Optical TTD compensation network-based phase precoding for THz massive MIMO systems.

In sixth generation (6G) communications, terahertz (THz) communication is one of the most important technologies in the future due to its ultra-bandwidth, where hybrid beamforming has been widely used to solve the severe transmission attenuation in the THz band. However, the use of frequency-flat phase shifters in hybrid beamforming leads to the beam split effect. To solve the beam split influence, we propose a novel optical true time delay compensation network (OTTDCN)-based phase precoding structure with low power consumption. In the proposed scheme, the OTTDCN pre-generates multiple beam compensation modes to achieve phase compensation for different frequencies. As a result, the compensated beams can be reoriented toward the target direction at different frequencies. Moreover, a low-complexity beam compensation mode-based hybrid precoding algorithm is proposed, where the selection of the optimal beam compensation modes used for all radio-frequency (RF) chains with finite beam compensation modes is considered. The results show that the OTTDCN-based phase precoding scheme can effectively alleviate the beam split effect with low power consumption and achieve near-optimal performance.

Transverse mode switchable mode-locked laser with narrow bandwidth.

Transverse mode switchable ultrashort optical pulses with narrow bandwidths can create potential for exploring what we believe are new physical effects. We demonstrate the generation of transverse mode switchable ultrashort pulses with narrow bandwidths in an all-fiber mode-locked laser by exploring a mode-selective photonic lantern (MSPL). The laser cavity serves not only as a ring resonator but also as an intrinsic spectral filter. For mode-locking with the LP01, LP11a, and LP11b modes, the bandwidths are 3.0 nm, 86.7 pm and 101.7 pm, respectively. The narrowband pulses with higher-order modes are generated by an intrinsic spectral filter due to the spectral-domain intermodal interference. Mode-locked pulses with a signal-to-noise ratio better than 60 dB for LP01, LP11a, and LP11b modes are independently generated, i.e., transverse mode switchable by changing the input port of the MSPL. The mode-locked wavelength can be tuned for the LP11a mode and LP11b mode by adjusting the state of polarization. Furthermore, our experimental results also show that, the slope efficiency of LP11a and LP11b modes can be improved, by the use of LP11 mode pump scheme. We anticipate that, narrowband pulses with complex mode profiles can be generated by simultaneously phase-locked transverse and longitudinal modes.

On the multipath interference mitigation for high-speed IM-DD transmissions with Nyquist subcarrier modulation.

The performance of high-speed intensity modulation direct detection (IM-DD) transmissions is severely degraded due to the occurrence of multipath interference (MPI), especially when a higher-order modulation format is utilized. Here, we propose and demonstrate, for the first time to the best of our knowledge, that a Nyquist subcarrier modulation (Nyquist-SCM) format inherently exhibits resistance to the MPI. We experimentally evaluate the MPI tolerance by transmitting 56 Gbit/s PAM-4 signals and Nyquist-SCM 16QAM signals over the 2 km standard single-mode fiber (SSMF) when the C-band semiconductor laser with a linewidth of 1.7 MHz is utilized. In comparison with the PAM-4 format, the Nyquist-SCM 16QAM format can lead to an enhanced MPI tolerance of 4 dB at the KP4-FEC threshold of BER = 2 × 10-4. In addition, even with the help of MPI mitigation for the PAM-4 signals based on two newly reported methods, the utilization of Nyquist-SCM 16QAM signal can still guarantee an improved MPI tolerance of 1 dB.

Single wavelength MDM-PDM 800-Gb/s net data rate transmission over a standard multimode fiber employing a Kramers-Kronig receiver.

We propose a high-speed multimode fiber short-reach optical interconnect system based on a Kramers-Kronig (KK) field reconstruction with the mode division multiplexing (MDM) and polarization division multiplexing (PDM) technology. In this work, the LP01, LP21a, LP21b, and LP02 modes are selected as independent channels to carry information. The demonstration achieved the 800 Gb/s net data rate per wavelength with a bit-rate-distance-product (BDP) of 8 Tb/s·km. To the best of our knowledge, this is the highest experimental record of a single wavelength BDP over the SMMF with KK detection. In addition, we discuss the system performance after all multiple-input multiple-output (MIMO) and partial MIMO processing and give guidance on the trade-off between system performance and computational resource.

Low-loss and compact photonic lantern based on a step-index double cladding fiber.

The fulfilment of the adiabatic criterion is indispensable for the realization of a low-loss photonic lantern (PL), concurrently imposing a stringent restriction on the taper transition length of the PL. Here, by relaxing the adiabatic criterion, a low-loss and compact PL based on a step-index double cladding fiber (SI-DCF) is theoretically proposed and experimentally demonstrated. The use of SI-DCF can reduce the mode field diameter (MFD) expansion ratio during the tapering processing and greatly decrease the taper transition length required for adiabatic tapering. We initially evaluate the variation of both MFD and effective refractive index (RI) along the fiber tapering based on three types of fiber structures, including the modified standard single-mode fiber (SSMF), the graded-index fiber (GIF), and the proposed SI-DCF. In comparison with the commonly used fiber geometry, the SI-DCF can reduce the MFD expansion ratio from 77.73% to 38.81%, leading to more than half reduction of the tapering length for both 3-mode and 6-mode PLs. Then, two kinds of SI-DCF with different core diameters are fabricated to realize a 3-mode PL. The fabricated PL possesses a 1.5 cm tapering length and less than 0.2 dB insertion loss (IL). After splicing with the commercial few-mode fiber, the PL has an average IL of 0.6 dB and more than 13 dB LP11 mode purity over the C-band. Finally, a transfer matrix measurement indicates that the fabricated PLs have a mode coupling of less than -10 dB at 1550 nm.

Blind frequency offset estimation method based on a minimum phase correction error for a full spectrum modulated NFDM system.

To improve the spectral efficiency of a full spectrum modulated nonlinear frequency division multiplexing (FS-NFDM) system, a blind frequency offset estimation (FOE) method has been proposed. The approach based on the minimum phase correction error can achieve high estimation accuracy of sub-MHz without need of any training symbols. Furthermore, in order to reduce the computational complexity, an eigenvalue-shift method is used to get a coarse search interval of FO, and then the one-dimensional optimization algorithm based on golden section search and parabolic interpolation is used to get the optimal FOE for the coarse search interval. The feasibility and reliability of the proposed blind FOE approach have been demonstrated in both BTB and fiber transmission scenarios. Compared with the grid search method, the proposed solving scheme can save hundreds of times of the searches. The experimental results reveal that the proposed method is robust to the amplified spontaneous emission noise and phase noise and has the capabilities of a wide FOE range and a high FOE accuracy.

Experimental demonstration of optical waveform transmission with multiple-eigenvalue 16-APSK modulation.

The multi-eigenvalue multiplexing-based discrete spectrum-modulated nonlinear frequency-division multiplexing (DS-NFDM) system with higher-order modulation format has been demonstrated experimentally. After designing the coefficients of the eigenvalue set and the constellation point distribution of 16-amplitude phase shift keying (16-APSK), the realizations of 14-, 30-, and 46-eigenvalue multiplexed DS-NFDM signals have been implemented. The results show that 46-eigenvalue and 30-eigenvalue multiplexed DS-NFDM signals can transmit 50 km and 400 km over a nonzero dispersion-shifted fiber (NZDSF) under soft-decision forward error correction (SD-FEC) threshold of 2.4E-2, respectively. This demonstration shows for the first time, to the best of our knowledge, the record for multiplexed eigenvalue number and data rate of the multiple-eigenvalue-based DS-NFDM system.

Bias-drift insensitive full-field frequency response characterization of a thin-film lithium niobate-based intensity modulator.

We propose a rapid and precise scheme for characterizing the full-field frequency response of a thin-film lithium niobate-based intensity modulator (TFLN-IM) via a specially designed multi-tone microwave signal. Our proposed scheme remains insensitive to the bias-drift of IM. Experimental verification is implemented with a self-packaged TFLN-IM with a 3 dB bandwidth of 30 GHz. In comparison with the vector network analyzer (VNA) characterization results, the deviation values of the amplitude-frequency response (AFR) and phase-frequency response (PFR) within the 50 GHz bandwidth are below 0.3 dB and 0.15 rad, respectively. When the bias is drifted within 90% of the Vπ range, the deviation fluctuation values of AFR and PFR are less than 0.3 dB and 0.05 rad, respectively. With the help of the full-field response results, we can pre-compensate the TFLN-IM for the 64 Gbaud PAM-4 signals under the back-to-back (B2B) transmission, achieving a received optical power (ROP) gain of 2.3 dB. The versatility of our proposed full-field response characterization scheme can extend to various optical transceivers, offering the advantage of low cost, robust operation, and flexible implementation.

Pairwise Tomlinson-Harashima precoding for multiple-lane IM-DD transmissions.

As for the photonic interconnection based on the multiple-lane intensity modulation direct detection (IM-DD) transmission, both intra-channel inter-symbol-interference (ISI) originating from bandwidth constraint, and inter-channel performance discrepancy emerging from inter-channel component differences are the major bottleneck for the throughput enhancement. Here, we propose a pairwise Tomlinson-Harshima precoding (P-THP) scheme, in order to simultaneously deal with both intra-channel ISI and inter-channel performance discrepancy. The effective function of the proposed P-THP scheme is experimentally evaluated by transmitting 4-channel 81-GBaud PAM4 signals over 2 km standard single-mode fiber (SSMF). Compared with the conventional scheme with only applying THP on individual wavelength channel, the required optical received power (ROP) under the back-to-back (B2B) transmission can be reduced by 0.75∼1 dB with the help of proposed P-THP in different experimental component configurations, at the 7% hard decision forward error correction (HD-FEC) threshold of BER = 3.8 × 10-3. After the 2 km SSMF transmission, only the use of proposed P-THP can guarantee to reach the designated HD-FEC threshold, leading to a net rate of >600 Gbit/s.

Simultaneous realization of high gain and low DMG of four-mode EDFA under bidirectional hybrid-mode pump.

We theoretically and experimentally verify that, the bidirectional hybrid-mode pumping scheme can address the optimization problem of trade-off between high gain and low differential modal gain (DMG) of four-mode erbium-doped fiber amplifier (4M-EDFA), in comparison with traditional both forward and backward hybrid-mode pumping scheme. It is noticed that, when the total pump power is fixed, the bidirectional hybrid-mode pumping scheme can not only achieve higher gain, but also suppress DMG due to different overlap integrals for the forward and backward pumping schemes. The bidirectional hybrid-mode pumped 4M-EDFA is developed with the forward pumping at LP02 mode and the backward pumping at LP21 mode, under a pump power ratio of 30%:70%. Thus, we can achieve an average gain of up to 21.16 dB and a low DMG of 0.43 dB at 1550 nm, and an average gain of up to 20.64 dB with a DMG of less than 1.6 dB over the C-band. In particular, the bidirectional hybrid-mode pumping scheme allows us to tailor the gain characteristics of the few-mode erbium-doped fiber amplifiers (FM-EDFAs), by adjusting the power ratio between forward and backward pumps.

Non-sweep DC component estimation method for a virtual-carrier assisted Kramers-Kronig receiver.

The Kramers-Kronig (KK) receiver has attracted much attention in short-range optical interconnection because of its ability to recover the phase of the signal from the intensity information through KK algorithm. In high-speed KK systems, such as virtual-carrier (VC) assisted ones, an alternating current (AC) coupled photo-detector (PD) is preferred due to relaxing the requirements of analog-to-digital converter (ADC) and electronic amplifier by filtering direct current (DC) component. However, the loss of the DC component will cause the KK algorithm to break down, so it is necessary to accurately recover DC value in the digital domain with multiple-sweep. In this paper, we propose what we believe is a novel non-sweep DC component estimation scheme based on optimized digital carrier-to-signal power ratio (OD-CSPR) method, which can accurately estimate the DC component with only 3-4 iterations in the scenario of VC-assisted KK receiver optical transmission. The scheme utilizes the one-dimensional search optimization algorithm based on golden section search and parabolic interpolation without sweeping. The simulation and experimental results of the proposed non-sweep OD-CSPR method show that the DC component can be estimated accurately in a large CSPR range, and the system performance is close to that of the conventional DC-sweep method. Compared with the typical defined digital CSPR (DD-CSPR) based optimization method, the proposed one can realize optical signal-to-noise ratio (OSNR) gains of 0.9 dB in the back-to-back (B2B) and 0.7 dB under 80 km fiber transmission scenarios respectively with a total bit rate of 160Gb/s.

High-speed secure key distribution based on interference spectrum-shift keying with signal mutual modulation in commonly driven chaos synchronization.

The secure key generation and distribution (SKGD) are unprecedentedly important for a modern secure communication system. This paper proposes what we believe to be a novel scheme of high-speed key distribution based on interference spectrum-shift keying with signal mutual modulation in commonly driven chaos synchronization. In this scheme, delay line interferometers (DLI) are utilized to generate two low-correlation interference spectra from commonly driven synchronous chaos, and then a 2 × 2 optical switch can effectively change the relationship between the two interference spectra in post-processing by shifting the states of the switch. The signals then undergo electro-optic nonlinear transformation through a hardware module, which includes a signal mutually modulating module (SMMM) and a dispersion component. This optimization significantly enhances the entropy source rate of synchronized chaos from both legitimate users. Moreover, thanks to the introduction of DLIs and electro-optic nonlinear transformation module, the key space of the proposed scheme is remarkably improved. In comparison to traditional chaotic drive-response architectures, the scheme effectively suppresses residual correlation. A 6.7 Gbit/s key distribution rate with a bit error rate below 3.8 × 10-3 is experimentally demonstrated over a 40 km single-mode fiber (SMF).

Transmitter dispersion eye closure quaternary assessment based on linear CNN with 1 × 1 convolutional kernel.

Transmitter dispersion eye closure quaternary (TDECQ) is a vital metric to characterize the quality of four-level pulse amplitude modulation (PAM-4) optical signals. However, the traditional TDECQ assessment scheme is complex and time consuming, with heavy iterative operations. Therefore, accelerating the TDECQ assessment has great significance for photonic data-center interconnection (DCI) applications. Here, we propose and experimentally demonstrate a TDECQ assessment based on linear-convolutional neural network (L-CNN) with the 1 × 1 convolutional kernel to reduce the implementation complexity. Our experimental results verify that the lightweight L-CNN can realize the accurate TDECQ assessment, without the involvement of nonlinear activation functions (NAFs). The mean absolute error (MAE) of 26.5625 and 53.125 GBaud PAM-4 signals are 0.16 dB and 0.18 dB, respectively, over a TDECQ range from 1.5 to 4.0 dB. Meanwhile, in comparison with existing CNN-based schemes, the L-CNN based TDECQ assessment scheme only needs 2048 multiplications, which have been reduced by five orders of magnitude.

Frequency offset estimation for nonlinear frequency division multiplexing with continuous spectrum modulation.

Carrier frequency offset (CFO) estimation is very important for the optical fiber communications and has been studied widely in linear coherent systems, while only a few works have been reported for nonlinear Fourier transform (NFT) based systems. In continuous spectrum (CS) modulation nonlinear frequency division multiplexing (CS-NFDM) systems, frequency offset (FO) has a great influence on its performance, requiring an improved frequency offset estimation (FOE) method. We found that the oversampling rate R0 adopted in NFDM to ensure the accuracy of the NFT and inverse NFT (INFT) calculations, would cause the estimation accuracy of the traditional FFT-FOE method to decrease by R0 times. Moreover, CS-NFDM signals with higher baud rate require more subcarriers and then result in an oversampling factor greater than 16. This makes the traditional FFT-FOE method be ineffective to use the common training sequence (TS) overhead to meet the FOE error requirement of CS-NFDM system. Therefore, a modified FOE method based on FFT assisted by TS and autocorrelation has been proposed. The theoretical analysis and simulation results show that the proposed method is applicable to CS-NFDM system, no matter what modulation format is used. For 512 subcarriers, with a high rate of 70GBaud and the TS length of 8192, the proposed method can obtain a minimum FO estimation error about 0.1 MHz, which is better than the other two typical FFT-FOE and Schmidl & Cox methods. In addition, the proposed method can save at least 87.5% and 50% overhead. Thus, the proposed method has obvious improvement for CS-NFDM system with requiring high oversampling rate.

Low latency microwave photonic RTFT processing based on bandwidth slicing and equivalent dispersion.

Microwave photonic real-time Fourier transformation (RTFT) processing based on optical dispersion is a promising solution for microwave spectrum analysis. However, it usually brings the drawbacks of limited frequency resolution and large processing latency. Here, we demonstrate a low-latency microwave photonic RTFT processing based on bandwidth slicing and equivalent dispersion. The input RF signal is first divided into different channels with the help of bandwidth slicing technique, and then finely analyzed by the fiber-loop based frequency-to-time mapping. In the proof-of-concept experiment, a 0.44-m fiber-loop offers an equivalent dispersion as high as 6 × 105 ps/nm with a small transmission latency of 50 ns. As a result, we can realize a wide instantaneous bandwidth of 1.35 GHz, a high frequency resolution of approximately 20 MHz, and a high acquisition frame rate of approximately 450 MHz, together with a total latency of less than 200 ns.

2.4-Tb/s/λ*km fundamental mode transmission over OM2 fiber employing Kramers-Kronig receiver.

We propose a digital-carrier Kramers-Kronig (DC-KK) scheme based high-speed multimode fiber short-reach optical interconnect system with fundamental mode transmission. After optimization of the parameters, including the roll-off factor of the root-raised-cosine (RRC) filter, and the guard interval (GI) between signal and carrier tone, as well as the carrier signal power ratio (CSPR), 200-Gb/s 32-quadrature amplitude modulation (32QAM) signal transmission over 12-km OM2 fiber has been experimentally demonstrated with a bit error ratio (BER) below the soft-decision forward error correction (SD-FEC) threshold of 4 × 10-2. To the best of our knowledge, this is the highest experimental record of single lambda bitrate-distance-product (SLBDP) achieved by direct-detection (DD)-based transmission over a standard multimode fiber (MMF). The proposed scheme has potential to improve the system performance without replacing massive deployed legacy MMFs for future large-capacity data center interconnects (DCIs).

Characterization of differential modal group delay in few-mode fiber using a digital re-sampling technique.

We characterize the differential modal group delay (DMGD) arising in few-mode fibers (FMFs) based on the digital re-sampling technique, which is commonly used in current digital signal processing flow at the receiver-side. When the DMGD of a 291-m two-mode fiber is characterized over the C-band by using a 500-Mb/s non-return-to-zero (NRZ) signal and 1-GSa/s real-time oscilloscope, the experimental results are consistent with the DMGD obtained from the traditional time-of-flight (TOF) method. However, the wide-bandwidth instruments of the traditional TOF method can be replaced by cheap ones with a bandwidth of only a few hundred MHz, but the same temporal precision is achieved. Moreover, our proposed DMGD characterization method is not limited by the number of guided modes arising in the FMF, together with the capability to obtain both the DMGD value and its sign between two arbitrary guided modes.

C-band 100-GBaud/λ PAM-4 transmission enabled by hardware-efficient nonlinear equalizer with weight-sharing pruning.

A Volterra nonlinear equalizer (VNLE) can effectively mitigate both linear and nonlinear impairments arising in intensity-modulation direct-detection (IM-DD) transmission systems. However, the high computational complexity of the VNLE hinders its applications. Here, we propose a hardware-efficient Volterra equalizer with weight-sharing pruning (VE-WSP). Such an equalizer first uses the k-means++ clustering algorithm for the weight sharing within the same cluster, and then ranks the cluster centroid weight for pruning, leading to a significant computational complexity reduction without sacrificing any equalization performance. We experimentally verified that the use of VE-WSP can enable the C-band 100-GBaud/λ PAM-4 transmission over a 1-km standard single-mode fiber (SSMF), to reach the 20% soft-decision forward error correction (SD-FEC) threshold. Meanwhile, the proposed VE-WSP can reduce the computational complexity by 51% and 21%, in terms of multiplication and addition, respectively, in comparison with the VNLE.

In-line mode-dependent loss equalizer with femtosecond laser induced refractive index modification.

We demonstrate an in-line all-fiber mode-dependent loss (MDL) equalizer with femtosecond laser induced refractive index (RI) modification. By inscribing an RI-modified structure into the core of a few-mode fiber (FMF), a differential mode attenuation (DMA) can be achieved for LP01 and LP11 modes. The DMA can serve as an in-line MDL equalizer for the long-haul mode-division multiplexing transmission system. Through numerical simulations, we identify that the LP01 mode has a larger attenuation than that of higher-order modes, where the sign of DMA is contrary to that of the conventional FMF links and devices. Finally, a proof-of-concept experiment is implemented by inscribing an RI modified region with a width of 4 µm, a height of 13 µm, and a length of 200 µm into the FMF core. An average additional attenuation of 8.4 dB and 3 dB can be applied to the LP01 and LP11 modes over the C-band, respectively, leading to an MDL equalization range of 5.4 dB. Meanwhile, the average polarization dependent loss (PDL) of the LP01 and LP11 modes induced by the in-line MDL equalizer is approximately 0.3 dB over the C-band. Power matrix measurement indicates that the in-line MDL equalizer has a negligible mode coupling. The proposed in-line MDL equalizer with a wider range and low insertion loss is feasible by precise manipulation of femtosecond laser inscription.

Towards high spectral efficiency UDWDM-PON based on the simplified coherent reception of pulsed shaped PAM-4 signals.

Enhancing spectral efficiency (SE) of ultra-dense wavelength division multiplexing passive optical network (UDWDM-PON) is vital to providing broadband access for massive users. Here, we experimentally demonstrate a high SE UDWDM-PON in the C-band, based on the simplified coherent reception of 10 Gb/s 4-level pulse-amplitude modulation (PAM-4) signals. We investigate the WDM signal reception by mathematical derivation and propose to enhance the SE by adopting both intradyne detection and pulse shaping techniques. Then, both approaches are numerically evaluated, with an identification that there occurs a trade-off between SE and power budget improvements. Finally, we experimentally achieve a SE of 0.83 (bit/s)/Hz and a power budget of 25 dB for a proof-of-concept 3 × 10 Gb/s PAM-4 downstream transmission over 20 km standard single mode fiber (SSMF).