Open Access
Add to BibSonomyAbstract
The requirements for higher data rates in optical communication systems lead to the use of more efficient modulation formats. In the networks the all optical synchronization and storage of these signals is still a major challenge in order to enable higher transmittable data rates and reduce the energy consumption. In this contribution we show for the first time, to the best of our knowledge, the tunable storage of phase modulated optical data packets with up to 60 pulse widths. This opens the way to the optical storage of data packets modulated with highly efficient modulation formats.
© 2012 Optical Society of America
1. Introduction
In recent years the requirements on existing communications networks have increased steadily due to the demand for higher data rates. Therefore, among other enabling technologies, advanced optical modulation formats have become key to the design of modern fiber based communication systems [1]. All optical buffering and data storage for applications like packet synchronisation, label processing and contention management is still a major bottleneck in optical communication networks. Conventional solutions need to convert the optical packet into an electrical signal and convert it back after processing. This double conversion increases the complexity, restricts the bandwidth and is power consuming. Therefore, by eliminating optical-electronic-optical conversion, the transmission speed of the network can be improved and the energy consumption can be reduced [2]. Hence, several methods for the delay of optical pulses have been shown. Among them are especially slow light systems which use the optically controlled change of the group velocity to slow down or advance the optical pulses. Slow-light buffers based on stimulated Brillouin scattering (SBS) can provide variable delay as a function of the pump power [3, 4]. But, unfortunately they typically have a rather low delay-bandwidth product [5]. The delay can be enhanced by modifying the shape of the Brillouin gain [6]. But the maximum delay of 4 Bit, that can be achieved with slow-light buffers, is still rather low. Additionally, almost all of these experiments are restricted to amplitude shift keyed signals. However, for increasing the spectral efficiency of modulation especially the phase of the signal is modulated. Optical delay of phase-modulated signals has been demonstrated experimentally in [7]. A 10.7 Gb/s signal encoded in a differential-phase-shift-keying (DPSK) format was delayed by up to 42 ps or around 0.5 pulse widths using broadband SBS-based slow-light.
A very promising method for the storage of optical packets is the so called Quasi-Light-Storage (QLS) [8, 9]. The QLS is based on the frequency sampling of the packet spectrum and can be used for all optical, tunable storage of data packets in optical fibers. The QLS works with standard telecommunication equipment, requires rather low optical power, works in the entire telecommunications range, and enables storage capacities of several thousand bits. For amplitude modulated signals the mechanism of QLS is well known, was further developed and results in almost distortion free storage of data packets up to 140 ns [10].
In this article we apply the QLS for the all-optical tunable storage of 8 Bit 1 Gbps binary phase shift keying (BPSK) modulated data packets. In our proof of concept experiments we achieved maximum storage times of around 60 ns, showing that the QLS can be incorporated for the buffering of data packets modulated with spectraly efficient modulation formats.
2. Theory
Advanced, spectrally efficient optical modulation formats have become a key ingredient to the design of modern communication systems, because it is possible to minimize both the linear and the nonlinear impairments over the transmission fiber, e.g. chromatic dispersion and four-wave mixing. In optical communications especially phase encoded signals are becoming evermore important due to their potential for increased receiver sensitivity, tolerance to various fiber impairments and better spectral efficiency [11]. The BPSK for a 11001101 data packet can be seen in Fig. 1 on the left side. In the frequency domain the BPSK can be characterized by its spectral density, represented by the dashed line in Fig. 1 on the right side.
Fig. 1 Representation of the BPSK modulated data packet in time (left side) and frequency domain (right side). The dashed line shows the power spectral density function of the BPSK modulated packet whereas the solid line shows the QLS applied to the spectrum. The frequency components resulting from the packet length are not shown.
The main idea of QLS is based on the duality between the time and frequency representation of each signal. A signal in the time domain equals a spectral distribution in the frequency domain. If the spectrum is multiplied by a narrow spaced frequency comb, the result is a sampled version of the spectrum, as can be seen in Fig. 1 on the right side. The multiplication in the frequency domain leads to a convolution in the time domain and - if we neglect the energy restrictions - to a theoretically unlimited train of copies of the original signal in the time domain [12]. One of these copies is extracted and the signal is quasi-stored. The multiplication of the frequency comb with the spectrum in the frequency-domain is carried out via SBS [13, 14] in a standard single mode fiber (SSMF). In principle, a spectral narrow comb filter is generated. The filter bandwidth of each line and therefore the maximum storage time depends on the bandwidth of SBS. The maximum data packet length that can be stored is restricted to the distance between the frequency components in the comb, and therefore the distance of the copies.
SBS itself is a nonlinear effect, which occurs through an interaction between an incident pump wave and an acoustic wave in the medium, e.g. an optical fiber. The acoustic wave is generated through the pump wave and a part of the pump power is scattered back. Due to the relative speed between pump and acoustic wave, this results in a gain and loss region inside the fiber with a frequency shift to the pump wave of ± fSBS. Therefore, a counter propagating signal wave can be amplified or attenuated within the gain and the loss region. This behavior can be theoretically explained by three coupled differential equations [15]. All the described interactions depend on the complex phase mismatch term Δk, which can be written as:
with ΔfB as the Brillouin gain Bandwidth, fSBS as the Brilloiun shift, αa as the attenuation coefficient and va as the velocity of the acoustic wave with 5.96 km/s [16]. The real part describes the bandwidth in which the Brillouin interaction takes place and the imaginary part determines the accompanied phase shift. For the used AllWave fiber we measured a Brillouin gain bandwidth of around 12 MHz at room temperature for a pump wavelength of 1550 nm. Contrary to SBS based Slow Light, for the QLS the low bandwidth is an advantage since it enables higher storage times [12]. The normalized real and imaginary parts of the Brillouin gain according to Eq. (1) are shown in Fig. 2.As can be seen, in the center of the Brillouin gain the phase change is zero. Since the QLS extracts narrow frequency components out of the spectrum, the phase information of the signal is not changed. However, even if the whole bandwidth of the signal is affected by the SBS like in a Slow-Light system, the phase in phase modulated signals is preserved [7]. Hence, the SBS based QLS can be utilized for the all optical storage of spectral efficient phase modulated signals.
3. Experiment
The theoretical predictions are verified within a proof of concept experiment. The experimental setup can be seen in Fig. 3. The data pattern is generated with an arbitrary waveform generator (AWG). The signal is converted into the optical domain with a DFB laser diode (LD1) at 1550 nm and a phase modulator (PM). The PM is driven at Vπ in order to achieve the best modulation results. Additionally, an isolator (not shown) is used to protect the equipement from destruction by the pump wave, before the data signal is coupled into a 20 km AllWave fiber. The AllWave fiber is used due to the lower SBS bandwidth. The pump wave is coupled into the fiber from the opposite side with a circulator (C). The electrical signal for the frequency comb is generated by the AWG as well and is transferred into the optical domain with a Mach-Zehnder modulator (MZM1) and LD2. The signal power and bias voltage for MZM1 are adjusted to the point where the frequency comb is almost flat. The comb consists of 10 branches to cover the spectrum of the signal. The optical frequency comb is amplified by an erbium doped fiber amplifier (EDFA) to provide sufficient pump power for the QLS.
Fig. 3 Experimental setup for the storage of phase modulated signals. LD: laser diode, PM: phase modulator, AWG: arbitrary waveform generator, MZM: Mach-Zehnder modulator, EDFA: erbium doped fiber amplifier, C: circulator, LO: local oscillator, Osci: oscilloscope, OSA: optical spectrum analyzer.
The resulting copies of the QLS are combined with a local oscillator (LO), in form of a fiber laser, by a 3 dB coupler in order to perform coherent optical detection [17]. The polarization matching between the LO and the received signal, as well as the relative phases of the signal and the LO are ensured. Afterwards the desired copy is extracted with MZM2 which is driven by a rectangular signal generated by the AWG. The final signal is splitted by an 90/10 coupler and indicated with an oscilloscope. The oscilloscope works with an average of 4 times. With help of the optical spectrum analyzer (OSA) the correct wavelenghts of the lasers are monitored.
4. Results
Within the experiment we have used a 8 Bit pattern with the sequence 11001101 at a bitrate of 1 Gbps. The measurement results can be seen in Fig. 4. The reference pattern without QLS can be seen on the left side (black) as well as the stored packets. We were able to achieve a storage time of up to 60 ns with acceptable distortions which are displayed in detail in Fig. 5. According to the theory, the delays are within an envelope [8]. The width of the envelope is proportional to the inverse of the width of the SBS gain spectrum. The SNR, calculated as 10log(Psignal/Pnoise) with Psignal as the average power of the signal and Pnoise as the average noise power, for the specific delays can be found in Fig. 4. Like the delays themselves, it follows the mentioned envelope.
Fig. 4 Measurement results for a 11001101 BPSK modulated bit sequence with the reference signal on the left side (black) and the different extracted copies of the SBS based QLS.
Fig. 5 The extracted patterns overlapped together with the reference (black line) in order to highlight the distortions.
The existing distortions can be addressed to laser drifts and changes due to the modulator over the time. Thus, the shape of the frequency comb and the power distribution of the several lines is changed. Also small variations of the polarization during the measurement can cause distortions. Another non-ideal behavior is the finite SBS bandwidth, which limits the maximum achievable storage time. During the measurement, the maximum available pump power was 20 dBm at the fiber input. Therefore the pump power was not sufficient enough to reduce the gain down to 11 MHz in an AllWave fiber. This results in a limited storage time of 60 ns. The copies after this point disappear in the noise floor. To overcome this limitation higher pump powers should be used as well as a reduced Brillouin gain bandwidth [18, 19]. With this improvement much higher storage times should be possible. Additionally, the width of the comb was restricted to 0.9 GHz in our experiment. Therefore, not the whole bandwidth of the spectrum can be sampled correctly and some distortions occur. The frequency comb can be broadened, e.g. with an additional modulator. Since the QLS is simply a sampling in the frequency domain, with an ideal flat comb and sufficient bandwidth the QLS is distortion free. The measurement shows that the QLS works well for BPSK modulated signals. Therefore, the results can be generalized for other spectral efficient modulation formats, like differential phase shift keying, as well. Further investigations with spectral efficient modulation formats and a mixed amplitude and phase modulation like QAM are necessary.
5. Conclusion
In conlcusion we have shown that the QLS can be applied for the all optical storage of phase encoded signals. This opens the way for the all optical tunable storage of data packets modulated with spectrally efficient formats. The theory of SBS based QLS could be validated for a DPSK modulated signal. Within the experiment we used a packet with the sequence 11001101 at a bitrate of 1 Gbps. The stored signals show rather low distortions. With moderate pump powers a maximum storage time of 60 ns or 60 pulse widths was achieved. We believe that the absolute storage time could be easiliy enhanced with higher pump powers and/or a reduction of the SBS bandwidth. For higher bit rate data packets with much higher relative storage times should be possible. For 10 Gbps DPSK data packets the 60 ns would result in a relative storage time of 600 Bits. For 20 Gbps DPSK or 10 Gbps QPSK modulated packets this would result in 1200 Bit and so on.
Acknowledgments
The authors would like to acknowledge the financial support of the German Research Foundation (reference number: SCHN 716/6-2). Additionally the authors would like to thank K. Jamshidi, A. Wiatrek and J. Klinger of HfT Leipzig.
References and links
1. P. J. Winzer and R.-J. Essiambre, “Advanced optical modulation formats,” Proc. IEEE 94, 952–985 (2006). [CrossRef]
2. R. S. Tucker, “Green optical communicationspart II: energy limitations in networks,” IEEE J. Sel. Top. Quantum Electron. 17, 261–274 (2011). [CrossRef]
3. Y. Okawachi, M. S. Bigelow, J. E. Sharping, Z. M. Zhu, A. Schweinsberg, D. J. Gauthier, R. W. Boyd, and A. L. Gaeta, “Tuneable all-optical delays via Brillouin slow light in an optical fiber,” Phys. Rev. Lett. 94, 153902–153906 (2005). [CrossRef] [PubMed]
4. K. Y. Song, M. G. Herráez, and L. Thévenaz, “Observation of pulse delaying and advancement in optical fibers using stimulated Brillouin scattering,” Opt. Express 13, 82–88 (2005). [CrossRef] [PubMed]
5. T. Schneider, “Time Delay limits of stimulated-Brillouin-scattering based slow light systems,” Opt. Lett. 33, 1398–1400 (2008). [CrossRef] [PubMed]
6. T. Schneider, M. Junker, and K.-U. Lauterbach, “Time delay enhancement in stimulated Brillouin scattering based slow light systems,” Opt. Lett. 32, 220–222 (2007). [CrossRef] [PubMed]
7. B. Zhang, L. Yan, I. Fazal, L. Zhang, A. E. Willner, Z. Zhu, and D. J. Gauthier, “Slow light on Gbps differential-phase-shift-keying signals,” Opt. Express 15, 1878–1883 (2007). [CrossRef] [PubMed]
8. K. Jamshidi, S. Preussler, A. Wiatrek, and T. Schneider, “A review to the all-optical quasi-light storage,” IEEE J. Sel. Top. Quantum Electron. 18, 884–890 (2012). [CrossRef]
9. T. Schneider, K. Jamshidi, and S. Preussler, “Quasi-light storage: a method for the tunable storage of optical packets with a potential delay-bandwidth product of several thousand bits,” J. Lightwave Technol. 28, 2586–2592 (2010). [CrossRef]
10. S. Preussler, A. Wiatrek, K. Jamshidi, and T. Schneider, “Quasi-light-storage enhancement by reducing the Brillouin gain bandwidth,” Appl. Opt. 50, 4252–4256 (2011). [CrossRef] [PubMed]
11. A. H. Gnauck and P. J. Winzer, “Optical phase-shift-keyed transmission,” J. Lightwave Technol. 23, 115–130 (2005). [CrossRef]
12. S. Preussler, K. Jamshidi, A. Wiatrek, R. Henker, C. Bunge, and T. Schneider, “Quasi-Light-Storage based on time-frequency coherence,” Opt. Express 17, 15790–15798 (2009). [CrossRef] [PubMed]
13. T. Schneider, Nonlinear Optics in Telecommunications (Springer-Verlag, 2004).
14. A. Kobyakov, M. Sauer, and D. Chowdhury, “Stimulated Brillouin scattering in optical fibers,” Adv. Opt. Photon. 2, 1–59 (2010). [CrossRef]
15. T. Schneider, M. Junker, and K.-U. Lauterbach, “Theoretical and experimental investigation of Brillouin scattering for the generation of Millimeter waves,” J. Opt. Soc. Am. B 23, 1012–1019 (2006). [CrossRef]
16. D. Cotter, “Stimulated Brillouin scattering in monomode optical fiber,” J. Opt. Commun. 4, 10–19 (1983). [CrossRef]
17. E. Voges and K. Petermann, Handbuch der Optischen Kommunikationstechnik (Springer-Verlag, 2002).
18. S. Preussler, A. Wiatrek, K. Jamshidi, and T. Schneider, “Brillouin scattering gain bandwidth reduction down to 3.4MHz,” Opt. Express 19, 8565–8570 (2011). [CrossRef] [PubMed]
19. S. Preussler, A. Wiatrek, K. Jamshidi, and T. Schneider, “Quasi-light-storage enhancement by reducing the brillouin gain bandwidth,” Appl. Opt. 50, 4252–4256 (2011). [CrossRef] [PubMed]
