Open Access
Add to BibSonomyAbstract
An RF-assisted 112 Gbit/s transmitter based on optical phase modulation is presented. The system uses two closely spaced sub-channels generated using radio frequency (RF) electronics. Numerical simulations as well as experiments over 824 km of installed fiber are used for evaluation. The analysis shows that performance almost as good as for a conventional Mach-Zehnder modulator can be obtained. Although a relatively high OSNR is required due to the large fraction of power residing in the optical carrier, very small penalty is observed for optimum power levels in the transmission link.
©2011 Optical Society of America
1. Introduction
Advanced modulation formats, such as QPSK and 16-QAM, are now considered indispensable for next generation optical communication systems in order to increase spectral efficiency in the transmission link and maintain a reasonably low baudrate for electronic processing. Dual polarization (DP)-QPSK is already in practical use within 40G and 100G transceivers [1,2] and 16-QAM will probably be required in order to increase the channel capacity and spectral efficiency further [3–5]. Usually these systems employ nested Mach-Zehnder modulators (MZM) for electro-optical conversion since these modulators allow independent modulation of the inphase (I) and quadrature (Q) components of the optical field. However, these modulators are fairly complex and today have a conversion bandwidth limited to about 30 GHz.
In order to enable low-cost implementation of transceivers for advanced modulation formats, we have recently proposed the use of microwave electronic modulation of the data on radio frequency (RF) carriers before converting the electronic signal into the optical domain using conventional electro-optical amplitude modulation techniques, such as MZMs [6,7] or even electro-absorption modulators. For low-cost implementation the laser and modulator should preferably be integrated [8]. However, also the conventional MZM is a fairly complicated device to monolithically integrate with the laser.
In this paper we demonstrate and evaluate the use of an optical phase modulator for electro-optical modulation of the RF sub-channels onto a laser carrier. An optical phase modulator constitutes the simplest possible optical modulator structure since it only relies on the Pockels-effect [9] where the refractive index depends linearly on the applied electrical field and thus the phase of the optical field becomes linearly dependent on the applied electrical drive signal. In addition, for a phase modulator no bias control is required, which further reduces cost and complexity of the transmitter. As with an optical amplitude modulator, the optical output signal will contain two mirrors of the signal spectrum, one on each side of the laser carrier, and one of these should preferably be selected by an optical filter before transmission. This can normally be accommodated by the multiplexer filter of the dense wavelength division multiplexing system (DWDM). Even though in principle all energy from the laser carrier can be transferred to the side-bands, this is not possible without significant energy also arising in higher harmonic spectral components. In reality the carrier cannot be substantially depleted in a multi-channel system, since harmonic components will then induce cross-talk between the optical channels [10].
2. Theory
The generation of higher order spectral components when using a phase modulator can be illustrated by considering a sinusoidal microwave signal, A sin(ωCt), at a sub-carrier angular frequency, ωC. After phase modulation by such a signal, the resulting optical field is
where EO is the input optical field, ωO the angular frequency of the optical carrier, m is the modulation index defined as m = πA/Vπ, and Vπ is the drive voltage of the phase modulator for π phase shift. The optical signal in Eq. (1) can be expanded aswhere Jk is the Bessel function of order k. From Eq. (2) it can be concluded that if the amplitude, A, of the microwave signal is small relative to Vπ, the signal appears relatively undisturbed in the first order sideband. For small m, we can express the ratio between the wanted first order sideband and the unwanted second order sideband as3. Experimental setup
In order to test the concept of electro-optical conversion of multiple RF signals to the optical domain using a phase modulator, a dual channel RF transmitter generating 7 Gbaud 16-QAM (28 Gbit/s) per channel was used. Figure 1 shows the experimental setup of the system under test. Data at 7 Gbaud was encoded on two RF sub-carriers at 14 GHz and 24 GHz using commercial electronic I/Q-upconverters designed for use in e.g. high-speed microwave radio links. Here, binary drive signals (210-1 PRBS) were injected to the two inputs of the I/Q upconverters resulting in QPSK modulation at the output of the electrical modulators. 16-QAM modulation was emulated by splitting the electrical QPSK signals, decorrelating and attenuating one branch by 6 dB, before combining them to obtain a 16-QAM signal. The two RF carriers were subsequently combined before being injected to the optical phase modulator modulating light from a tunable external cavity laser (TECL). The phase modulator used was a LiNbO3 based phase modulator with 27.5 GHz electro-optical 3 dB-bandwidth. A 112 Gbit/s dual polarization signal was obtained by emulating a polarization multiplexed system by splitting the output light from the modulator and decorrelating the two paths before combining them in a polarization beam combiner (PBC). The lower mirror side-band was then filtered out by properly aligning the transmit laser wavelength to the 50 GHz spacing optical add-drop multiplexer (OADM). The resulting spectrum at the input of the transmission link is shown in Fig. 2 .
Fig. 1 Schematic of the 100G RF assisted test system with optical phase modulator. Two RF carriers are modulated with 7 Gbaud 16-QAM using electrical IQ-modulators with subsequent electro-optical conversion using a phase modulator. 112 Gbit/s data is obtained by using a polarization multiplexer emulator.
Fig. 2 Optical spectrum from the transmitter. Green trace: single channel only; Pink trace: Dual channels high modulation depth; Yellow trace: Dual channels low modulation depth.
At the receiver, conventional polarization-diverse intradyne coherent detection was used involving another TECL as local oscillator (LO), a polarization beam splitter (PBS), optical hybrids and balanced photodetectors. The two sub-carriers were detected separately by tuning the LO to approximately the center wavelength of each channel.
The outputs of the four balanced detectors were digitized by asynchronous sampling in a real-time oscilloscope and subsequently processed in a computer. Data was recovered using a number of DSP blocks (see [6] for more details) including sampling skew and I/Q-phase offset correction, frequency domain compensation of chromatic dispersion, resampling and retiming to synchronous two samples per symbol, adaptive equalization and phase recovery. A number of symbols needed for convergence of all filters are discarded before applying Gray mapping to bits and counting bit errors. About 200k symbols in each polarization are used for the BER computation.
4. Results
Figure 2 shows the optical spectra after the phase modulator and subsequent OADM filter for single sub-channel (56 Gbit/s) and dual sub-channels for two different drive amplitudes. The two sub-channels are centered in the pass-band of the OADM filter and the mirrors on the left side of the laser carrier are completely suppressed since the distance between the two closest mirrors are 28 GHz. However, the OADM filter cannot completely suppress the laser carrier that is only 19 GHz from the center of the passband. Thus a substantial fraction of the total optical energy resides in the carrier. For higher modulation depth the laser carrier is more depleted, but then higher harmonics of the sub-channels are present in the spectrum. At lower modulation depth, harmonic distortion is less of a problem, but more energy remains in the laser carrier. The potential of crosstalk between the channels are clearly seen by the higher harmonic generation of the single channel, as shown by the green trace in Fig. 2. A reasonable tradeoff between carrier depletion and harmonic crosstalk is shown by the yellow trace in Fig. 2.
As a consequence of the non-depleted carrier it is difficult to obtain a relevant measure of the optical signal-to-noise ratio (OSNR) since the energy in the carrier may dominate the measured signal energy. Here, we measure BER after a transmission link and compare with a conventional MZM based transmitter. The transmission link comprised 12 spans of installed G.652 fiber with an average amplifier span length of about 70 km for a total distance of 824 km. No inline dispersion compensation was used.
Figure 3b shows measured BER versus launch power into each span for the transmitter using a phase modulator (red trace). For comparison, the phase modulator was replaced with a zero chirp MZM biased for carrier suppression with the resulting BER versus launch power shown as the blue trace in Fig. 3b. We notice that the phase modulator trace is shifted to about 5 dB higher launch power, which indicates that about 70% of the total power resides in the laser carrier. Also the optimum BER is slightly higher with the phase modulator, which reveals some cross-talk penalty. The 24 GHz sub-channel always suffered from higher BER value, which is most likely due to the second order harmonic of the 14 GHz channel, giving rise to some cross talk on the 24 GHz sub-channel. Figure 3a shows constellations before and after transmission for the 24 GHz sub-channel.
Fig. 3 Experimental results: (a) 16-QAM constellations before and after the 824 km transmission link and (b) measured BER versus span launch power after 824 km with phase modulator (red) and conventional MZM (blue).
Another interesting observation made during the experiments was that the ratio between the remaining laser carrier and the data carrying RF channels at the transmitter was reduced by about 1.6 dB after 824 km transmission. This reduction is believed to occur due to stimulated Brillouin scattering (SBS) in the transmission fiber. A fraction of the narrow band laser carrier is scattered in the backward direction due to SBS and thus depleting the forward propagating light.
In order to further investigate penalty from nonlinear crosstalk resulting from the phase modulator and fiber transmission, numerical simulations were performed using VPItransmissionMaker. Figure 4a shows required OSNR for BER = 10−3 as a function of root-mean-square (RMS) amplitude of the electrical drive signal for both a phase modulator and a MZM. Note that only the useful signal power in the data sub-channels was taken into account when measuring OSNR, excluding the relatively large fraction of power residing in the carrier, which would otherwise cause a large OSNR penalty in the case of the phase modulator. Even so, the penalty increases sharply with increased drive amplitude for the phase modulator, mainly due to the second-order nonlinearity not present with the MZM.
Fig. 4 Simulation results: (a) Required OSNR for BER = 10−3 for phase modulator (blue) and conventional MZM (red) versus RMS amplitude of the driving signal, (b) BER versus launch power for MZM and phase modulator with different levels of carrier suppression in the OADM filter preceding the link.
Figure 4b shows simulated BER versus launch power for the same transmission link configuration that was used in the experiment. For the phase modulator, the drive amplitude was 0.03 Vπ and different levels of carrier suppression were tested by varying the alignment of the signal to the OADM filter preceding the link. When the carrier is strongly suppressed, roughly the same performance as for the MZM is obtained with a small penalty caused by the phase modulator nonlinearity. As carrier suppression is decreased we observe the same behavior as in the experiment; the optimum power is shifted to higher levels due to the residual power in the carrier and the optimum BER increases slightly which we believe is caused by nonlinear effects in the fiber supported by the presence of the carrier.
5. Conclusions
The concept of an RF-assisted optical 16-QAM transmitter using an optical phase modulator was demonstrated at 112 Gbit/s. The system was tested over an 824 km installed fiber link with small penalty compared to a conventional MZM in the transmitter. The main problem was the residual energy in the laser carrier that co-propagates through the link. However, this can be mitigated either by using a steeper optical filter, or by increasing the RF frequencies of the microwave sub-channels, which would potentially position the laser carrier outside the filter pass-band. Our results show that the demonstrated concept is a promising option a for cost-efficient transmitter implementation.
References and links
1. K. Roberts, M. O'Sullivan, K.-T. Wu, H. Sun, A. Awadalla, D. J. Krause, and C. Laperle, “Performance of Dual-Polarization QPSK for Optical Transport Systems,” J. Lightwave Technol. 27(16), 3546–3559 (2009). [CrossRef]
2. OIF, “100G Ultra Long Haul DWDM Framework Document”, June 2009, http://www.oiforum.com/public/documents/OIF-FD-100G-DWDM-01.0.pdf
3. P. J. Winzer, A. H. Gnauck, C. R. Doerr, M. Magarini, and L. L. Buhl, “Spectrally Efficient Long-Haul Optical Networking Using 112-Gb/s Polarization-Multiplexed 16-QAM,” J. Lightwave Technol. 28(4), 547–556 (2010). [CrossRef]
4. M. S. Alfiad, M. Kuschnerov, S. L. Jansen, T. Wuth, D. van den Borne, and H. de Waardt, “11 x 224-Gb/s POLMUX-RZ-16QAM Transmission Over 670 km of SSMF With 50-GHz Channel Spacing,” IEEE Photon. Technol. Lett. 22(15), 1150–1152 (2010). [CrossRef]
5. M. Nölle, J. Hilt, L. Molle, M. Seimetz, and R. Freund, “8×224 Gbit/s PDM 16QAM WDM Transmission with Real-Time Signal Processing at the Transmitter,” in Proc. ECOC 2010, Paper We.8.C.4.
6. B. E. Olsson, J. Mårtensson, A. Kristiansson, and A. Alping, “RF-Assisted Optical Dual-Carrier 112 Gbit/s Polarization- Multiplexed 16-QAM Transmitter,” in Proc. OFC/NFOEC 2010, Paper OMK5.
7. B. E. Olsson, J. Mårtensson, A. Alping, A. Djupsjöbacka, and A. Rhodin, “112-Gb/s RF-Assisted Multicarrier DP-16-QAM Optical Transmission Over Field Deployed Fiber Link,” IEEE Photon. Technol. Lett. 23(19), 1367–1369 (2011). [CrossRef]
8. C. Rolland, “InGaAsP-based Mach-Zehnder modulators for high-speed transmission systems”, in Proc. OFC 1998, Paper ThH1.
9. A. Alping, X. S. Wu, T. R. Hausken, and L. A. Coldren, “Highly efficient waveguide phase modulator for integrated optoelectronics,” Appl. Phys. Lett. 48(19), 1243–1245 (1986). [CrossRef]
10. P. O. Hedekvist, B. E. Olsson, and A. Wiberg, “Microwave harmonic frequency generation utilizing the properties of an optical phase Modulator,” J. Lightwave Technol. 22(3), 882–886 (2004). [CrossRef]
