SYSTEM AND METHOD FOR 400G SIGNAL GENERATION

AND COHERENT DETECTION

FIELD OF THE INVENTION

[0001 ] The field of the present invention is communication systems, and particularly, systems and method for 400G Signal generation and coherent detection.

Introduction

[0002] With 00G coherent transport technologies based on PM-QPSK being commercialized at a rapid pace, it is believed that 400G transport per channel is likely one of the potential successors for the next-step capacity expansion ¹ . The technology options for 400G transmission are being intensively studied among industrial research communities, which can be realized by increasing the symbol rate, increasing the number of bonded channels, increasing the number of modulation levels, or their combinations ^2"4 . The approach using singe-carrier PM- 16QAM at 56 Gbaud per wavelength is mainly limited by the insufficient bandwidth of a series of devices from end to end such as electrical multiplexers, optical modulators, and analog to digital converters ⁵ . Although methods with reduced or low bandwidth requirement like dual-carrier PM-16QAM at 28 Gbaud or 5-carrier PM- 32QAM at 9 Gbaud have been presented with spectral efficiency (SE) up to 8 b/s/Hz, unfortunately, the demonstrated maximum transmission distance ranging from 800 to 1200 km can hardly be achieved without employing special ultra-large-area fibers (ULAF) and Raman amplification to deal with the reduced nonlinearity tolerance and higher OSNR requirement. On the other hand, packing a number of established PM- QPSK channels in the way of Nyquist wavelength division multiplexing (N-WDM) is considered as a practical beyond- 00G solution. N-WDM technology relies on pulse shaping via spectral filtering to optimize the trade-off between the SE and the impairment induced by inter-channel interference (ICI) and inter-symbol interference (ISI), and the suggested ratio of channel spacing to symbol rate is around 1.1. [xxx] Further reducing that ratio to achieve a higher SE will require innovation on DSP algorithms at the receiver to suppress the undesired noise and linear crosstalk linear equalization process .

[0003] In this disclosure, a 400G generation and transmission solution based on quad-carrier PM-QPSK at the total channel line rate of 512 Gb/s are

demonstrated. For the first time, the channel/carrier-spacing-to-symbol-rate ratio was down to only 0.78, yielding a net SE of 4 b/s/Hz. Two 400G signals with no guard band in between were delivered over a re-circulating loop comprising SMF-28 fiber spans, EDFAs, and a 100-GHz-wide wavelength selective switch (WSS) to emulate the bandwidth narrowing effect as a result of cascade inline ROADMs. The results show that, even for the channel at the passband edge of WSS, over 2400-km reach was achieved with a BER below the soft-decision pre-FEC limit of 2x10 ^'2 .

SUMMARY OF THE INVENTION

[0004] Aspects of the present invention include systems and methods for generating and detecting signals in communication systems, in one aspect, a system and method for 400G signal generation and coherent detection.

[0005] In another particular aspect, a novel digital signal processing scheme (DSP) for quadrature duobinary (QDB) spectrum shaped polarization multiplexed quadrature phase shift keying (PM-QPSK) based on multi-modulus blind

equalizations (MMBE) is demonstrated with both simulation and experimental results. The key algorithms for this novel digital signal processing scheme include the cascaded multi-modulus algorithm (CMMA) for blind polarization de-multiplexing, multi-modulus QPSK partitioning frequency offset estimation (FOE) and two stage carrier phase recovery (CPR) with maximum likelihood phase estimation. The final signal is detected by maximum-likelihood sequence detection (MLSD) for data BER measurement. The feasibility of the proposed digital signal processing scheme is demonstrated by the experiment of 1 12Gb/s QDB spectrum shaped PM-QPSK signal with a 25GHz bandwidth waveshaper for Nyquist WDM channels.

BRIEF DESCRIPTION OF THE DRAWINGS

[0006] Figure 1 illustrates the experiment setup for 400G transmission based on quad-carrier PM=QPSK with 4-b/s/Hz net SE.

[0007] Figure 2 is the illustration of 9-QAM signal generation by digital T-DAF.

[0008] Figure 3 is the illustration of Back-to-back BER Curves. [0009] Figure 4 illustrates the measured Q ² factor and OSNR degradation as a function of wavelength offset.

[0010] Figure 5 illustrates the BER performance vs. transmission distance.

[001 1 ] Figure 6 illustrate the setup of the 400G transmitter and receiver.

[0012] Figure 7 illustrates component 101 of the setup of the 400G transmitter and receiver shown in Figure 6.

Experimental Setup and Results

[0013] Fig. 1 shows the experimental setup for 400G transmission based on quad-carrier PM-QPSK at 4 b/s/Hz, which contains 100 GHz inline WSS in recirculating loop to study on the maximum transmission distance and also the highest number of ROADMs can be potentially achieved and passed. 8 external cavity lasers (ECLs) are used as a CW light source array and grouped them into even and odd channels spaced at 25 GHz. Each of them has a linewidth less than 100 kHz and the total output power of 14.5 dBm. The even/odd light sources were combined using a 50:50 polarization-maintaining optical couplers (OCp _M ). The channel modulation comprises in-phase/quadrature-phase modulators (IQ Mod.) and polarization multiplexers (P-MUX). Each IQ Mod. was driven by two sets of de-correlated pseudorandom bit sequence (PRBS) at 32 Gb/s with a word length of 2 ¹ -1. After that, the line rate of each QPSK channels was doubled by using P-MUX. All the PM- QPSK channels at 128 Gb/s, under the assumption of both 24% SD-FEC and 4% protocol overheads, were aggressively shaped in spectrum domain and

simultaneously combined via a 25-GHz wavelength-selective switch (WSS). The aggregated channels were boosted by an Erbium doped fiber amplifier (EDFA) giving a launch power of -1 dBm per channel. In the end, there were 8x128 Gb/s channels generated on a 25-GHz grid. This will be viewed as two independent 400G (512-Gb/s quad-carrier PM-QPSK) channels, as illustrated in the inset of Fig. 1 , with no guard band in between, and each of which occupied 100-GHz spectral width.

[0014] The re-circulating loop consists of 5 spans of 80-km SMF-28 with an average span loss of 17.5 dB and the chromatic dispersion coefficient of 17- ps/km/nm, loop control switch, and dual-stage C-band EDFAs It is note that there are no any inline dispersion compensation modules. More importantly, in order to study the tolerance of the 512-Gb/s quad-carrier PM-QPSK signal against cascade narrowing filter effect along the transmission path, another WSS was placed in the loop, which was programmed to be a 100-GHz add/drop filter. The two 512-Gb/s signals were fed into the common port of the WSS, and they were then filtered and routed to port 1 and port 2, respectively. Both signals were re-combined again by using a 50:50 OC with one-arm delayed for de-correlation.

[0015] Prior to the coherent receiver (Co. Rx), a tunable optical bandpass filter (OBPF) with 3-dB bandwidth of 0.4 nm was employed to choose the interest of PM- QPSK channel. The optical front-end of the Co. Rx consisted of a fiber-laser local oscillator (LO) with a linewidth less than 100 kHz, a 90-degree hybrid for polarization and phase diversities, and four balanced photo-detectors. The output waveforms were captured using a real-time scope with an analog to digital converter (ADC) operated at 50 GSa/s, and a bandwidth of 18 GHz. For the digital signal processing (DSP), first, the clock was extracted by using "square and filter" method, and then the digitized signal was resampled at twice of the symbol rate based on the recovered clock. Second, a 772-spaced time-domain finite impulse response (FIR) filter was utilized for electronic dispersion compensation (EDC). Third, the

polarization recovery and residual chromatic dispersion compensation by using classic constant modulus algorithm (CMA) was done with 23-tap, T/2-spaced adaptive FIR filters. Finally, carrier recovery including frequency offset estimation by fast Fourier transform method and carrier phase recovery by 4th power Viterbi- Viterbi algorithm were performed.

[0016] In addition to those typical algorithms for PM-QPSK demodulation, in this particular research work we additionally incorporated a linear T-spaced delay- and-add filter (Γ-DAF) to the DSP flow after the carrier phase estimation stage to suppress the undesired noise and linear crosstalk enhancement caused by the use of linear equalizers in the presence of aggressive channel filtering ⁶ . Such T-DAF also made possible the employment of MLSE with a short memory length of only 1 , further compensating transmission impairment. From the constellation point of view, the effect of this 7-DAF digital filter turns the originally recovered 4-point QPSK to 9- point quadrature duobinary signals. The evolution of this transformation is illustrated in Fig. 2. As a result of the delay-and-add effect, the 2-ASK in-phase and quadrature components disappear and independently change into two 3-ASK symbol series. The generation mechanism of 9-QAM signals can be considered as the superposition of two 3-ASK vectors on a complex plane, and the size of constellation points represents the relative number of points generated after the T-DAF digital filtering. Fig. 3 shows the BTB BER curves with and without adjacent channels at 32GBaud. At 2x10 ^'2 BER, the required OSNR are 14.9 and 15.8 dB for single- channel and N-WDM cases, respectively. The insets show the received

constellations of original QPSK and duobinary-shaped QPSK, respectively, at 19.2 dB OSNR in N-WDM case. Since in the proposed N-WDM transmission, the ratio of channel spacing to symbol rate is as low as 0.78, it is then important to know the impact of wavelength offset between each PM-QPSK channel and corresponding passband window of WSS on the system performance. Fig. 4 shows the Q ² factor and OSNR degradation as a result of the wavelength offset (minus offset show symmetry impact) in a BTB single-channel measurement. With wavelength offset less than 3 GHz, the Q ² factor drops by 0.5 dB/GHz in average, and it will deteriorate to 1.1 dB/GHz beyond 3 GHz. The inset shows the measured optical spectra of filtered PM-QPSK channel at 128 Gb/s with and without the wavelength offset. Fig. 5 shows the measured BER at multiple distance measuring points for different channels. Below 2x10 ^'2 BER, Ch #2 can be delivered over 2800 km, while in the case of the Ch #4, the maximum distance was shortened by 400 km. This is because Ch #4 was located at the WSS passband edge, which inherently suffers from more bandwidth narrowing penalty. However, evidently 400G transmission using quad- carrier PM-QPSK approach can at least reach 2400 km on a 100-GHz grid with a net spectral efficiency of 4 b/s/Hz. The inset shows the received optical spectrum after 2400 km transmission with 20.4-dB delivered OSNR.

[00 7] Setup for the 400G transmitter and receiver are shown in Fig. 6 having components identified as 101 , 102, 103, 104, 105, and 106. In more detail, the component 101 in Fig. 6 is illustrated in Fig. 7. More detailed discretions of the components are as follows:

[0018] For 101 , as shown in Fig. 7, the laser can be any laser, such as distributed feedback laser diode, external cavity laser with linewidth smaller than 10MHz. I/Q modulator is driven by I and Q electrical signal. The bit rate of I or Q can be 25~ 34Gb/s binary on/off keying data. Polarization beam combiner is used to combine the two polarization signal. The neighboring laser spacing is 25GHz so that it will occupy 100GHz spacing for 400Gb/s signal. [0019] For 102, it can be spliced optical coupler, optical interleaver, arrayed waveguide grating or other optical coupler.

[0020] For 103, optical amplifier, it can be erbium doped fiber amplifier, Raman amplifier or other optical amplifier.

[0021] For 104, optical fiber, it can be standard single mode fiber or other special fiber.

[0022] For 105, optical filter, it be tunable optical filter, WDM coupler, arrayed waveguide grating or other optical filter. The system works even without any optical filter.

[0023] For 106, coherent detector, details on regular coherent detection with special digital signal process are provided in US provisional application 61/581 ,946, filed December 30, 2011 , entitled DIGITAL FILTER, PARTIAL RESPONSE

EQUALIZER, AND DIGITAL COHERENT RECEIVER DEVICE AND METHOD, incorporated hereinwith in its entirety.

Conclusions

[0024] A novel digital signal processing scheme (DSP) for quadrature duobinary (QDB) spectrum shaped polarization multiplexed quadrature phase shift keying (PM-QPSK) based on multi-modulus blind equalizations (MMBE) is demonstrated with both simulation and experimental results. The key algorithms for this novel digital signal processing scheme include the cascaded multi-modulus algorithm (CMMA) for blind polarization de-multiplexing, multi-modulus QPSK partitioning frequency offset estimation (FOE) and two stage carrier phase recovery (CPR) with maximum likelihood phase estimation. The final signal is detected by maximum-likelihood sequence detection (MLSD) for data BER measurement. The feasibility of the proposed digital signal processing scheme is demonstrated by the experiment of 1 12Gb/s QDB spectrum shaped PM-QPSK signal with a 25GHz bandwidth waveshaper for Nyquist WDM channels.

[0025] In another aspect, experimental demonstration is made for a long- reach 400G transmission using Nyquist quad-carrier PM-QPSK technique with a net spectral efficiency of 4 b/s/Hz. With T-spaced delay-and-add filter (T-DAF) and 1-bit MLSE added to the DSP flow at the coherent receiver, the enhanced noise and linear crosstalk can be effectively suppressed even at a low channel-spacing-to- symbol-rate ratio of 0.78. The results show that, even with bandwidth narrowing effect, 2400-km SMF-28 transmission of 512-Gb/s quad-carrier PM-QPSK signals with BER less than the SD pre-FEC limit of 2x10 ^~2 was realized.

[0026] It should be understood that the methods and devices of the present invention may be executed employing machines and apparatus including simple and complex computers. Moreover, the architecture and methods described above can be stored, in part or in full, on forms of machine-readable media. For example, the operations of the present invention could be stored on machine-readable media, such as magnetic disks or optical disks, which are accessible via a disk drive (or computer-readable medium drive). Alternatively, the logic to perform the operations as discussed above, could be implemented in additional computer and/or machine readable media, such as discrete hardware components as large-scale integrated circuits (LSI's), application-specific integrated circuits (ASIC's), firmware such as electrically erasable programmable read-only only memory (EEPROM's); and the like. Implementations of certain embodiments may further take the form of machine- implemented, including web-implemented, computer software.

[0027] While aspects of this invention have been shown and described, it will be apparent to those skilled in the art that many more modifications are possible without departing from the inventive concepts herein. The invention, therefore, is not to be restricted except in the spirit of the following claims.

Acknowledgements

[0028] This research study is supported by National High Technology

Research and Development Program of China (No. 2012AA011303).

References

[1 ]X. Zhou et al., Proc. ECOC'1 1 , We.8.B.2, (201 1 ).

[2]V.A.J.M. Sleiffer et al., Proc. OFC'12, OW4C.3, (2012).

[3] F. Buchali et al. , Proc. OFC'12, OW4C.4, (2012).

[4]X. Liu et al., Proc. OFC'1 1 , JThA37 (201 1 ).

[5]P. J. Winzer et al., Proc. ECOC'10, PDP 2.2, (2010).

[6] J. Li et al., Proc. OFC'12, OM3H.2, (2012).