We propose a new optical transmitter which is capable of changing flexibly the modulation format of the optical signal. By using this transmitter, we can handle and assign various modulation formats: binary phase shift keying (BPSK), quadrature phase shift keying (QPSK), 8-ary quadrature amplitude modulation (8QAM), and 16QAM. The proposed transmitter is based on a combination of a dual-drive Mach-Zehnder modulator (DD-MZM) and a dual-parallel MZM (DP-MZM) with electrical binary drive signals. DD-MZM is a key element to produce the 8QAM and 16QAM formats where each arm of DD-MZM is driven by independent binary data. This is because we can modulate the amplitude and phase of the optical signal by using a frequency chirp of the modulator when we adjust properly the amplitudes of the electrical drive signals. In addition, we show an algorithm by which the proposed transmitter can intelligently select the modulation format in accordance with the signal quality.
© 2012 OSA
In order to efficiently utilize the limited spectrum resources in a dynamically reconfigurable wavelength-division-multiplexing (WDM) network, there have been numerous strategies regarding the adaptive spectrum allocation [1–4]. In other words, the assigned bandwidth of each channel (i.e., lightpath) will not be fixed any longer. Therefore, to maximize the spectral efficiency of arbitrary bandwidth channels, the optical transmitter is highly desired to provide the flexibility (i.e., the modulation format and/or bit rate of the optical signal can be flexibly changed without modifying the transmitter hardware). For this purpose, several technologies for realizing the flexible optical transmitter have been reported [5–7]. For example, a new transmitter based on the silica planer lightwave circuits and LiNbO3 (PLC-LN) hybrid integration technology was proposed and demonstrated . By using the highly integrated modulators in parallel, this transmitter can generate the multi-flow and multi-rate optical signals. However, for its proper and stable operation, it is necessary to control precisely many parameters such as tunable filters, variable couplers, phase shifters, and modulator biases. Besides, the software-defined multi-format transmitters have been reported [6,7]. To flexibly switch the modulation format of the optical signal, the electrical drive signals applied to an IQ modulator are changed from binary to multilevel signals through the software. For this operation, the drive signals are generated by a digital-to-analog converter (DAC) controlled by the field-programmable gate array (FPGA) chip  or an electrical-optical-electrical (E-O-E) method . However, the multilevel generation methods either require high computational power or are complicated in hardware. In addition, these techniques demand the superior quality of the driver amplifiers for the electrical multilevel signals.
In this paper, we propose a new flexible-format optical transmitter using the electrical binary drive signals. This transmitter produces binary phase shift keying (BPSK), quadrature phase shift keying (QPSK), 8-ary quadrature amplitude modulation (8QAM), and 16QAM. The proposed transmitter consists of a dual-drive Mach-Zehnder modulator (DD-MZM) and a dual-parallel MZM (DP-MZM) in tandem. Each modulator is driven by the binary data. For BPSK or QPSK generation, DD-MZM is turned off (i.e., it is transparent) and DP-MZM is driven by a single binary data or two independent binary data, respectively. 8QAM and 16QAM formats are generated by adjusting a bias voltage and amplitudes of the electrical signals applied to DD-MZM while DP-MZM produces a QPSK output. As a result, the proposed transmitter can create various formats without the hardware modification. In addition, due to the use of the binary drive signals, this transmitter has a potential to easily increase the symbol rate and it does not require either the sophisticated electronics or pre-equalization to cope with the modulator’s nonlinear characteristic. We also present how to intelligently assign the modulation format and change flexibly it. By using this approach, we experimentally demonstrate that the proposed flexible transmitter can switch automatically the modulation format in accordance with the bit-error rate (BER) of the optical signal.
2. Proposed transmitter
Figure 1 shows the configuration of the proposed transmitter which is capable of generating BPSK, QPSK, 8QAM and 16QAM. This transmitter consists of DD-MZM and DP-MZM and every drive signal (i.e., Data1 ~Data4) is an electrical binary data. First, each arm of DD-MZM is driven by two independent binary data, Data1 ( = A1v1(t)) and Data2 ( = A2v2(t)) where vi(t) (i = 1, 2, 3, and 4) represents a binary data taken values between 0.5 and −0.5. If we adjust the amplitudes of these drive signals (i.e., A1 and A2), we are able to manipulate arbitrarily the amplitude and phase of the optical signal by using a frequency chirp of the modulator. When the splitting/combining ratios of DD-MZM are even, the transfer function H(t) of DD-MZM is expressed as [8,9]
For BPSK generation, DD-MZM is switched off (i.e., it is transparent) and MZMB of DP-MZM is driven by the electrical binary signal with the driving amplitude of 2Vπ while a drive signal (Data3) of MZMA is null, as shown in Table 1 . In this table, the bias positions of DP-MZM are represented by peak/null/quadrature (maximum/minimum/half-position of the typical transfer curve of MZM) except vb.
For QPSK generation, DD-MZM is also turned off and DP-MZM operates typically to make the QPSK signal, as shown in Table 1. In other words, MZMA and MZMB produce BPSK signals and two BPSK signals are then combined orthogonally in MZMC.
The optimum constellation of 8QAM has two moduli with the amplitude ratio of , as illustrated in Fig. 2(c) . For the optimum 8QAM generation, the amplitude and phase of the optical signal are first modulated by controlling the bias voltage and amplitude of the drive signal of DD-MZM as shown in Fig. 2(b). For this operation, we select the modulation conditions described in Table 1 although there are a couple of combinations of a bias and drive signals . This condition enables to realize easily the flexible functionality of the proposed transmitter. Figure 2(a) shows the modulation procedure in DD-MZM where blue empty and red filled circles refer to the phase modulation at the upper and lower arms of DD-MZM, respectively. From Eq. (1), Data1 experiences the phase shift by the bias adjustment and two signals of upper and lower arms are then added (denoted by black circles). As a result, the desired constellation of Fig. 2(b) can be obtained. In addition, DP-MZM produces a typical QPSK output and it results in 8QAM as shown in Fig. 2(c).
Table 1 and Fig. 3 also introduce the modulation condition for 16QAM generation. In the same manner as 8QAM, DD-MZM is used to make the desired four symbols of 16QAM by applying two independent data to each arm of DD-MZM . When A1 = A2 = 0.3Vπ and vb = 0.5Vπ, we can make four symbols placed in a quadrant of the 16QAM constellation as shown in Fig. 3(a). Blue empty and red filled circles represent the phase modulation at the upper and lower arms of DD-MZM, respectively, and each individual signal is added to each other (denoted by black circles). Thus, DD-MZM produces the desired four symbols like Fig. 3(b). Then, through QPSK modulation by DP-MZM, 16QAM is finally generated as illustrated in Fig. 3(c).
3. Experimental demonstration
We introduce an algorithm in order to assign the modulation format of the proposed transmitter. We believe it is a very important functionality in future dynamic transparent optical networks since, for example, when an optical path is rerouted in transparent optical networks, the restoration path may have different physical characteristics such as transmission distance, optical signal-to-noise ratio (OSNR), and number of the optical filters. In order to either guarantee the signal quality at the receiver side or maximize the spectral efficiency, it is necessary to automatically assign a proper modulation format and/or a transmission rate of the optical signal in accordance with the link’s condition. For this purpose, first, we continuously check the signal’s quality by monitoring BER of the optical signal. When BER is degraded below the pre-defined BER threshold, we select a lower-order QAM format than the current modulation format to improve the OSNR margin. For example, if BER of the 16QAM signal becomes worse than the upper limit of the BER threshold in the restoration path, 8QAM or QPSK is then chosen. On the other hand, if BER becomes better than the lower limit of the BER threshold, a higher-order QAM format is selected to provide the high spectral efficiency. Figure 4 shows the flow chart of this algorithm regarding how to assign the modulation format of the optical signal. At the initial stage, we set the modulation format to be 16QAM for the high spectral efficiency and define the upper and lower limits of the BER threshold. In addition, here, we assume that the fiber loss of the signal path is varied.
By using this procedure, we experimentally demonstrated the flexible optical transmitter together with a digital coherent receiver, as illustrated in Fig. 5 . To remove the additional effects caused by the modulator chirp, we applied return-to-zero (RZ) pulse shaping . In addition, to emulate the dynamic change of the fiber link, we varied the insertion loss between transmitter and receiver by using an optical attenuator. The self-homodyne coherent receiver was implemented by using a 90° optical hybrid and balanced photodetectors (BPDs) followed by a 40-GS/s digital storage oscilloscope with 12-GHz bandwidth as an analog-to-digital (A/D) converter. The captured data were processed offline . The signal processing algorithm contained the correction of the receiver imperfections, half-symbol-spaced linear equalization adapted by constant-modulus algorithm (CMA) and radius-directed algorithm (RDA), carrier recovery, decision-directed least-mean-square (LMS) equalization, and BER calculation. In addition, the identical signal processing was used for every modulation format (i.e, same tap length and step-size parameter for equalizers). In this experiment, we utilized a clock of 12.5 GHz (i.e., the symbol rate of the modulated signals was 12.5 Gbaud) and it was fixed. A laser was modulated with 12.5-Gbaud RZ-QPSK, RZ-8QAM, or RZ-16QAM and every drive signal was a pseudo-random bit sequence (PRBS) with a sequence length of 215-1. The transmitter, optical attenuator, and receiver were connected to a computer by using a general purpose interface bus (GPIB). In the transmitter, to change the modulation format, the gain of the driver amplifiers was controlled by adjusting a dc bias voltage. The dc voltage was provided by a typical dc power supply connected by GPIB. When the modulation format was changed from QPSK to 8QAM or 16QAM, additional losses occurred because of different bias position and amplitudes of the drive signals applied to DD-MZM. In our experiment, the additional losses for 8QAM and 16QAM generation were 2.28 dB and 3.06 dB, respectively. First, we measured the BER performances of 12.5-Gbaud RZ-QPSK, RZ-8QAM, and RZ-16QAM signals, as shown in Fig. 6 . The OSNR sensitivities to achieve BER of 10−3 were 6.4 dB, 11.8 dB, and 15.2 dB, respectively, which were about 1~2 dB off the theoretical limit . These implementation penalties were caused by the transmitter imperfections such as an imbalance between the amplitudes of drive signals and error in bias voltage.
We also demonstrated experimentally the feasibility and effectiveness of the proposed flexible-format transmitter. As described above, to decide the signal’s format, the upper and lower BER thresholds were set to be 2x10−3 and 2x10−4, respectively. Thus, if the measured BER was out of this range, the modulation format of the optical signal was switched. For this evaluation, we varied the insertion loss in every 5 steps and the loss increased or decreased. Figure 7 shows the experimental results. When we increased the insertion loss from 11 dB to 13 dB, the measured BER of the 16QAM signal was degraded correspondingly. However, at the loss of 15dB, BER of the 16QAM signal was out of the BER threshold and resulted in switching the modulation format (from 16QAM to 8QAM) automatically. Consequently, we could observe that BER was back within the threshold. In the same manner, when the insertion loss was changed to 21 dB, the 8QAM signal’s BER was worse than the upper BER threshold and, as a result, the modulation format was changed to QPSK. However, when the insertion loss was reduced to 11 dB again, the modulation format was switched to 8QAM and finally switched to 16QAM because the measured BER was better than the lower threshold.
A new flexible optical transmitter was proposed and successfully demonstrated. This transmitter could generate BPSK, QPSK, 8QAM and 16QAM formats without any modification of hardware. The proposed transmitter was implemented by combining DD-MZM and DP-MZM in tandem and each modulator was driven by the electrical binary signal. Thus, it has a potential to increase the symbol rate of the multilevel modulation format. In addition, we experimentally evaluated the feasibility and effectiveness of the proposed BER-adaptive flexible-format transmitter.
This work was partly supported by the Ministry of Internal Affairs and Communications, Japan.
References and links
1. M. Jinno, H. Takara, B. Kozicki, Y. Tsukishima, Y. Sone, and S. Matsuoka, “Spectrum-efficient and scalable elastic optical path network: architecture, benefits, and enabling technologies,” IEEE Commun. Mag. 47(11), 66–73 (2009). [CrossRef]
2. B. Kozicki, H. Takara, Y. Sone, A. Watanabe, and M. Jinno, “Distance-adaptive spectrum allocation in elastic optical path network (SLICE) with bit per symbol adjustment,” in Proceedings of OFC2010, paper OMU3 (2010).
3. N. Amaya, M. Irfan, G. Zervas, K. Banias, M. Garrich, I. Henning, D. Simeonidou, Y. R. Zhou, A. Lord, K. Smith, V. J. F. Rancano, S. Liu, P. Petropoulos, and D. J. Richardson, “Gridless optical networking field trial: flexible spectrum switching, degragmentation and transport of 10G/40G/100G/555G over 620-km field fiber,” in Proceedings of ECOC2011, paper Th.13.K.1 (2011).
4. M. Eiselt, “Flexible optical transport solutions,” in Proceedings of ECOC2010, workshop WS2 (2010).
5. H. Takara, T. Goh, K. Shibahara, K. Yonenaga, S. Kawai, and M. Jinno, “Experimental demonstration of 400 Gb/s multi-flow, multi-rate, multi-reach optical transmitter for efficient elastic spectrum routing,” in Proceedings of ECOC2011, paper Th.5.A.4 (2011).
6. R. Schmogrow, D. Hillerkuss, M. Dreschmann, M. Huebner, M. Winter, J. Meyer, B. Nebendahl, C. Koos, J. Becker, W. Freude, and J. Leuthould, “Real-time software-defined multiformat transmitter generating 64QAM at 28 GBd,” IEEE Photon. Technol. Lett. 22(21), 1601–1603 (2010). [CrossRef]
7. Y.-K. Huang, E. Ip, P. N. Ji, Y. Shao, T. Wang, Y. Aono, Y. Yano, and T. Tajima, “Terabit/s optical superchannel with flexble modulation format for dynamic distance/route transmission,” in Proceedings of OFC2012, paper OM3H.4 (2012).
8. H. Y. Choi, T. Tsuritani, and I. Morita, “Method to generate 112-Gb/s polarization-multiplexed 8QAM signal,” Electron. Lett. 48(9), 511–512 (2012). [CrossRef]
9. H. Y. Choi, T. Tsuritani, and I. Morita, “A novel transmitter for 320-Gb/s PDM-RZ-16QAM generation using electrical binary drive signals,” in Proceedings of ECOC2012, paper Tu.4.A.2 (2012).
10. H. Y. Choi, T. Tsuritani, and I. Morita, “Effects of LN modulator chirp on performance of digital coherent optical transmission system,” in Proceedings of COIN2012, paper TuF.2 (2012).
11. J. G. Proakis, Digital Communications, 4th ed. (McGraw-Hill, New York, 2001).