Open Access
Add to BibSonomyAbstract
In this paper, field trial on optical code division multiplexing through the commercial-used fiber line is presented. We fabricated fiber Bragg grating en/decoders with time-spreading and wavelength-hopping scheme, considering environmental fluctuation of transmission fiber. 200 km-long transmission of 2-channel ×10 Gb/s signals was achieved on the field photonic network test bed of JGN II. Error free transmission was demonstrated in real field deployed single-mode transmission fiber.
©2006 Optical Society of America
1. Introduction
Optical code division multiplexing (OCDM) has attracted much attention as a promising technology for photonic packet switching and the optical metro and local area networks (MAN/LAN) system applications [1], due to its all-optical signal processing, flexible capacity, and highly secured transmission. Many techniques on code construction and its implementation to encoder device, and interference elimination have been proposed for enhancement on transmission rate and the number of channels. Total systems with transmitters and receivers using such techniques have been demonstrated in the past few years [2–4]. However, experiments using transmission line for commercial use have been discussed insufficiently. In such a transmission line, other factors of signal degradation are considered due to fluctuation of environmental condition, such as, variation of polarization state, timing jitter and signal drift of the transmitted signal through fiber. For practical use, their influences should be considered to design the real system.
In various OCDM systems proposed so far, OCDM with time-spreading and wavelength-hopping scheme [5] has expected as a simple and powerful solution. Fiber-Bragg-grating (FBG) devices are candidates of encoders incorporated in this scheme, because of features of simple structure, low loss and low polarization sensitivity. We have developed the OCDM using in-house FBG as en/decoder [6, 7]. In this paper, we will report a field trial of timespreading and wavelength-hopping OCDM on 10 Gb/s signals to verify the application to MAN. 200 km-long transmission of 2-channel multiplexing signal was achieved through the installed fiber.
2. System configuration
The OCDM field trial was demonstrated on the field photonic network test bed of JGN II [8] as shown in Fig. 1. The test bed link, as shown in Fig. 1(a), consisted of two spans of standard single-mode fiber (SMF) between Otemachi Reserarch Center in Tokyo and the Tsukuba Research Center, via Kashiwa Relay Station. All spans between the research centers and the relay station were about 50 km. We utilized the link in loop-back configuration, that is, four spans (Span 1–4) with the total link length of 200 km. Four optical amplifiers were employed to compensate power loss on the transmission line. Dispersion- and slope-compensating fiber (DCF) was placed in each repeater. The residual dispersion of the transmission line was approximately 0 ps/nm. The total output power of each repeater was optimized to 10 dBm.
Figures 1(b) and 1(c) show OCDM-transmitter and -receiver for 2-channel ×10 Gb/s OCDM transmission in Otemachi Research Center, respectively. The channel of 10 Gb/s was implemented at data rate enhancement scheme with 4 time-shifted codes at 2.5 Gb/s [6]. Then 10 Gb/s channel was encoded using the same FBG encoder as that for encoding of 2.5 Gb/s signals. Time-gating utilizing electro-absorption modulator (EAM) [6] was adopted as co-channel interference elimination. To gate signal after 200km-long transmission, clock signal for gating was extracted from the transmitted signal in the receiver.
The OCDM-transmitter consisted of a multi-wavelength source, an optical pulse generator, an EAM, FBG encoders, erbium-doped fiber amplifiers (EDFAs), and amplified spontaneous emission (ASE) rejection filters. The multi-wavelength source and the optical pulse generator were employed to generate 10 GHz optical pulse train, which had five carrier wavelengths of 1556.8–1560.1 nm with 0.8 nm spacing. The optical pulse train was amplified by an EDFA and modulated by the EAM to a return-to-zero (RZ) formatted signal with a 223-1 pseudorandom binary sequence (PRBS) at 10 Gb/s. The signal was split into two, then two-channel signals were obtained. One of the channel signals was delayed to create two pseudo-independent data. FBG encoders encoded the signals with time-spreading and wavelength-hopping patterns. The two encoded signals are combined and amplified by the EDFA to compensate power loss on the code multiplexing process.
The OCDM-receiver consisted of an FBG decoder, a coupler, an EDFA, a clock extraction circuit, an EAM, and a conventional 3R receiver. Transmitted signal was launched into the FBG decoder, which had the same code pattern as one of the encoders. The decoded signal was split into two by the coupler. The EAM, which was concatenated to one of the output ports of the coupler, time-gated the decoded signal to eliminate interference channel noise. Clock signal driving the EAM was obtained by the clock extraction circuit which was concatenated to another output ports of the coupler. In this way, OCDM signal was de-multiplexed by decoding and time-gating. The conventional 3R receiver converted the de-multiplexed optical signal into electrical signal for error detection.
Fig. 1. System configuration of field trial for 2-channel ×10 Gb/s OCDM transmission; (a) field photonic network test bed of JGN II, (b) OCDM-transmitter, and (c) OCDM-receiver.
3. FBG based encoder and its design
FBGs in the en/decoders were composed of five-chirped gratings in the optical fiber based on the prime-hop sequences of length 25. The chip duration, Tc, was 16 ps, and thus the chip rate was 62.5 Gchip/s. The spreading time of encoded signal, Ts, was 400 ps. The reflectivity wavelengths of the FBGs and the spacing were the same as that of the multi-wavelength light source. Figures 2(a) and 2(b) show the code patterns of encoders for desired and undesired channels, respectively, in two-dimensional matrix description. Rows and columns of the matrices represent wavelength bins and time slots in the unit of Tc. Filled and opened squares represent presence and absence of chip pulses with the corresponding wavelength bins and time slots. In the figures, λ0 is the longest center wavelength of the chip pulses and λi=λ0-iΔλ, where Δλ is wavelength spacing. T 0 is the lowest delay through the encoder and Tj=T 0+jTc. Since Ts is 4-time larger than the basic signal period of 100ps, effects of inter-symbol interference (ISI) and beat noise arises [9]. As mentioned in Ref. [9], they are caused by overlapping sidelobes of decoded waveforms from a desired channel and undecoded waveforms from undesired channels into a correctly decoded symbol in coherent time-spreading scheme. In time-spreading and wavelength-hopping scheme, the above effects are originated from the only overlapping between the correctly decoded symbol and the undecoded waveforms, since the sidelobes are not generated in the decoded waveform from the desired channel. Then, we selected the code shown in Fig. 2(b) as an undesired channel to minimize the overlapping between a correctly decoded symbol and undecoded waveforms.
The FBGs were athermal devices by passively compensating the variation of the gratings in temperature changes [10]. The maximum variation of the FBGs on reflectivity wavelength was 0.05 nm over the range of 0 °C to 80 °C.
Index profiles of all gratings were designed to optimize the reflectivity spectral widths of the gratings, so that the reflectivity spectrum width of each grating was wider than width of each spectral component composing signal, and narrower than wavelength-hopping spacing. Especially, it was enough wide to tolerate variation of source wavelength. Lengths of the gratings, which determined both temporal and spectral mainlobe width of reflected pulses, were 3 mm. The apodization technique was applied in FBGs to suppress of pulse distortion due to reflection from spectrum sidelobes [11].
Figures 3(a), 3(b), and 3(c) show measured reflectivity and group delay (GD) of the FBG encoder for desired channel, the FBG encoder for undesired channel, and the FBG decoder, respectively. The values of GD represent relative values to GD at λ0. Full width at half maximum and peak reflectivity of the reflection spectrum for each grating was 0.5 nm and -3 dB, respectively. Table 1 shows averages of relative GD obtained by averaging the measured values of relative GD, shown in Fig. 3, within a 3dB reflectivity bandwidth for each spectral component. By dividing the averages by chip duration of 16 ps, it was confirmed that all FBGs implemented code patterns shown in Fig. 2. Also, as designed for the desired channel, results from adding the average for the desired channel encoder to that for the decoder in every spectral component were all zero. Figures 4(a) and 4(b) show measured polarization dependent loss (PDL) and differential group delay (DGD) of the encoder, respectively. The PDL and DGD were less than 0.2 dB and 5 ps in the range of 3 dB reflectivity bandwidth of the FBG. Similar results were obtained for another FBG encoder and decoder.
Fig. 3. Measured reflectivity and group delay (GD) of (a) FBG encoder for desired channel, (b) FBG encoder for undesired channel, and (c) FBG decoder.
Table 1. Averages of relative GD of all gratings; (a) FBG encoder for desired channel, (b) FBG encoder for undesired channel, and (c) FBG decoder.
Fig. 4. Measured polarization dependency of an FBG encoder; (a) polarization dependent loss (PDL), and (b) differential group delay (DGD).
4. Experimental results of OCDM transmission
Figure 5 shows measured eye diagrams. Figure 5(a) shows the eye diagram of initial 10Gb/s RZ formatted signal measured after the EAM in the OCDM-transmitter. Pulse width of the signal was 15 ps. Figure 5(b) shows the eye diagram of multiplexed signal before transmitting to the fiber of 200 km. The waveform was spread temporally without clear eye-opening, so it is difficult to recognize the original data of individual channel. Figure 5(c) shows the eye diagram of de-multiplexed signal in the OCDM-receiver after 200 km-long transmission. Clear eye-opening was observed.
Fig. 5. Eye diagrams of signals; (a) 10Gb/s PRBS signal before encoding, (b) encoded multiplexed signals without transmission, and (c) de-multiplexed signal after 200 km transmission with time gating.
Figure 6 shows measured bit error rate (BER) of received signals. The received power difference at the BER of 10-9 of the de-multiplexed signal from 2-channel OCDM without transmission from single-channel signal without encoding and transmission (back-to-back), which means power penalty due to OCDM, was 1dB. The received power difference of the de-multiplexed signal with 200 km transmission from that without transmission, which means power penalty due to 200 km-long transmnission, was almost 0 dB. In both cases of OCDM signal transmission, the value of BER was achieved to ~10-12 at the received power of -32dBm. Furthermore, we measured BER of de-multiplexed signals every 30 minutes for four hours at the received power of -32 dBm. Stable BERs were obtained within 2–4×10-12. From these results, our OCDM architecture is applicable to transmission though the installed fiber up to 200 km-long.
Fig. 6. Measured bit error rate. Filled circle: single-channel signal without encoding and transmission (back-to-back), filled triangle: De-multiplexed signal from 2-channel OCDM without transmission, and opened square: De-multiplexed signal from 2-channel OCDM with 200 km-long transmission.
5. Conclusion
We demonstrated field trial of time-spreading and wavelength-hopping OCDM using FBG en/decoder on 2-channel ×10 Gb/s signals on the field photonic network test bed of JGN II. The BER measurement results showed that the power penalty due to transmission was sufficiently small and 200 km-long transmission was achieved on installed fiber. These results indicate that our OCDM architecture is applicable to practical fiber-optic systems both in the LAN and MAN applications.
References and links
1. K. Kitayama, N. Wada, and H. Sotobayashi, “Architectural considerations for photonic IP router based upon optical code correlation,” J. Lightwave Technol. 18, 1834–1844 (2000). [CrossRef]
2. S.-J. Kim, T. J. Eom, T.-Y. Kim, B. H. Lee, and C.-S. Park, “Experimental demonstration of 2×10 Gb/s OCDMA system using cascaded long-period fiber gratings formed in dispersion compensating fiber,” in Proceedings of Optical Fiber Communications Conference/National Fiber Optic Engineers Conference (OFC/NFOEC) 2005, Paper JWA48 (2005).
3. Z. Jiang, D. S. Seo, S.-D. Yang, D. E. Leaird, R. V. Roussev, C. Langrock, M. M. Fejer, and A. M. Weiner, “Four-User, 2.5-Gb/s, Spectrally Coded OCDMA System Demonstration Using Low-Power Nonlinear Processing,” J. Lightwave Technol. 23, 143–158 (2005). [CrossRef]
4. V. J. Hernandez, W. Cong, R. P. Scott, C. Yang, N. K. Fontaine, B. H. Kolner, J. P. Heritage, and S. J. B. Yoo, “320-Gb/s Capacity (32 Users ×10 Gb/s) SPECTS O-CDMA Local Area Network Testbed,” Optical Fiber Communications Conference/National Fiber Optic Engineers Conference (OFC/NFOEC) 2006, PD Paper PDP45 (2006).
5. N. Wada, H. Sotobayashi, and K. Kitayama, “2.5 Gbit/s time -spread/wavelength-hop optical code division multiplexing using fibre Bragg grating with supercontinuum light source,” Electron. Lett. 36, 815–817 (2000). [CrossRef]
6. S. Kutsuzawa, N. Minato, S. Oshiba, A. Nishiki, and K. Kitayama, “10 Gb/s x 2ch signal unrepeated transmission over 100km of data rate enhanced time-spread/wavelength-hoping OCDM using 2.5 Gb/s-FBG en/decoder,” IEEE Photon. Technol. Lett. 15, 317–319 (2003). [CrossRef]
7. H. Tamai, H. Iwamura, N. Minato, and S. Oshiba, “Experimental study on time-spread wavelength-hop Optical code division multiplexing with group delay compensating en/decoder,” IEEE Photon. Technol. Lett. 16, 335–337 (2004). [CrossRef]
8. T. Miyazaki, M. Daikoku, I Morita, T. Otani, Y. Nagao, M. Suzuki, and F. Kubota, “Stable 160-Gb/s DPSK transmission using a simple PMD compensator on the field photonic network test bed of JGN II,” in Proceedings of 9th Optoelectronics and Communications Conference/3rd International Conference on Optical Internet (OECC/COIN 2004), Paper PD1-3 (2004).
9. X. Wang, N. Wada, and K. Kitayama, “Inter-symbol interference and beat noise in flexible data-rate coherent OCDMA and the BER improvement by using optical thresholding,” Opt. Express 13, 10469–10474 (2005) [CrossRef] [PubMed]
10. G. W. Yoffe, P. A. Krug, F. Ouellette, and D. Thorncraft, “Temperature-compensated optical-fiber Bragg gratings,” in Proceedings of Optical Fiber Communications Conference (OFC) ‘95, Paper WI4 (1995).
11. H. Iwamura, N. Minato, H. Tamai, S. Oshiba, and A. Nishiki, “FBG based optical code en/decoder for long distance transmission without dispersion compensating devices,” in Proceedings of Optical Fiber Communications Conference (OFC) 2004, Paper WK6 (2004).
