PAM-4 Signal Research Articles

An 80-Gb/s PAM-4 Simultaneous Bidirectional Transceiver With Hybrid Adaptation Scheme

This paper presents a simultaneous bidirectional (SBD) transceiver with four-level pulse amplitude modulation (PAM-4), employing a novel hybrid adaptation scheme. The possibility of extending bandwidth is explored by applying PAM-4 signaling to SBD. Furthermore, a mismatch compensation method for a hybrid circuit is proposed, which is essential for removing the outbound signal in SBD. The hybrid adaptation is easily implemented by applying the locked condition of the Mueller M|ller phase detector (MMPD) with the data level adaptation. By sharing one error sampler with MMPD and adaptation engine, the presented SBD transceiver is efficient in terms of clocking power consumption and bandwidth of a receiver front-end. A wide linear range hybrid and a data alignment technique guarantee the robust PAM-4 SBD operation. Fabricated in the 28-nm CMOS, the 80-Gb/s SBD transceiver achieves a bit error rate (BER) of less than 10-12 with an energy efficiency of 2.65 pJ/b.

High-Data-Rate and Wide-Steering-Range Optical Wireless Communication via Nonuniform-Space Optical Phased Array

The silicon photonic optical phased array (OPA) featured by high integration and agile beam steering exhibits greatly potential application in optical wireless communication (OWC). In this paper, we demonstrate high-data-rate and long-range OWC using an integrated Silicon-Silicon Nitride OPA chip with 12 mm <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> large aperture. The OPA chip constituted by 128 waveguide grating antennas at nonuniform spacing is employed for wide-steering and small-divergence solid beam steering. In the experiment, we demonstrated 100° wide range communications in horizontal direction via antenna phase tuning. To our best knowledge, 100° is the widest cover range in the reported OPA-based OWC systems. Moreover, the OPA chip is also utilized to achieve up to 32 Gb/s low BER transmission of non-return-to-zero (NRZ) signal over 54 m free space distance. In addition, we also successfully demonstrated more than 50 Gb/s PAM4 signal transmission by measuring clear eye diagrams at the horizontal direction of 0° and 50° over 10 m. Thanks to agile beam steering of the OPA, switching time among arbitrary positions in multi-targets communication can be less than 27 μs. By virtue of tiny beam divergence 0.021° × 0.029° emitted from the OPA, data transmission switching between two targets spaced at 3.5 cm over 20 m can be achieved accurately. The experimental measurement shows the sidelobe crosstalk from the adjacent target channel is less than -25.2 dB. Our demonstrations validate the feasible implementation of the silicon-based OPA in the high-data-rate and wide-steering-range OWC system.

Adaptive Hybrid Iterative Linearization Algorithms for IM/DD Optical Transmission Systems

For optical field recovery and linear dispersion compensation, we propose a performance-enhanced linearization algorithm, termed adaptive hybrid multi-constraint iteration algorithm (MCIA), which does not require any physical modifications to standard configurations of intensity-modulation and direct-detection (IM/DD) transmission systems. To improve the sensitivity to the residual inter-symbol interference (ISI) effect, we introduce, after fiber backward-propagation, a linear feed-forward equalizer (FFE) pair into the proposed algorithm. To improve the sensitivity to fiber dispersion estimation errors, we utilize a two-stage dispersion estimator coupled with the G-S iteration. After 100-Gb/s PAM-4 signal transmissions over 400-km fibers, the simulation results show that the MCIA offers a 1.5-dB optical signal-to-noise ratio (OSNR) gain and a 1-dB optical power budget improvement compared with the decision-directed data-aided iterative algorithm (DD-DIA), for highly dispersive IM/DD transmissions. By performing adaptive dispersion estimation, the MCIA has higher tolerance to estimation errors in fiber length. Moreover, for cases subject to large dispersion, the usage of the embedded FFE pair not only desensitizes the MCIA on the limited bandwidth effect, but also accelerates the convergence performance for reaching lower BERs. We experimentally demonstrate that the proposed algorithm can support 150-Gb/s PAM-4 transmissions over 25-km standard single mode fibers (SSMF), where just a 7-tap FFE-pair is required. For 150 Gb/s transmissions, the tolerance to fiber length estimation error is increased from 0.9 km to 20 km.

A Decision Feedback Equalization Algorithm Based on Simplified Volterra Structure for PAM4 IM-DD Optical Communication Systems

A novel simplifying Volterra structure algorithm is proposed for an intensity modulation direct detection (IM-DD) optical fiber short distance communication system using the decision feedback equalization algorithm (DFE). Based on this algorithm, the signal damage for the four-level pulse amplitude modulation signal (PAM-4) is compensated, which is caused by device bandwidth limitation and dispersion during transmission. Experiments have been carried out using a 25 GHz Electro-absorption Modulated Laser (EML), showing that PAM-4 signals can transmit over 10 km in standard single-mode fiber (SSMF). The 112 Gbps and 128 Gbps signals can reach the error rate threshold of KP4-FEC (BER = 2 × 10−4) and HD-FEC (BER = 3.8 × 10−3), respectively. The simplified principle and process of the proposed Volterra-based equalization algorithm are presented. Experimental results show that the algorithm complexity is greatly reduced by 75%, which provides effective theoretical support for the commercial application of this algorithm.

Accelerating Transmitter Dispersion Eye Closure Quaternary Assessment by Deep Transfer Learning Technique

Transmitter dispersion eye closure quaternary (TDECQ) is one of the most important parameters that can systematically evaluate the signal quality of four-level pulse amplitude modulation format (PAM-4) transmitters. We experimentally demonstrate the acceleration of TDECQ assessment by deep transfer learning (DTL) technique. With the help of amplitude histograms (AHs) generated by the received PAM-4 signal, we are able to characterize the TDECQ value directly by the deep neural network (DNN). To further reduce both the computation time and the experimental data used for the training, simulation data can be used to provide the suitable source domain for the DTL implementation. The experimental results of 25 Gbaud PAM-4 signals indicate that, the mean absolute error of TDECQ assessment is below 0.18 dB over a TDECQ range from 1.9 to 3.6 dB. Meanwhile, 33.3&#x0025; amount of the used experimental data and 40&#x0025; epochs can be simultaneously reduced during the training process, due to the use of DTL.

A PAM4 transceiver design scheme with threshold adaptive and tap adaptive

To meet the demand of low bit error rate and high bandwidth for high-speed links, a reliable 112 Gb/s four-level pulse amplitude modulation (PAM4) transceiver design scheme with adaptive threshold voltage and adaptive decision feedback equalizer is proposed in this paper. In this scheme, three continuous time linear equalizers (CTLEs) at the front end of receiver are used to compensate the high-frequency, mid-frequency and low-frequency signals, respectively, and the variable gain amplifier (VGA) and saturation amplifier (SatAmp) are used to scale the signal amplitude. In addition to the three data samplers, four auxiliary samplers are also used for threshold adaptation. The sign-sign least mean squares algorithm uses the offset between the data sampler and the auxiliary sampler at the receiver side to drive the auxiliary reference voltage to converge to the signal constellation level, thus ensuring that the eye diagram of the PAM4 received signal has equal spacing and a constant signal–noise ratio for the three eyes in the vertical direction. In addition, the adaptive DFE for PAM4 signaling allows the transceiver to better adapt to the channel and thus achieve better equalizer performance. The simulation results show that the PAM4 transceiver design can compensate up to 25 dB of channel loss with an average eye height of 59.6 mv and an average eye width of 0.27 UI at a bit error rate of 10−12 under the condition of 3-tap feedforward equalizer (FFE) transmitter.

Plasmonic Ferroelectric Modulator Monolithically Integrated on SiN for 216 GBd Data Transmission

A high-speed plasmonic barium titanate (BTO, BaTiO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sub> ) Mach-Zehnder modulator is presented. We combine nanoscale plasmonics with BTO as solid-state active material and silicon nitride (SiN) for versatile and low loss waveguiding, and integrate them in a monolithic platform. We demonstrate a plasmonic BTO modulator processed onto foundry-produced SiN. The 15 µm long high-speed modulator features a flat electro-optic frequency response up to 70 GHz and is expected to be flat way beyond. A low V <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">π</sub> L product of 144 Vµm is shown. Data experiments reaching 216 Gbit/s with a 216 GBd 2PAM signal and 256 Gbit/s with a 128 GBd 4PAM signal are demonstrated. The merger of the versatile silicon nitride platform with high-speed plasmonics using the highly nonlinear ferroelectric BTO is an attractive solution as a future Tb/s optical interconnect platform.

Generation of high-speed PAM-4 signal with 3-bit DAC enabled by CRD-NS in optical interconnect.

In this paper, we experimentally demonstrated a 2-km high-speed optical interconnection with pulse-shaped pre-equalized four-level pulse amplitude modulation (PAM-4) signal generated by a 3-bit digital-to-analog converter (DAC) with the aid of in-band quantization noise suppression techniques under different oversampling ratios (OSRs) to reduce the influence of quantization noise. The simulation results show that the quantization noise suppression capability of high computational complexity digital resolution enhancer (DRE) is sensitive to taps number of the estimated channel and match filter (MF) response when OSR is sufficient, which will lead to further significant computational complexity increase. To better accommodate this issue, channel response-dependent noise shaping (CRD-NS) which also takes channel response into consideration when optimizing quantization noise distribution is proposed to suppress the in-band quantization noise instead of DRE. Experimental results show that about 2 dB receiver sensitivity improvement can be achieved at the hard-decision forward error correction (HD-FEC) threshold for 110 Gb/s pre-equalized PAM-4 signal generated by 3-bit DAC when the traditional NS technique is replaced by the CRD-NS technique. Compared to the high computational complexity DRE technique, in which channel response is also considered, negligible receiver sensitivity penalty is observed for 110 Gb/s PAM-4 signal, when the CRD-NS technique is utilized. Considering both the system cost and bit error ratio (BER) performance, the generation of high-speed PAM signal with 3-bit DAC enabled by the CRD-NS technique is regarded as a promising scheme for optical interconnection.

FPGA-Based Demonstration of WDM-PAM4-PON System With Low-Complexity CMA

We propose and experimentally demonstrate a low-complexity constant modulus algorithm (CMA) based on FPGA for 4-channel wavelength-division-multiplexing passive optical network (WDM-PON) employing 4-level pulse amplitude modulation (PAM4). The FPGA processes the 4 wavelength channels sequentially. In order to reduce the logic resources required by parallel equalization algorithm, an optimization based on parallel structure of constant mode algorithm (PCMA) was proposed in this paper. Different from the classic all-parallel structure CMA, our proposed PCMA adopts a new tap update structure, which only needs to update few sets of taps. With the help of the proposed structure, the resource of DSP is reduced to 21% on average. The PAM4 signal with a net rate of 110.24Gb/s transmission over 25-km SSMF can be achieved with the bit error rate (BER) less than 3.8×10 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">-3</sup> .

A switched-capacitor based track-and-hold amplifier suitable for PAM4 signaling in 45-nm CMOS

A switched-capacitor based track-and-hold amplifier suitable for PAM4 signaling in 45-nm CMOS

Burst errors suppression for MLSE in high-speed IM/DD transmission systems using PAM.

Maximum likelihood sequence estimation (MLSE) is the optimal signal sequence detection that can remove the inter-symbol interference (ISI). However, we find that the MLSE causes burst consecutive errors alternating between +2 and -2 in M-ary pulse amplitude modulation (PAM-M) IM/DD systems with large ISI. In this paper, we propose to use precoding to suppress the burst consecutive errors resulted from MLSE. A 2 M modulo operation is employed to guarantee that the probability distribution as well as the peak-to-average power ratio (PAPR) of encoded signal remain unchanged. After the receiver-side MLSE, the decoding process that involves adding the current MLSE output to the previous one and applying a 2 M modulo is implemented to break the burst consecutive errors. We conduct experiments to transmit 112/150-Gb/s PAM-4 or beyond 200-Gb/s PAM-8 signals at C-band to investigate the performance of the proposed MLSE integrated with precoding. The results show that the precoding can break burst errors effectively. For 201-Gb/s PAM-8 signal transmission, the precoding MLSE can achieve 1.4-dB receiver sensitivity gain and reduce the maximum length of burst consecutive errors from 16 to 3.

Performance evaluation of plastic optical fiber communication using micro-lens coupling and computational temporal ghost imaging algorithm.

Plastic optical fiber communication (POFC) systems are particularly sensitive to signal performance and power budget. In this paper, we propose what we belive to be a novel scheme to jointly enhance the bit-error-ratio (BER) performance and coupling efficiency for multi-level pulse amplitude modulation (PAM-M) based POFC systems. The computational temporal ghost imaging (CTGI) algorithm is developed for PAM4 modulation for the first time to resist the system distortion. The simulation results reveal that enhanced BER performance and clear eye diagrams are acquired by using CTGI algorithm with an optimized modulation basis. Experimental results also investigate and show, with CTGI algorithm, the BER performance for 180 Mb/s PAM4 signals is enhanced from 2.2 × 10-2 to 8.4 × 10-4 over 10 m POF by using a 40 MHz photodetector. The POF link is equipped with micro-lenses at its end faces by using a ball-burning technique, which helps to increase the coupling efficiency from 28.64% to 70.61%. Both simulation and experimental results show that the proposed scheme is feasible to achieve a cost-effective and high-speed POFC system with short reach.

C-band 100-GBaud PS-PAM-4 transmission over 50-km SSMF enabled by FIR-filter-based pre-electronic dispersion compensation.

Chromatic dispersion (CD) is always an obstacle to C-band high-speed intensity modulation and direct detection (IM/DD) transmissions, especially with a fiber reach of > 20 km. To reach beyond net-100-Gb/s IM/DD transmission over 50-km standard single mode fiber (SSMF), we for the first time present a CD-aware probabilistically shaped four-ary pulse amplitude modulation (PS-PAM-4) signal transmission scheme with a FIR-filter-based pre-electronic dispersion compensation (FIR-EDC) for C-band IM/DD transmission system. With the help of the FIR-EDC at the transmitter, 100-GBaud PS-PAM-4 signal transmission at 150-Gb/s line rate and 115.2-Gb/s net rate over 50-km SSMF is realized with only feed-forward equalization (FFE) at the receiver side. The superiority of the CD-aware PS-PAM-4 signal transmission scheme over other benchmark schemes has been successfully verified by experiments. Experimental results show that 24.5% improvement of system capacity is obtained by the FIR-EDC-based PS-PAM-4 signal transmission scheme in comparison to the FIR-EDC-based on-off keying (OOK) signal transmission scheme. Compared with the FIR-EDC-based uniform PAM-4 signal transmission scheme or the PS-PAM-4 signal transmission scheme without EDC, the capacity improvement obtained by the FIR-EDC-based PS-PAM-4 signal transmission scheme becomes more profound. The results show the potential and feasibility of such CD-aware PS-PAM-4 signal transmission scheme applied in CD-constrained IM/DD datacenter interconnects.

A low jitter 50 Gb/s PAM4 optical receiver in 130 nm SiGe BiCMOS

This paper analyzed the causes of phase jitter in four-level pulse amplitude modulation (PAM4) optical receiver (ORX), and a modified architecture was proposed. An optimized shunt-feedback transimpedance amplifier was employed to suppress the random jitter caused by the noise of analog front-end, a three-lane limiting amplifier with inter-stage feedback was introduced to extend the bandwidth, and retimers aligning the edges of thermometer codes with an external clock were also incorporated. The ORX fabricated in 130 nm SiGe BiCMOS technology occupies an area of 1.125 mm2 while dissipating 630 mW. The preliminary measured results demonstrated that the phase jitter of the most significant bit (MSB) was less than 0.08 UI for 25 Gb/s NRZ data. Post-layout simulation results indicate that our ORX can correctly decode 50 Gb/s PAM4 signal under a supply voltage of 3.3 V, and the peak-to-peak jitters for the MSB and the least significant bit (LSB) are only 1.5 ps and 2.9 ps, respectively.

Optical amplification-free deep reservoir computing-assisted high-baudrate short-reach communication.

An optical amplification-free deep reservoir computing (RC)-assisted high-baudrate intensity modulation direct detection (IM/DD) system is experimentally demonstrated using a 100G externally modulated laser operated in C-band. We transmit 112 Gbaud 4-level pulse amplitude modulation (PAM4) and 100 Gbaud 6-level PAM (PAM6) signals over a 200-m single-mode fiber (SMF) link without any optical amplification. The decision feedback equalizer (DFE), shallow RC, and deep RC are adopted in the IM/DD system to mitigate impairment and improve transmission performance. Both PAM transmissions over a 200-m SMF with bit error rate (BER) performance below 6.25% overhead hard-decision forward error correction (HD-FEC) threshold are achieved. In addition, the BER of the PAM4 signal is below the KP4-FEC limit after 200-m SMF transmission enabled by the RC schemes. Thanks to the use of a multiple-layer structure, the number of weights in deep RC has been reduced by approximately 50% compared with the shallow RC, whereas the performance is comparable. We believe that the optical amplification-free deep RC-assisted high-baudrate link has a promising application in intra-data center communications.

Transmission characteristics of all-optically converted QPSK and 16QAM signals

Various types of all-optical modulation format conversions have been proposed for realizing low latency and efficient connection of networks that have different modulation formats. All-optical PAM4 to QPSK and 16QAM modulation format conversions using an HNLF and NOLM have been demonstrated for connecting data center networks. However, there has been no study on the transmission characteristics of the converted signal over a long distance. In this study, we numerically investigate the spectral efficiency and transmission characteristics of the QPSK and 16QAM signals generated by the all-optical modulation format converters. The all-optically converted 50 Gbps class QPSK and 100 Gbps class 16QAM signals require a bandwidth of 55 GHz to maintain the EVM below 10%. We demonstrate that the converted QPSK and 16QAM signals can be transmitted up to 19,500 km and 4,000 km, respectively, in both single and 8 channels transmission by preserving the sufficient channel spacing of 62.5 GHz. Because the transmission distance of the typical inter-data-center network is shorter than 200 km, the converted signals are applicable to inter-datacenter networks. Moreover, the QPSK and 16QAM signals converted from PAM4 signals after 40 km transmission in an intra-data center network are valid for transmission in inter-data-center networks.

FPGA-based 4 × 29.4912 Gbit/s PS-PAM4 signal transmission with a low-complexity probabilistic shaping scheme.

In this experiment, we demonstrate a real-time intensity modulation and direct detection (IM/DD) system based on a field programmable gate array (FPGA). For high-speed parallel signal processing, we propose and implement the simplified parallel-constant modulus algorithm (CMA) and decision-directed least mean square (DDLMS) equalizers with low complexity and low latency. Moreover, the bit-class probabilistic shaping (PS) scheme is adopted with very few hardware resources. The digital signal processing (DSP) steps are implemented in the XCVU9P-FLGB2104-2-I Xilinx FPGA with a clock frequency of 230.4 MHz. Based on the experimental results, 4 × 29.4912 Gbit/s PS-pulse amplitude modulation (PAM4) signals can be successfully transmitted over 25 km of standard single-mode fiber (SSMF) while satisfying the hard-decision forward error correction (HD-FEC) threshold at 3.8 × 10-3. Compared with the uniformly distributed PAM4 signal, the low-complexity PS scheme can improve the receiver sensitivity by more than 1 dB.

Beyond 300 Gbps Short-Reach Links Using TFLN MZMs With 500 mVpp and Linear Equalization

Driven by the growing data demand and the stringent power constraints on short-reach datacenter interconnects (DCI), this work compares the driver-less (RF driver-free) transmission performance of thin-film lithium niobate (TFLN) Mach-Zehnder modulators (MZM). Two C-band MZMs (18 / 23 mm) fabricated in a commercial foundry are considered. Using the longer MZM with only linear equalization and ~500 mVpp driving signal, we transmit: (1) 132 Gbaud PAM6 signal (net 309 Gbps) below the 6.7% overhead (OH) hard-decision (HD)-FEC threshold; and (2) 136 Gbaud PAM8 signal (net 342 Gbps) above the 19.02% OH soft-decision (SD)-FEC normalized generalized mutual information (NGMI) threshold of 0.8798 over 500 m of standard single-mode fiber (SSMF). This work demonstrates the feasibility of operating beyond 300 G/lane using TFLN MZMs with CMOS-compatible driving levels.

Integrated sensing and communication in an optical fibre

The integration of high-speed optical communication and distributed sensing could bring intelligent functionalities to ubiquitous optical fibre networks, such as urban structure imaging, ocean seismic detection, and safety monitoring of underground embedded pipelines. This work demonstrates a scheme of integrated sensing and communication in an optical fibre (ISAC-OF) using the same wavelength channel for simultaneous data transmission and distributed vibration sensing. The scheme not only extends the intelligent functionality for optical fibre communication system, but also improves its transmission performance. A periodic linear frequency modulation (LFM) light is generated to act as the optical carrier and sensing probe in PAM4 signal transmission and phase-sensitive optical time-domain reflectometry (Φ-OTDR), respectively. After a 24.5 km fibre transmission, the forward PAM4 signal and the carrier-correspondence Rayleigh backscattering signal are detected and demodulated. Experimental results show that the integrated solution achieves better transmission performance (~1.3 dB improvement) and a larger launching power (7 dB enhancement) at a 56 Gbit/s bit rate compared to a conventional PAM4 signal transmission. Meanwhile, a 4 m spatial resolution, 4.32-nε/sqrt {Hz} strain resolution, and over 21 kHz frequency response for the vibration sensing are obtained. The proposed solution offers a new path to further explore the potential of existing or future fibre-optic networks by the convergence of data transmission and status sensing. In addition, such a scheme of using shared spectrum in communication and distributed optical fibre sensing may be used to measure non-linear parameters in coherent optical communications, offering possible benefits for data transmission.

Adaptive Bayesian neural networks nonlinear equalizer in a 300-Gbit/s PAM8 transmission for IM/DD OAM mode division multiplexing.

The strong stochastic nonlinear impairment induced by random mode coupling appears to be a long-standing performance-limiting problem in the orbital angular momentum (OAM) mode division multiplexing (MDM) of intensity modulation direct detection (IM/DD) transmission systems. In this Letter, we propose a Bayesian neural network (BNN) nonlinear equalizer for an OAM-MDM IM/DD transmission with three modes. Unlike conventional Volterra and convolutional neural network (CNN) equalizers with fixed weight coefficients, the weights and biases of the BNN nonlinear equalizer are regarded as probability distributions, which can accurately match the stochastic nonlinear model of the OAM-MDM. The BNN nonlinear equalizer is capable of adaptively updating its weights and biases sample-by-sample, according to the probability distribution. An experiment was conducted on a 300-Gbit/s PAM8 signal with three modes over a 2.6-km OAM-MDM RCF transmission. The experimental results demonstrate that the proposed BNN nonlinear equalizer exhibits promising solutions to effectively mitigate nonlinear distortions, which outperforms conventional Volterra and CNN equalizers with receiver sensitivity improvements of 1.0 dBm and 2.5 dBm, respectively, under hard-decision forward error correction (HD-FEC) thresholds. Moreover, compared with the Volterra and CNN equalizers, the complexity of the OAM-MDM is significantly improved through the BNN nonlinear equalizer. The proposed BNN nonlinear equalizer is a promising candidate for the high capacity inter-data center interconnects.