Open Access
Add to BibSonomyAbstract
In this paper, a FPGAs-based real-time adaptively modulated 256/64/16QAM-encoded base-band OFDM transceiver with a high spectral efficiency up to 5.76bit/s/Hz is successfully developed, and experimentally demonstrated in a simple intensity-modulated direct-detection optical communication system. Experimental results show that it is feasible to transmit a raw signal bit rate of 7.19Gbps adaptively modulated real-time optical OFDM signal over 20km and 50km single mode fibers (SMFs). The performance comparison between real-time and off-line digital signal processing is performed, and the results show that there is a negligible power penalty. In addition, to obtain the best transmission performance, direct-current (DC) bias voltage for MZM and launch power into optical fiber links are explored in the real-time optical OFDM systems.
© 2014 Optical Society of America
1. Introduction
With the rapid growth of high application services like IPTV and HDTV in current network, a further increase of bandwidth in the optical access networks (OANs) has to be considered [1–4]. However, the bandwidth requirement increased the cost of transmitter and receiver. It is well known that the highly cost-effective passive optical network (PON) architecture is a key solution for low cost optical access networks. At present, there are two key technologies used in PONs such as time division multiplexing (TDM)-based PON and wavelength division multiplexing (WDM)-based PON. Currently, TDM-PON architecture needs complex scheduling algorithms and framing technology to support a variety of services. On the other hand, WDM-PON assigns different high-speed data to the appoint wavelength via arrayed waveguide grating or optical filter distribute wavelengths to the corresponding receivers, which will increase both system cost and complexity [2,3]. Recently, optical orthogonal frequency division multiplexing (OFDM) systems have attracted a lot of attention in both academic and industrial sectors, and it is regarded as one of the most promising techniques for the next-generation OANs [3] due to its some advantages such as high spectral efficiency (SE) of M-QAM modulation format, allowing the use of conventional low bandwidth devices, powerful digital signal processing (DSP) and high tolerance to both chromatic dispersion and polarization-mode dispersion [1–4]. Furthermore, compared to regular OFDM, the use of adaptively modulated optical OFDM (AMOOFDM) can further improve signal transmission capacity, network flexibility and performance robustness in a cost-effective manner [1]. In adaptive modulation scheme, modulation formats on the data-carrying sub-carriers can be adjusted according to the transmission characteristics of each subcarrier. The spectral efficiency is improved by assigning a high modulation format to “high quality” sub-carriers while low modulation formats are assigned to “low quality” sub-carriers [4].
Among optical OFDM systems, the intensity-modulation and direct-detection (IMDD) optical OFDM transceivers is very attractive for its compactness and cost-effectiveness. However, most of the experimental works reported so far [4–8] have been undertaken using off-line digital signal processing approaches, which do not consider the limitations imposed by the precision and speed of practical DSP hardware required for real-time optical OFDM transceiver. In [8], the performance of 11.85Gbps 256QAM-encoded IMDD optical OFDM communication system is investigated by off-line DSPs. The sample rates of DAC in arbitrary waveform generator (AWG) and ADC in the real-time oscilloscope are set to 4GSps and 20GSps, respectively. In this way, ADC over-sampling can improve ADC resolution and reduce ADC quantization noise. Recently, there are some real-time FPGA-based OFDM transmitters or/and receivers have been demonstrated in optical transmission systems [9–13]. To our best knowledge, 128QAM-encoded optical OFDM with a SE of 5.25bit/s/Hz is the highest modulation format reported in the intensity modulation and direct detection optical OFDM (DDO-OFDM) real-time system [11].
In this paper, a FPGAs-based real-time adaptively modulated 256/64/16AQM encoded base-band OFDM transceiver with a high SE up to 5.76bit/s/Hz is developed and experimentally investigated in the IMDD optical communication system. In the transceiver, 2.5GSps digital-to-analog converter (DAC) and analog-to-digital converter (ADC), and an external Mach-Zehnder modulator (MZM) are used. Experimental results show that it is feasible to transmit a raw signal bit rate of 7.19Gbps adaptively modulated real-time optical OFDM signal over 20km and 50km SMFs. Meanwhile, the performance comparison between real-time and off-line DSP is performed, and the results show that there is a negligible power penalty. In addition, in order to obtain the best transmission performance, the optimal direct-current (DC) bias voltage for MZM and launch power into optical fiber links are explored by real-time monitoring measured bit error rate (BER) performance at the receiver.
2. Real-time base-band OFDM transceiver
The field programmable gate arrays (FPGAs)-based real-time base-band transceiver is shown in Fig. 1, most of core DSP functions can be found in our previous work [12,13]. It is should be noted that, in this paper, two Xilinx FPGA boards, namely ML605 equipped with a Virtex-6 XC6VLX240T FPGA and VC707 equipped with a Virtex-7 XC7VX485T FPGA are used to implement digital signal processing (DSP) algorithms in transmitter and receiver, respectively. For the real-time transmitter, the DSP functions consist of training sequence (TS) generation, adaptive modulation, pilot insertion, 128-point inverse fast Fourier transform (IFFT) with input complex data being arranged to satisfy the Hermitian symmetry, the real outputs after IFFT process are clipped at a optimal digital clipping ratio of 11.52dB and then scaled to obtain the negligible clipping noise and quantization noise, and cyclic prefix (CP) addition. On the other hand, for the real-time receiver, the corresponding DSP functions are include simple TS-based symbol synchronization, CP removal, 128-point FFT, TS-based channel estimation and only phase compensation, pilot-assisted sampling clock frequency offset (SFO) estimation and SFO-caused phase compensation, sampling frequency synchronization based on the estimated SFO, adaptive demodulation and real-time BER analysis.
The 2.5GSps DAC and ADC are used in the real-time transceiver. However, the state-of-the-art FPGA can only operate at about a few hundred megahertz. Instead of serial data processing, the parallel signal processing technique is used to reduce operation frequency. In our real-time transceiver, the clock frequencies for both transmitter and receiver are 156.25MHz by utilizing 16-channel parallel digital signal processing. Some key parameters for the 7.19Gbps real-time base-band OFDM transceiver are given in Table 1.
Table 1. Some Key Parameters for the Real-Time Base-band OFDM Transceiver
To enhance the performance of the real-time optical OFDM communication system, adaptive modulation scheme is introduced into the real-time base-band transceiver. In adaptive modulation scheme, modulation formats taken on the data-carrying sub-carriers are adjusted according to the transmission characteristics of each subcarrier. In generally, to identify the highest level modulation format that should be on each data-carrying sub-carrier, the M-QAM mapping is chosen so that high (low) modulation format is used on a sub-carrier having a high (low) signal-to-noise ratio (SNR). In our real-time optical OFDM system, the signal modulation formats are 256-QAM, 64-QAM and 16-QAM. Due to the sinc roll-off effect of the DAC [14] and imperfect frequency response of low-pass filter, there is a high power fading at the high-frequency sub-carriers. It will result in much worse SNRs loss after optical fiber transmissions. The normalized amplitude responses on data-carrying and the pilot sub-carriers obtained by the TS-based channel estimation technique are shown in Fig. 2.It can be seen clearly that the power attenuation at the highest-frequency sub-carrier is about 10dB. While the low-frequency sub-carriers are mainly suffered from the signal-to-signal beating interference (SSBI) after square-law detection via the photodiode in the receiver side [7, 15]. So the low modulation formats 64-QAM and 16-QAM is modulated on these low-SNR sub-carriers. Other sub-carriers with high SNRs are modulated with 256-QAM. In this way, the SE and transmission performance of the system can be improved efficiently. The adaptive modulation scheme for the real-time system is also shown in Fig. 2.
3. Experimental setup
The experimental setup of the proposed real-time IMDD optical OFDM communication system is shown in Fig. 1. The electrical base-band adaptively modulated OFDM signal with a peak-to-peak voltage of 700mV is first generated by a 2.5GSps DAC. Then it is used to drive the MZM at a DC bias voltage of 2.6V which modulate an optical carrier generated by an external cavity laser (ECL) with 100 kHz line-width. The polarization controller (PC) is rotated to maximize the MZM output optical power at about 3.1dBm. The generated optical OFDM signal is boosted up to 6.0dBm by an erbium-doped fiber amplifier (EDFA-1), and then transmission over 20km and 50km SMFs respectively. At the receiver, another EDFA (EDFA-2) is used to boost the optical power of the OFDM signal up to 6.1dBm. Before the photoelectric conversion, the power of the detected optical OFDM signal can be changed by a variable optical attenuator (VOA). The received base-band OFDM signal can be obtained via a direct-current (DC) coupled photo-detector (PD) with a bandwidth of 10GHz and a maximal optical input power of 4dBm. An anti-aliasing filter with a 3dB bandwidth of 1.1GHz is used to avoid aliasing at the ADC. The power spectrums of the received OFDM signal before and after the electrical low-pass filter (EPLF) are measured by a digital storage oscilloscope (DSO) (Tektronix TDS6804B) at a sample rate of 5GSaps. And it is shown in Fig. 1(a) and 1(b), respectively. It can be clearly that the image of DAC output signal is completely suppressed by the EPLF. However, there is a peak on the power spectrums at the frequency of 1.25GHz comes from the sampling clocks used for the DAC and the DSO. Then a DC block is utilized to remove the PD-output DC component. The samples captured by a 2.5GSps ADC are sent to FPGA for real-time DSP. Some of internal signals of FPGA at the receiver including the ADC captured samples for off-line processing, error bits count and estimated SFO, are sampled by the Xilinx ChipScope module and uploaded to a computer via a Joint Test Action Group (JTAG) cable for a further analysis. It is should be mentioned that the DSO is only used to measure the power spectrums of received signal, while all the other off-line results are obtained by processing the samples that captured by the 2.5GSps ADC using a Matlab program.
4. Experimental results and discussion
4.1 DC bias voltage for MZM optimization
The transfer curve of the MZM is experimentally measured, as shown in Fig. 3.Meanwhile, at optical back-to-back (OBTB) case, the real-time measured BER performance versus applied bias voltage for MZM is also shown in Fig. 3. The results show that, the half-wave voltage of the MZM is about 3.5V, and the optimal DC bias voltage for MZM is 2.6V. When a low or high DC bias voltage is applied to MZM, it will make the BER performance of the system even worse due to the nonlinear transfer function of MZM.
Fig. 3 Transfer curve of the MZM and real-time measured BER performance versus applied bias voltage for MZM
Figure 4 shows that the EVM values on all of the data-carrying sub-carriers which are calculated by processing the ADC captured samples off-line. The severely degraded EVM performance on low-frequency sub-carriers from electrical back-to-back (EBTB) to OBTB is mainly attributed to the heavy SSBI, while the high-frequency sub-carriers suffering a high power fading for both EBTB and OBTB cases. It is due to the imperfect responses of the DAC and LPF.
4.2 Launch power optimization
The optimal launch power for the real-time system is experimental investigated by on-line monitoring the BER values over different optical launch power. After 50km SMF transmission, the real-time measured BER performance with different launched power is shown in Fig. 5.It can be seen that the optimal launch power is about 6dBm for the real-time system. For degraded BER performance at low launch power, the reason is mainly due to the noise figure of the EDFA-2, while the deteriorated BER performance at high launch power is mainly caused by optical fiber nonlinear effects.
4.3 Transmission performance over SMF Links
To make use of limited bandwidth in access networks, the 7.19Gbps FPGAs-based adaptively modulated 256/64/16QAM-encoded real-time base-band OFDM transceiver is developed. Then, the transmission performance over 20km and 50km SMFs is experimentally investigated in IMDD optical communication systems. The optimal bias voltage for MZM and optimal launch power are set to 2.6V and 6.0dBm, respectively. The BER value for EBTB, at a received optical power of 4.0dBm for OBTB, and after 20km and 50km SMFs transmissions, are summarized in Table 2.
Table 2. 7.19Gbps Real-Time DDO-OFDM Systems Performance
Note that, for channel equalization, only phase compensation is done in our real-time receiver to avoid the complex division operations [12]. The 256/64/16QAM constellations of the adaptively modulated OFDM signal are obtained by using off-line DSP approaches, are shown in Figs. 6(a)-6(i). From EBTB to OBTB case, the degrade EVM performances are mainly due to the abovementioned non-ideal optical modulation and detection, while the noise figure of EDFA-2 is the major factor for the EVMs degradation from OBTB to post-20km SMF transmission.
Fig. 6 256/64/16QAM constellations for (a-c) electrical back-to-back, (d-f) optical back-to-back, and (g-i) after 20km SMF transmission
In order to obtain the best BER performance, the sampling clock frequency synchronization should be achieved by adjusting the clock frequency of receiver according to the estimated SFO in the initial stage of establishing a connection between transmitter and receiver. The real-time measured BER performances of the 7.19Gbps adaptively modulated OFDM signal over OBTB, 20km and 50km SMF transmissions and with a total of 95,420,416 transmitted bits versus received optical power is shown in Fig. 7.The BER performance degrades with the decrease of received optical power and the increase of fiber span, but the BER of post-20km SMF is still less than 7% hard-decision forward-error-correction (HD-FEC) threshold of 3.8 × 10−3. In addition, the BER of the system after 50km SMF transmission is far less than 20% soft-decision FEC (SD-FEC) threshold of 2.7 × 10−2 [8] and near HD-FEC threshold when the received optical power is 4dBm. In order to avoid electrical amplifiers (EA) nonlinearity, we do not use EA after DAC and before ADC. However, it reduces the receiver sensitivity. Compared with the OBTB case, at a BER of 3.8 × 10−3, there are 1.6dB and about 3dB power penalties after 20km and 50km SMF transmissions, respectively. For comparison, the BER performance from off-line DSPs is also given in Fig. 7. It is obvious that there is a negligible power penalty between real-time and off-line digital signal processing results.
5. Conclusions
In this paper, the aggregated 7.19Gbps real-time adaptively modulated 256/64/16QAM-encoded OFDM transceiver with a high SE of 5.76bit/s/Hz has been successfully developed based on off-the-shelf FPGAs. It has also been experimentally investigated in IMDD optical communication system. The experimental results showed that the BER of adaptively modulated OFDM signal with a net signal bit rate of 6.30Gbps after 20km SMF transmission is less than 3.8 × 10−3. Moreover, there is a negligible power penalty between real-time and off-line digital signal processing results.
Acknowledgments
This work is supported by the National Natural Science Foundation of China (61307087, 61377079), by Hunan Provincial Natural Science Foundation of China (12JJ3070), by the National “863” High Tech Research and Development Program of China (2011AA010203), by the Open Fund of State Key Laboratory of Information Photonics and Optical Communications (Beijing University of Posts and Telecommunications) and by the Fundamental Research Funds for the Central Universities and Young Teachers Program of Hunan University.
References and links
1. J. M. Tang and K. A. Shore, “30-Gb/s signal transmission over 40-km directly modulated DFB-laser-based single-mode-fiber Links without optical amplification and dispersion compensation,” J. Lightwave Technol. 24(6), 2318–2327 (2006). [CrossRef]
2. Y.-M. Lin and P.-L. Tien, “Next-generation OFDMA-based passive optical network architecture supporting radio-over-fiber,” IEEE J. Sel. Areas Comm. 28(6), 791–799 (2010). [CrossRef]
3. N. Cvijetic, “OFDM for Next-generation optical access networks,” J. Lightwave Technol. 30(4), 384–398 (2012). [CrossRef]
4. T. Takahara, T. Tanaka, M. Nishihara, Y. Kai, L. Li, Z. Tao, and J. Rasmussen, “Discrete Multi-Tone for 100 Gb/s Optical Access Networks,” in Optical Fiber Communication Conference, OSA Technical Digest (online) (Optical Society of America, 2014), paper M2I.1. [CrossRef]
5. W. Shieh and C. Athaudage, “Coherent optical orthogonal frequency division multiplexing,” Electron. Lett. 42(10), 587–589 (2006). [CrossRef]
6. D. Qian, J. Hu, J. Yu, P. N. Ji, L. Xu, T. Wang, M. Cvijetic, and T. Kusano, “Experimental demonstration of a novel OFDM-A based 10Gb/s PON architecture,” in Proceedings of European Conference and Exhibition on Optical Communication, (Berlin, 2007), paper Tu5.4.1.
7. F. Li, Z. Cao, J. Yu, X. Li, and L. Chen, “SSMI cancellation in direct-detection optical OFDM with novel half-cycled OFDM,” Opt. Express 21(23), 28543–28549 (2013). [CrossRef] [PubMed]
8. F. Li, J. Yu, Y. Fang, Z. Dong, X. Li, and L. Chen, “Demonstration of DFT-spread 256QAM-OFDM signal transmission with cost-effective directly modulated laser,” Opt. Express 22(7), 8742–8748 (2014). [CrossRef] [PubMed]
9. Y. Benlachtar, P. M. Watts, R. Bouziane, P. Milder, D. Rangaraj, A. Cartolano, R. Koutsoyannis, J. C. Hoe, M. Püschel, M. Glick, and R. I. Killey, “Generation of optical OFDM signals using 21.4 GS/s real time digital signal processing,” Opt. Express 17(20), 17658–17668 (2009). [CrossRef] [PubMed]
10. R. Schmogrow, M. Winter, B. Nebendahl, J. Meyer, M. Dreschmann, M. Huebner, J. Becker, C. Koos, W. Freude, and J. Leuthold, “101.5 Gbit/s real-time OFDM transmitter with 16QAM modulated subcarriers,” in Proceedings of Optical Fiber Communication Conference and Exposition and the National Fiber Optic Engineers Conference (Los Angeles, 2011), paper OWE5. [CrossRef]
11. X. Q. Jin, R. P. Giddings, E. Hugues-Salas, and J. M. Tang, “Real-time demonstration of 128-QAM-encoded optical OFDM transmission with a 5.25bit/s/Hz spectral efficiency in simple IMDD systems utilizing directly modulated DFB lasers,” Opt. Express 17(22), 20484–20493 (2009). [CrossRef] [PubMed]
12. M. Chen, J. He, and L. Chen, “Real-time optical OFDM long-reach PON system over 100-km SSMF using a directly modulated DFB laser,” J. Opt. Commun. Netw. 6(1), 18–25 (2014). [CrossRef]
13. M. Chen, J. He, Z. Cao, J. Tang, L. Chen, and X. Wu, “Symbol synchronization and sampling frequency synchronization techniques in real-time DDO-OFDM systems,” Opt. Commun. 326, 80–87 (2014). [CrossRef]
14. E. C. Ifeachor and B. W. Jervis, Digital Signal Processing: A Practical Approach (Addison-Wesley, Boston, 1993).
15. Y. Gao, J. Yu, J. Xiao, Z. Cao, F. Li, and L. Chen, “Direct-detection optical OFDM transmission system with pre-emphasis technique,” J. Lightwave Technol. 29(14), 2138–2145 (2011). [CrossRef]
