Open Access
Add to BibSonomy
Abstract
A twin-single-sideband (twin-SSB) signal single-photodiode (PD) detection system without optical bandpass filter is experimentally demonstrated for the first time. After direct detection by a single-ended PD at the receiver side, we can directly separate the optical left sideband (LSB) and right sideband (RSB) using a simple one-path digital signal processing algorithm without separating the two sideband signals using an optical bandpass filter (OBPF), thus achieving lower complexity and low cost while doubling the spectral efficiency. Using our proposed twin-SSB scheme, we demonstrate 1-, 2-, and 4-Gbaud LSB geometric shaping 4-quadrature amplitude modulation and RSB quadrature phase shift keying signal transmission over 10 km of single-mode fiber (SMF). Our experimental results demonstrate that the bit-error rate (BER) of the 4-Gbaud LSB geometric shaping 4-quadrature amplitude modulation (GS-4QAM) and RSB quadrature phase shift keying (QPSK) transmission system is below the 7% hard decision forward error correction threshold.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Distributed data centre architectures are increasingly being used to support performance-critical applications such as commercial cloud services and big data [1]. Many machine-to-machine inputs and outputs are generated among servers. Global content network and network-scale companies need to connect these data centers in the form of a distributed cluster (within 80 km), requiring a large connection bandwidth and low cost [2]. Among the proposed solutions, coherent detection technologies have more degrees of freedom, whereas the intensity modulation and direct detection (IM/DD) technique has the advantages of low cost and simple structure [3]. However, the double sideband (DSB) signal will suffer serious frequency-selected power fading, caused by chromatic dispersion (CD), in a conventional IM/DD system, therefore limiting the transmission capacity and distance of the system and therefore it is unlikely to satisfy the transmission requirements of data center inter-connects [4]. In this context, optical single-sideband (SSB) transmission has recently attracted a great deal of attention, since this transmission scheme not only overcomes the problem of dispersion-induced radio frequency (RF) power fading but can also double the spectral efficiency (SE) [5–13]. Recently, optical twin-SSB modulation has been used as an extension to optical SSB modulation [14–21].
Twin-SSB signals can be generated based on an in-phase-quadrature (I/Q) modulator, and two independent left sideband (LSB) and right sideband (RSB) signals are used to carry individual data to effectively harvest the advantages of twin-SSB modulation, which yields higher spectral efficiency. However, the dual SSB signal at the receiving end needs to use two optical filters to separate and detect the LSB and the RSB separately, which increases the system cost and prevents the system from being applied in low-cost, short-distance optical communication systems [14,15].
Over the past five years, many researchers have studied the above challenges, and double-SSB transmission schemes with simplified structure have been presented through simulation and experiments. In 2018, Deng et al. proposed and experimentally demonstrated a heterodyne W-band fiber-wireless system using a twin-SSB orthogonal frequency division multiplexing (OFDM) transmission system with low-cost electrical filters, which could reach a 40.07 Gb/s data rate with a BER below the 7% hard decision forward error correction (HD-FEC) threshold of 3.8 × 10−3 after SSMF transmission over 22 km and wireless transmission over 1 m [14]. In 2020, Li et al. proposed a scheme for the asymmetric direct detection of twin-SSB signals based on a simple receiver front-end, which was composed of one optical filter and two photodiodes (PDs). Asymmetric direct detection exploits the difference in the photocurrent between a filtered and unfiltered signal pair to reconstruct and linearize the received twin-SSB signal with high electrical spectral efficiency [15]. In 2021, Nakagawa et al. first demonstrated the detection of an optical twin-SSB signal without an optical filter using a homodyne receiver through a new method of electrical butterfly operation [16]. In 2022, Zhou et al. designed a twin-SSB system with different modulation formats in the two sidebands, adopting GS-3 phase shift keying (3PSK) modulation for the LSB and quadrature phase shift keying (QPSK) modulation for the RSB [17]. Based on the beating characteristics of the LSB and RSB, their simulations showed that the proposed method could separate the two independent sideband signals using DSP algorithms at the receiver end.
In this paper, we experimentally demonstrate for the first time a single-PD detection system for twin-SSB signals without OBPF, and describe in more detail the inaccuracies noted in the simulation presented by Zhou et al [17]. Our experimental results show that in the absence of OBPF, the double-sideband signal can still achieve good transmission quality after being monitored by a single PD. For an ROP of 0 dBm, when the center frequency interval of the RSB and LSB vector signals is set to 8, 12, and 16 GHz, the BERs of 4-GBaud LSB- geometric shaping 4-quadrature amplitude modulation (GS-4QAM) and RSB-QSPK vector signals after 10 km of SMF transmission are 9.63 × 10−3, 1.03 × 10−2, 4.68 × 10−3, 8.01 × 10−3, 2.97 × 10−3, 3.36 × 10−3, respectively. When the bandwidth of the receiver oscilloscope is constant, changes in the symbol rates of the LSB-GS-4QAM and RSB-QSPK vector signals can affect the transmission quality of the system more than changes in the center frequency interval between LSB and RSB.
The remainder of this paper is organized as follows. Section 2 describes the principle of the two-sideband vector signal generation based on an I/Q modulator and theoretical derivations for separating the LSB and RSB signals by DSP. Section 3 introduces the experimental setup for the system. In Section 4, evaluations of the BER performance for LSB GS-4QAM and RSB QPSK signals are provided for different SMF transmission distances and different baud rates. Finally, our conclusions are presented in Section 5.
2. Principle of operation of our twin-SSB scheme without OBPFs separating the two sidebands
Figure 1 shows our scheme in which two independent LSB and RSB vector signals are generated based on an I/Q modulator. The driving signal I-path and Q-path for the I/Q modulator were generated via DSP. At the transmitter end, two sets of twin-SSB pseudo-random binary sequences (PRBSs) of the same length, denoted by PRBS1 and PRBS2, are individually vector mapped, up-sampled, and raised cosine (RC)-shaped in order to generate independent LSB and RSB vector signals. They are then used to modulate intermediate-frequency carriers in the form of complex sinusoidals, $\exp ({j2\pi {f_c}t} )$ and $\exp ({ - j2\pi {f_c}t} )$, to generate the LSB and RSB signals, respectively [18–21].
Fig. 1. LSB and RSB signal generation based on optical carrier suppression (OCS) and twin-SSB modulation by an I/Q modulator (PRBS: pseudo-random binary sequence; RC: raised cosine; ECL: external cavity laser; PM: phase modulator; MZM: Mach–Zehnder modulator): (a) schematic diagram of the RSB signal; (b) schematic diagram of the LSB signal; (c) schematic diagram of the received signal after the PD
Here, the absolute value of the carrier frequency is the same for the left and right bands. In practice, the carrier frequencies for the LSB and RSB vector signal may be identical or different. The LSB and RSB carrier signals can be expressed as
where ${E_l}(t )$ and ${E_r}(t )$ represent the LSB and RSB signals, respectively. Schematics of the upconverted LSB and RSB signals are given in Figs. 1(a) and (b), respectively. The two sideband signals are then added, and can be expressed asAs shown in Fig. 1, the real and imaginary parts of $E(t )$ are added to the I and Q ports of the I/Q modulation through the DAC. The on-wave light generated by the ECL is represented as ${E_{CW}}(t )$, its center frequency is ${f_0}$, and the angular frequency of the LSB is ${\omega _l}$, the angular frequency of the RSB is ${\omega _r}$, the phase of the LSB is ${\varphi _l}$, and the phase of the RSB is ${\varphi _r}$. The modulated optical signal can then be expressed as
It can be observed from Eq. (5) that the phase of the received electric vector signal (${\varphi _l}(t )+ {\varphi _r}(t )$) is the sum of the phases of the LSB and RSB signals ${\varphi _l}(t )$ and ${\varphi _r}(t )$, and the amplitude is related to ${J_{ - 1}}({\beta {A_l}} ){J_1}({\beta {A_r}} )$. In this paper, we choose LSB GS-4QAM and RSB QPSK for our experiments, and the constellation diagram is shown in Fig. 2. Figures 2(a)–(c) illustrate the phase relationship between the 16QAM vector signal and the LSB GS-4QAM and RSB QPSK vector signals, using different colors. Since the color correspondence reflects the mapping relationship between the received signal and the two sideband signals, the received signal can be separated at the receiver. For example, consider the constellation point within the red box in Fig. 2(c). It is composed of black and yellow colors, which indicates that it is generated by the beating of the LSB GS-4QAM (yellow) signal constellation point and the RSB QPSK (black) signal constellation point. Hence, at the receiver side, after the 16QAM signal is equalized to remove the noise of other frequency components, the 16QAM signal can be separated into (1,0) GS-4QSM and (1,1) QPSK signals according to the corresponding relationship shown in Fig. 2. This means that two OBPFs and PDs are not required to separate the LSB and RSB signals. In addition, since only one mixed-signal path is sent to MATLAB for DSP, the complexity of the received-side calculation is reduced.
Fig. 2. Constellations of GS-4QAM, QPSK, and 16QAM vector signals: (a) constellations of the GS-4QAM vector signal, denoted by different colors; (b) constellations of the QPSK vector signal, denoted by different colors; (c) constellations of the 16QAM vector signal after the PD, denoted by different colors.
3. Experimental setup and results
Figure 3 shows the experimental setup for our novel twin-SSB scheme without OBPFs separating the two sideband signals. The CW is generated by ECL with a line width of less than 100 kHz and an output power of less than 15 dBm. The 1550 nm light source is then modulated by the I/Q modulator to realize photoelectric conversion. The offline DSP operation was implemented using MATLAB, with a PRBS word length of 214. GS-4QAM modulation was adopted for the LSB vector signal (see Fig. 3 for a constellation diagram for GS-4QAM), while QPSK modulation was adopted for the RSB vector signal (see Fig. 3 for a constellation diagram for QPSK). A 64GSa/s AWG was used to convert the LSB and RSB signals into analog signals, which were loaded into the IQ modulator after passing through an electrical amplifier (RC IQ filters, roll-off = 0.2). The I/Q modulator had a 3 dB bandwidth of 33 GHz and an insertion loss of 6 dB. The measured LSB and RSB spectra are shown in Figs. 3(a) and 3(b), respectively. The transmission link was composed of 10 km SMF (G 652 with fiber loss, dispersion, and nonlinearity coefficient 0.18 dB/km, 17 ps/(nm.km), and 1.3 W−1 km−1, respectively). A variable optical attenuator (VOA) was used to adjust the optical signal power before the PD. The optical signals were detected by a single-ended PD, and Fig. 3(c) shows the calculated output electrical signal spectrum of the PD. A four-channel digital phosphor oscilloscope with a sampling rate of 100 GSa/s and a bandwidth of 20 GHz as ADC was used to receive the signal. The offline digital signal was then processed by MATLAB. Offline DSP included the filter, frequency shift, clock phase recovery, CMA, and BPS. The CMA algorithm has 31 taps with a step factor of 0.00001. The BPS is implemented with 64 test phases. The part of the signal that was not expected by the first term in Eq. (5) was removed by filtering, and the first term in Eq. (5) was then recovered by algorithms, and can be regarded as a 16QAM signal.
Fig. 3. Experimental setup for our novel twin-SSB system with OCS modulation enabled by an I/Q modulator: (a) spectrum of the LSB GS-4QAM signal; (b) spectrum of the RSB QPSK signal; (c) spectrum of the received 16QAM signal. EA: electrical amplifier; VOA: variable optical attenuator. CMA: constant modulus algorithm. BPS: blind phase search
Figure 4 shows the BER for LSB-GS-4QAM and RSB-QSPK vector signals after 10 km optical fiber transmission for ${f_c}$ of 4, 6 and 8 GHz and a transmission rate B of 4 GBaud. It can be clearly seen that the BER of LSB-GS-4QAM and RSB-QSPK vector signals decreases faster with increasing ROP as ${f_c}$ increases. Under these conditions, the BER of the LSB-GS-4QAM vector signals is slightly lower than that of the RSB-QSPK vector signals. It can be seen from the constellation diagram in Fig. 5 that when ${f_c}$ is 4 GHz, the center frequencies of the RSB and LSB signals are closely spaced, and the constellation diagram reflects the high level of nonlinear noise in the signals after the beat frequencies of the two at the receiving end. When the vector signal transmission rate B remains unchanged, ${f_c}$ increases and the nonlinear noise of the signal obtained by the receiving end is smaller. When ROP is 0dBm and ${f_c}$ is 4, 6, and 8 GHz, the BERs of the LSB-GS-4QAM and RSB-QSPK vector signals at 4 GBaud are: 9.63 × 10−3, 1.03 × 10−2, 4.68 × 10−3, 8.01 × 10−3, 2.97 × 10−3, 3.36 × 10−3, respectively.
Fig. 4. BER versus received optical power (ROP) with varying ${f_c}$ for SMF transmission over 10 km
Fig. 5. (a), (b), and (c) show constellation diagrams for 4 Gbaud LSB-GS-4QAM modulated and RSB-QPSK modulated vector signals after SMF transmission over 10 km for ${f_c}$ values of 4, 6, and 8 GHz at a 0dBm ROP.
Figure 6 shows the BERs for the LSB-GS-4QAM and RSB-QSPK vector signals after 10 km fiber transmission for transmission rates B of 1, 2, and 4 GBaud and ${f_c}$/B = 2 Hz/Baud. For a fixed value of ${f_c}$/B, the six curves in Fig. 6 show a similar decreasing trend in the BER with an increase of ROP. Unlike the results shown in Fig. 4, we do not see a lower overall BER of the double unilateral signal for a larger ${f_c}$. When the rate of the twin-SSB signal increases, even if ${f_c}$ increases linearly, its BER is still higher than that of the low-rate signal. It can be seen from Fig. 6 that when the symbol rate of the twin-single-sideband signals is 1 GBaud, for an ROP of −4 dBm, the BER is lower than the 7% HD-FEC threshold, and when the symbol rate is 4 GBaud, the ROP needs to be higher than −0.5 dBm. At the 7% HD-FEC threshold of 3.8 × 10−3, compared to the 1 Gbaud case, there are power penalties of 2.1 and 4.5 dB for the 2 and 4 Gbaud cases, respectively.
In Fig. 7, the center frequencies of the LSB vector signals and the RSB vector signals of the three groups of twin-SSB signals differ by 12 GHz, that is, ${f_c}$= 6 GHz. As can be seen from Fig. 7, the lower the LSB and RSB vector signal rates (i.e., the larger the guard interval between the two vector signal spectra), the lower the BER for a fixed value of the ROP. It can be seen from the experimental results in Figs. 5 to 7 that though OPBF is not used at the receiving end, the proposed twin-SSB signal single-PD detection system can separate the vector signals carried by LSB and RSB at the receiving end through the use of algorithms.
4. Conclusion
Existing twin-SSB schemes include two OBPFs and PDs to filter and detect LSB and RSB signals, which increases the system complexity, system cost, and algorithm complexity. We experimentally demonstrate a novel twin-SSB scheme in which no OBPFs are needed to separate the two independent sideband signals and only a single-ended PD is employed for direct detection. It can be seen from our experimental results that the system demonstrated in this paper can achieve low BER transmission below the 7% HD-FEC threshold when the center frequency spacing between the LSB and RSB signals is greater than twice the signal spectrum width. When the ROP is 0dBm and ${f_c}$ is 8 GHz, the BER values for the LSB-GS-4QAM and RSB-QSPK vector signals at 4 GBaud are 2.97 × 10−3 and 3.36 × 10−3, respectively. We believe that this twin-SSB system further reduces the complexity and cost of the system structure, meaning that it has broad applications for short-distance transmission in the future.
Funding
National Outstanding Youth Science Fund Project of National Natural Science Foundation of China (62022016); National Natural Science Foundation of China (61835002, 61727817); Beijing Municipal Natural Science Foundation (4222075); Funds for Creative Research Groups of China (62021005).
Disclosures
The authors declare no conflicts of interest.
Data availability
The data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. M. Filer, J. Gaudette, Y. Yin, D. Billor, Z. Bakhtiari, and J. L. Cox, “Low margin optical networking at cloud scale,” J. Opt. Commun. Netw. 11(10), C94–108 (2019). [CrossRef]
2. F. Wang, H. Yao, Q. Zhang, J. Wang, R. Gao, D. Guo, and M. Guizani, “Dynamic distributed multi-path aided load balancing for optical data center networks,” IEEE T. Netw. Serv. Man. 19(2), 991–1005 (2022). [CrossRef]
3. M. Xian, Q. Zhuge, Z. Xing, K. Zhang, T. M. Hoang, F. Zhang, and D. V. Plant, “Experimental study of performance enhanced IM/DD transmissions based on constellation switching,” Opt. Express 26(12), 15480–15489 (2018). [CrossRef]
4. B. Lin, J. Li, H. Yang, Y. Wan, Y. He, and Z. Chen, “Comparison of DSB and SSB transmission for OFDM-PON,” J. Opt. Commun. Netw. 4(11), B94–B100 (2012). [CrossRef]
5. W. Wang, D. Zou, Z Li, Q. Sui, Z. Cao, C. Lu, F. Li, and Z. Li, “Optical single sideband signal reconstruction based on time-domain iteration,” J. Lightwave Technol. 39(8), 2319–2326 (2021). [CrossRef]
6. X. Chen, C. Antonelli, S. Chandrasekhar, G. Raybon, A. Mecozzi, M. Shtaif, and P. Winzer, “Kramers–Kronig receivers for 100-km datacenter interconnects,” J. Lightwave Technol. 36(1), 79–89 (2018). [CrossRef]
7. T. Bo and H. Kim, “Kramers-Kronig receiver operable without digital upsampling,” Opt. Express 26(11), 13810–13818 (2018). [CrossRef]
8. T. Bo and H. Kim, “CSPR measurement method for optical single-sideband signal,” IEEE Photonics Technol. Lett. 31(13), 1100–1103 (2019). [CrossRef]
9. L. Shu, J. Li, Z. Wan, Z. Yu, X. Li, M. Luo, S. Fu, and K. Xue, “Single-photodiode 112-Gbit/s 16-QAM transmission over 960-km SSMF enabled by Kramers-Kronig detection and sparse I/Q Volterra filter,” Opt. Express 26(19), 24564–24576 (2018). [CrossRef]
10. T. M. Hoang, M. Y. S. Sowailem, Q. Zhuge, Z. Xing, M. Morsy-Osman, E. El-Fiky, S. Fan, M. Feng, and D. V. Plant, “Single wavelength 480 Gb/s direct detection over 80 km SSMF enabled by Stokes vector Kramers Kronig transceiver,” Opt. Express 25(26), 33534–33542 (2017). [CrossRef]
11. X. Li, Y. Xu, and J. Yu, “Single-sideband W-band photonic vector millimeter-wave signal generation by one single I/Q modulator,” Opt. Lett. 41(18), 4162–4165 (2016). [CrossRef]
12. Y. Xu, X. Li, J. Yu, and G.-K. Chang, “Simple and reconfigured single-sideband OFDM RoF system,” Opt. Express 24(20), 22830–22835 (2016). [CrossRef]
13. X. Li, J. Yu, J. Zhang, J. Xiao, Z. Zhang, Y. Xu, and L. Chen, “QAM vector signal generation by optical carrier suppression and precoding techniques,” IEEE Photonics Technol. Lett. 27(18), 1977–1980 (2015). [CrossRef]
14. R. Deng, J. Yu, and J. He, “Twin-SSB-OFDM transmission over heterodyne W-band fiber-wireless system with real-time implementable blind carrier recovery,” J. Lightwave Technol. 36(23), 5562–5572 (2018). [CrossRef]
15. X. Li, M. Zhu, Z. Xing, M. O’Sullivan, and D. V. Plant, “Asymmetric direct detection of twin-SSB signals,” Opt. Lett. 45(4), 844–847 (2020). [CrossRef]
16. . R. Nakagawa, Y. Kurokawa, K. Tsumura, and M. Nakamura, “Experimental demonstration of optical twin-SSB detection scheme using an electric butterfly operation,” in Optoelectronics and Communications Conference, Poster Session II: Transmission Systems and Technologies (Optical Society of America, 2021) paper JS2B. 7.
17. Y. Zhou, J. Xiao, C. Zhao, J. Zuo, J. Ming, and L. Zhao, “Simplified twin-single-sideband direct detection system separating left and right sideband signals using a digital signal processing algorithm instead of optical bandpass filters,” Opt. Express 30(1), 619–628 (2022). [CrossRef]
18. Z. Xu, R. Hui, and M. O’Sullivan, “Dual-band OOFDM system based on tandem single-sideband modulation transmitter,” Opt. Express 17(16), 13479–13486 (2009). [CrossRef]
19. L. Zhang, Q. Zhang, T. Zuo, E. Zhou, G. N. Liu, and X. Xu, “C-band single wavelength 100-Gb/s IM-DD transmission over 80-km SMF without CD compensation using SSB-DMT,” in Optical Fiber Communication Conference, OSA Technical Digest (online) (Optical Society of America, 2015), paper Th4A.2.
20. X. Pan, B. Zhang, and X. Liu, “Mitigation of crosstalk between RSB and LSB signals generated by one I/Q modulator,” Opt. Express 28(2), 2238–2250 (2020). [CrossRef]
21. X. Pan, X. Liu, H. Zhang, K. Wang, Y. Zhang, D. Ran, X. Wang, and C. Wang, “Independent dual single-sideband vector millimeter-wave signal generation by one single I/Q modulator,” Opt. Express 27(14), 19906–19914 (2019). [CrossRef]
