A high spectral efficiency scheme for all optical label switching (AOLS) network is introduced. The vestigial sideband (VSB) payload with wavelength labeling is proposed in beyond 40Gb/s AOLS networks. Two different label wavelength schemes are considered. The minimum 20dB channel bandwidth of payload and label is 0.40nm in 43Gb/s systems which confirm the payload power penalty less than 1dB at the receiver.
©2006 Optical Society of America
The explosion of Internet traffic demands the next generation optical network to support packet routing and forwarding operation at Terabits wire rates. All-optical label switching (AOLS) technology  which has the advantage of low latency, payload transparency, high scalability and flexibility is proposed in future optical networks. The important issue of AOLS system is the method of coding the label onto the packet, as it directly determines the structure and performance of AOLS nodes as well as the channel bandwidth and transmission quality of packet and its label. The spectral efficiency is often considered as main issue in WDM system with the bit rate of 40Gb/s or beyond.
There are some potential methods can to be used for high spectral efficiency transmission in AOLS. Some orthogonal modulation formats, such as DPSK/ASK, FSK/ASK, are introduced in [2–4], which is limited by the extinction ratio (ER) of payload and label. A method using optical carrier suppression and separation is demonstrated in 40Gb/s [5–6]. With wavelength conversion technique, multi-node label swapping is also shown in . However, the bandwidth of payload and label is still too large for 40Gb/s WDM system.
A high spectral efficiency scheme using vestigial sideband modulation (VSB) payload is given out to realize 40Gb/s all-optical label switching with optical wavelength labeling in . In this paper, two different label wavelength schemes are considered. The minimum 20dB channel bandwidth of payload and label is 0.40nm in 43Gb/s systems which confirm the payload power penalty less than 1dB at the receiver. This scheme can make the total channel bandwidth of payload and label less than 0.5nm in 43Gb/s bit rate, which can pass the traditional optical network devices, such as MUX/DMUX, with low interchannel crosstalk.
Because CSRZ is suitable for vestigial sideband modulation and the vestigial sideband CSRZ format also has the good performance for optically-routed networks . So in our scheme the payload signal is modulated as the CSRZ format firstly and then filtered by a tunable optical filter (TOF) to be VSB-CSRZ format. The experiment setup is shown in Fig. 1. The carrier-suppressed return-to-zero (CSRZ) signal generator is combined with two Mach-Zender modulators (MZM). The optical signal from a continuous wave laser (CWL) is modulated by the first MZM biased at Vπ/2 to generate the 43Gb/s non-return-to-zero (NRZ) code. The center wavelength of CWL is 1557.36nm. And the second MZM is biased at the minimal intensity-output point and driven with a 21.5GHz sine-clock signal. The phase deviation θ of the two modulator arms equals to π. Thus the output of MZM2 is 43Gb/s CSRZ code.
Another optical signal from a tunable CWL (T-CWL) is sent into MZM3. The range of the T-CWL is from 1520nm to 1590nm. The label is modulated as NRZ code with 1.25Gb/s label signal. Then the payload and label are combined together by an optical coupler. In the receiver node, a tunable fiber Bragg grating (T-FBG) is employed to separate the payload and label. The FBG is exactly tuned to filter out the payload signal and then detected by 43Gbps BER tester with an optical preamplifier. After EDFA there is an optical BPF to block amplified spontaneous emission (ASE) noise.
The interaction between the label and payload is a crucial consideration of this scheme. Due to the low bit rate of label signal, the payload suffers more impairment. So in this letter, we just focus on the performance of optical payload signal.
3. Results and discussion
3.1 Channel bandwidth
The wavelength difference between the VSB payload and the label is a key issue in this scheme. If the label wavelength λLabel is very close to the carrier wavelength λPayload, the payload signal quality will deteriorate. That is because the label signal may overlap the vestigial band of payload. Furthermore when label is separated from payload, the optical filter ramp will distort the payload signal. The narrower the wavelength difference between the label and the payload, the greater this distortion is. On the other hand, if the label wavelength is far from the carrier, the channel bandwidth will increase which will cause much interchannel crosstalk in high speed WDM systems. So there is a trade-off between the signal quality and channel bandwidth. Figure 2 shows the spectra of VSB-CSRZ payload and eye diagram. We define the spectra spacing from the top frequency point falling down to 20dB as the effective channel bandwidth (BW), for this part concentrates the most energy of the optical signal. It’s shown in Fig. 2 that the 20dB BW of VSB payload is 0.34nm, the BW range is from 1557.28 to 1557.62nm. Figure 3 shows the spectra of payload before and after short wavelength label being erased, and the corresponding eye diagrams are inserted. The 20dB channel BW of payload and label can be expressed as:
Generally speaking, this BW confirms the high speed optical signal to pass through the WDM network devices, such as MUX/DUMX, with little distortions. Figure 4 shows the spectra of payload before and after long wavelength label being erased, and the corresponding eye diagrams are inserted. The 20dB channel BW of payload and label can be expressed as:
3.2 Comparison of different label wavelengths
The relative spectra position of payload and label is also studied experimentally. When the label wavelength is placed in the low frequency or high frequency side of payload, the optical filtering influence at the receiver is different, especially for the VSB payload whose spectra is asymmetry. So in our experiment, the label wavelength is tuned from 1557.20–1557.30nm and 1557.62–1557.70.nm to test the performance of the received payload with different label wavelengths. First we change the short label wavelength as 1557.16nm, 1557.18nm, 1557.20nm, 1557.22nm and 1557.24nm respectively. The BER curves of payload after label erasing in different label wavelengths are shown in Fig. 5. The payload power penalty at BER=10-12 is less than 1dB when the label wavelength equal or less than 1557.18nm. We also change the long label wavelength as 1557.62nm, 1557.64nm, 1557.66nm, 1557.68nm and 1557.70nm respectively. The BER curves of payload after label erasing in different label wavelengths are shown in Fig. 6. The payload power penalty at BER=10-12 is less than 1dB when the label wavelength equal or more than 1557.68nm. Comparing the short label wavelength and long label wavelength schemes, the later one has narrower channel bandwidth with low power penalty than the former case. Figure 7 shows the power penalty versus channel bandwidth. If considering 1dB as a proper value of power penalty, the channel bandwidth of long wavelength scheme is 0.05nm less than the short wavelength scheme. Moreover, at the same bandwidth, such as 0.4nm, the power penalty of long wavelength scheme is 3dB lower than the short wavelength scheme. So in order to effectively reduce the total channel bandwidth with good payload reception performance, the long wavelength scheme is the better choice in DWDM optical networks.
In order to increase the spectral efficiency in beyond 40Gb/s AOLS networks, we proposed a VSB payload scheme combined with wavelength labeling. This scheme can effectively reduce the total channel bandwidth of payload and label. Two cases of short and long label wavelengths are experimentally demonstrated. The results show that long label wavelength case is more suitable for low payload power penalty. This scheme is proposing to be employed in the future high speed DWDM optical networks.
Reference and Links
1. D. J. Blumenthal, B. E. Olsson, G. Rossi, T. E. Dimmick, L. Rau, M. Masanovic, O. Lavrova, R. Doshi, O. Jerphagnon, J. E. Bowers, V. Kaman, L. A. Coldren, and J. Barton, “All-optical label swapping networks and technologies,” J. Lightwave Technol. 18, 2058–2075 (2000). [CrossRef]
2. X. Liu, Y. Su, X. Wei, J. Leuthold, and R. C. Gile, “Optical-label switching based on DPSK/ASK modulation format with balanced detection for DPSK payload,” Eur. Conf. Optical Communication (ECOC 2003), Rimini, Italy, Paper Tu4.4.3 (2003).
3. N. Chi, J. Zhang, P. V. Holm-Nielsen, C. Peucheret, and P. Jeppesen, “Optical label swapping and packet transmission based on ASK/DPSK orthogonal modulation format in IP-over-WDM networks,” in Optical Fiber Communication Conf. Tech. Dig., Postconference Edition, FS2 (2003).
4. J. Zhang, N. Chi, P. V. Holm-Nielsen, C. Peucheret, and P. Jeppensen, “A novel optical labeling scheme using a FSK modulated DFB laser integrated with an EA modulator,” in Optical Fiber Communication Conf. Tech. Dig., Postconference Edition, TuQ5 (2003).
5. Jianjun Yu and Gee-kung Chang, “A Novel Technique for Optical Label and Payload Generation and Multiplexing Using Optical Carrier Suppression and Separation,” IEEE Photonics Technol. Lett. 16, 320–322 (2004). [CrossRef]
6. G. K. Chang and J. Yu, “40 Gbit/s payload and 2.5 Gbit/s label generation using optical carrier suppression and separation,” Electron. Lett. , 40, 442–444 (2004). 8slkdjf [CrossRef]
7. J. Yu, G.-K. Chang, and Q. Yang, “Optical Label Swapping in a Packet-Switched Optical Network Using Optical Carrier Suppression, Separation, and Wavelength Conversion,” IEEE Photonics Technol. Lett. , 16, 2156–2158 (2004). [CrossRef]
9. A. Agarwal, S. Chandrasekhar, and R.-J. Essiambre, “VSB-CSRZ for Spectrally Efficient Optically-Routed Networks,” ECOC 2004 Proceedings, vol.3, Paper We3.4.4 (2004).