/nm, and then increases until the accumulated CD reaches 84 ps/nm. Therefore, the monitoring window of this technique for a 40-Gb/s RZ system is 42 ps/nm. However, it is noted that, unlike the periodic RF tone fluctuation [2], the other peaks . If using of the optical tone power are lower than that at CD this optical tone power as the monitoring signal for feedback control of the dispersion compensation, then it might avoid
Fig. 6.
CD monitoring sensitivity versus data signal power.
the case that the total accumulated CD after compensation is nonzero. The maximized optical tone power can only be achieved at zero accumulated dispersion. transAs shown in (2), the optical tone power for CD mission link increases with the average data power launching into the HNLF. Therefore, the monitoring sensitivity using this technique is dependent on the data power. Better sensitivity can be obtained by increasing the data signal power. We experimentally measured the dependence of the monitoring sensitivity on the average data signal power entering the HNLF. The result is shown in Fig. 6. As the average data power increases, the monitoring sensitivity increases as well. When the data power is amplified to 6 dBm, the monitoring sensitivity is around 0.5 dB/ps/nm in the 40-Gb/s RZ system. V. CONCLUSION We experimentally demonstrated and analyzed an all-optical CD monitoring technique using an XPM-generated optical tone in an HNLF. The monitoring window of the technique is 42 ps/nm for 40-Gb/s RZ systems. This technique needs no modification at the transmitter and can potentially work on ultrahigh-speed optical systems. REFERENCES [1] M. Tomizawa, Y. Yamabayashi, Y. Sato, and T. Kataoka, “Nonlinear influence on PM-AM conversion measurement of group velocity dispersion in optical fibers,” Electron. Lett., vol. 30, no. 17, pp. 1434–1435, 1994. [2] Z. Pan, Q. Yu, Y. Xie, S. A. Havstad, A. E. Willner, D. S. Starodubov, and J. Feinberg, “Chromatic dispersion monitoring and automated compensation for NRZ and RZ data using clock regeneration and fading without adding signaling,” in Proc. Conf. Optical Fiber Commun., vol. 3, Anaheim, CA, Mar. 2001. Paper WH5-1. [3] Q. Yu, Z. Pan, L.-S. Yan, and A. E. Willner, “Chromatic dispersion monitoring technique using sideband optical filtering and clock phase-shift detection,” J. Lightw. Technol., vol. 20, no. 12, pp. 2267–2271, Dec. 2002. [4] T. E. Dimmick, G. Rossi, and D. J. Blumenthal, “Optical dispersion monitoring technique using double sideband subcarriers,” IEEE Photon. Technol. Lett., vol. 12, no. 3, pp. 900–902, Mar. 2000. [5] S. Li and D. V. Kuksenkov, “A novel dispersion monitoring technique based on four-wave mixing in optical fiber,” IEEE Photon. Technol. Lett., vol. 16, no. 3, pp. 942–944, Mar. 2004. [6] P. S. Westbrook, B. J. Eggleton, G. Raybon, S. Hunsche, and T. Her, “Measurement of residual chromatic dispersion of a 40-Gb/s RZ signal via spectral broadening,” IEEE Photon. Technol. Lett., vol. 14, no. 3, pp. 346–348, Mar. 2002. [7] C. Dorrer and D. N. Maywar, “RF spectrum analysis of optical signals using nonlinear optics,” J. Lightw. Technol., vol. 22, no. 1, pp. 266–274, Jan. 2004. , “Ultra-high bandwidth RF spectrum analyzer for optical signals,” [8] Electron. Lett., vol. 39, no. 13, pp. 1004–1005, 2003.
