Neural-network-based end-to-end learning for adaptive optimization of two-dimensional signal generation in UVLC systems.

Visible light communication (VLC) benefits from the underwater blue-green window and holds immense potential for underwater wireless communication. In order to address the limitations of various equipment and harsh channel conditions in the underwater visible light communication (UVLC) system, the researchers proposed to use the method of autoencoder (AE) to tap the potential of the system. However, traditional AE schemes involve replacing the transmitting and receiving components of a communication system with a large multilayer perceptron (MLP) network, and they have significant drawbacks due to their reliance on a single network structure. In this paper, a novel 2D adaptive optimization autoencoder (2D-AOAE) framework is proposed to realize adaptive modulation and demodulation of two-dimensional signals. By implementing this scheme, we experimentally achieved a transmission rate of 2.85 Gbps over a 1.2-meter underwater VLC link. Compared to the traditional 32QAM UVLC system, the 2D-AOAE scheme demonstrated a 15.4% data rate increase. Moreover, the 2D-AOAE scheme exhibited a remarkable 73% improvement when compared to the UVLC system utilizing the traditional AE scheme. This significant enhancement highlights the superior performance and capabilities of the 2D-AOAE scheme in terms of transmission rate.

Multi-Gb/s visible light communication based on AlGaInP amber micro-LED.

Light-emitting diodes (LEDs), pivotal for solid-state illumination (SSL) and highly regarded as potential candidates in visible light communication (VLC) systems, have garnered significant interest as a solution to alleviate the congested radio frequency spectrum in next-generation communications. Addressing the challenge of extremely limited bandwidth due to the low response of phosphor in conventional illumination, our research focuses on an AlGaInP-based amber LED. This LED represents a promising avenue for phosphor-free, high-speed VLC applications when used in conjunction with the prevalent blue LED technology based on nitride materials. The fabricated AlGaInP amber LED, with a mesa diameter of 100 µm2, has undergone comprehensive optoelectronic property and transmission performance characterization. We have successfully demonstrated a proof-of-concept for VLC using the amber LED, achieving a data transmission rate of 2.94 Gb/s that complies with the forward-error-correction (FEC) standard of 3.8 × 10-3, utilizing adaptive bit and power loading with discrete multitone (BPL-DMT) modulation.

Feature decoupled knowledge distillation enabled lightweight image transmission through multimode fibers.

Multimode fibers (MMF) show tremendous potential in transmitting high-capacity spatial information. However, the quality of multimode transmission is quite sensitive to inherent scattering characteristics of MMF and almost inevitable external perturbations. Previous research has shown that deep learning may break through this limitation, while deep neural networks are intricately designed with huge computational complexity. In this study, we propose a novel feature decoupled knowledge distillation (KD) framework for lightweight image transmission through MMF. In this framework, the frequency-principle-inspired feature decoupled module significantly improves image transmission quality and the lightweight student model can reach the performance of the sophisticated teacher model through KD. This work represents the first effort, to the best of our knowledge, that successfully applies a KD-based framework for image transmission through scattering media. Experimental results demonstrate that even with up to 93.4% reduction in model computational complexity, we can still achieve averaged Structure Similarity Index Measure (SSIM) of 0.76, 0.85, and 0.90 in Fashion-MNIST, EMNIST, and MNIST images respectively, which are very close to the performance of cumbersome teacher models. This work dramatically reduces the complexity of high-fidelity image transmission through MMF and holds broad prospects for applications in resource-constrained environments and hardware implementations.

Modulation format recognition in a UVLC system based on an ultra-lightweight model with communication-informed knowledge distillation.

Modulation format recognition (MFR) is a key technology for adaptive optical systems, but it faces significant challenges in underwater visible light communication (UVLC) due to the complex channel environment. Recent advances in deep learning have enabled remarkable achievements in image recognition, owing to the powerful feature extraction of neural networks (NN). However, the high computational complexity of NN limits their practicality in UVLC systems. This paper proposes a communication-informed knowledge distillation (CIKD) method that achieves high-precision and low-latency MFR with an ultra-lightweight student model. The student model consists of only one linear dense layer under a communication-informed auxiliary system and is trained under the guidance of a high-complexity and high-precision teacher model. The MFR task involves eight modulation formats: PAM4, QPSK, 8QAM-CIR, 8QAM-DIA, 16QAM, 16APSK, 32QAM, and 32APSK. Experimental results show that the student model based on CIKD can achieve comparable accuracy to the teacher model. After knowledge transfer, the prediction accuracy of the student model can be increased by up to 87%. Besides, it is worth noting that CIKD's inference accuracy can reach up to 100%. Moreover, the parameters constituting the student model in CIKD correspond to merely 18% of the parameters found in the teacher model, which facilitates the hardware deployment and online data processing of MFR algorithms in UVLC systems.

Neural-network-based carrier-less amplitude phase modulated signal generation and end-to-end optimization for fiber-terahertz integrated communication system.

In fiber-terahertz integrated communication systems, nonlinear distortion and inter-symbol interference (ISI) will degrade transmission performance. Pre-compensation is an efficient method to handle the channel distortion as it can avoid noise boosting during channel compensation and reduce receiver side signal processing algorithmic complexity at user-end (UE) considering the asymmetric access scenario. In this paper, we propose and experimentally demonstrate a neural-network (NN)-based carrier-less amplitude phase (CAP) modulated signal generation and end-to-end optimization method for a fiber-terahertz integrated communication system. The CAP signal is generated directly from quadrature amplitude modulation symbols and pre-compensated through a transmitter NN, which allows the receiver to demodulate the signal with simple linear digital signal process (DSP). In generating the CAP signal, the NN based transmitter learns a group of filters, which can generate, up-convert, and pre-compensate the signals. Based on the proposed method, a fiber-terahertz integration access system at 220 GHz is demonstrated and a sensitivity gain of 1.2 dB is achieved at a transmission speed of 50 Gbps and the forward error correction (FEC) bit error rate (BER) threshold of 1 × 10-2 compared with the baseline after 10-km fiber transmission and 1-m wireless delivering.

113Gbps rainbow visible light laser communication system based on 10λ laser WDM emitting module in fiber-free space-fiber link.

With the advent of the sixth-generation mobile communication standard (6 G), the visible light communication (VLC) technology based on wavelength division multiplexing (WDM) technology can effectively solve the problem of shortage of spectrum resources and insufficient channel capacity. This paper introduces one of our technical achievements, namely the construction of a near-real-time visible light laser communication (VLLC) system based on WDM, which includes a self-designed 10-λ fully-packaged visible light laser emission module, 1 m multimode fiber - 0.175 m free space - 1 m multimode fiber optical transmission link, and receiver array. In the transmitter system, we adopt adaptive discrete multitone (DMT) modulation technique combined with Quadrature Amplitude Modulation (QAM) modulation scheme to obtain maximum spectral efficiency (SE). In the receiving system, we utilize the sparse-structured reservoir computing post-equalization algorithm to achieve superior equalization performance on the basis of the traditional post-equalization algorithm. The experimental results indicate that this quasi-real-time communication system has achieved a signal transmission rate of 113.175Gbps. To the best of our knowledge, this work has set a record in the field of high-speed visible light laser communication. Therefore, the laser communication system constructed by this work, with its flexibility in deployment and high-speed performance, demonstrates the significant potential application of visible light laser communication in data center interconnection and high-speed indoor access networks.

Net-432-Gb/s/λ PS-PAM transmission based on advanced DSP and single DAC for 400†G/λ IM/DD links.

The escalating surge in datacenter traffic creates a pressing demand for augmenting the capacity of cost-effective intensity modulation and direct detection (IM/DD) systems. In this Letter, we report the demonstration of the single-lane 128-GBaud probabilistically shaped (PS)-PAM-20 IM/DD transmission using only a single digital-to-analog converter (DAC) for a net 400 G/λ system. Based on the advanced digital signal processing (DSP), we achieve net bitrates of up to 437 Gb/s for optical back-to-back and 432 Gb/s after the 0.5-km SSMF transmission in the C-band with 128-Gbaud PS-PAM-20 signals. This work is the latest demonstration on ultra-high-order PS-PAM signals achieving net bitrates exceeding 400 Gb/s despite symbol rate limitations. Notably, to the best of our knowledge, the realized net information rate ([net bitrate]/[symbol rate]) of 3.37 marks a new achievement within the domain of 400 G/λ IM/DD systems, with promising implications for enhancing bandwidth efficiency in the upcoming 1.6-Tb Ethernet scenario.

Wide field-of-view laser-based white light transmitter for visible light communications.

The advancement demands of high-speed wireless data link ask for higher requirements on visible light communication (VLC), where wide coverage stands as a critical criterion. Here, we present the design and implementation of a transmitter structure capable of emitting a high-power wide-coverage white light laser. This laser source exhibits excellent stability, with an irradiation range extending to a half-angle of 20°. Its high brightness satisfies the needs of indoor illumination while maintaining excellent communication performance. Utilizing bit-loading discrete multi-tone modulation, a peak data transmission rate of 3.24 Gbps has been achieved, spanning 1 to 5 m. Remarkably, the data rates exceed 2.5 Gbps within a 40° range at a distance of 5 m, enabling a long-distance, wide coverage, high-speed VLC link for future mobile network applications.

Minimalist Deployment of Neural Network Equalizers in a Bandwidth-Limited Optical Wireless Communication System with Knowledge Distillation.

An equalizer based on a recurrent neural network (RNN), especially with a bidirectional gated recurrent unit (biGRU) structure, is a good choice to deal with nonlinear damage and inter-symbol interference (ISI) in optical communication systems because of its excellent performance in processing time series information. However, its recursive structure prevents the parallelization of the computation, resulting in a low equalization rate. In order to improve the speed without compromising the equalization performance, we propose a minimalist 1D convolutional neural network (CNN) equalizer, which is reconverted from a biGRU with knowledge distillation (KD). In this work, we applied KD to regression problems and explain how KD helps students learn from teachers in solving regression problems. In addition, we compared the biGRU, 1D-CNN after KD and 1D-CNN without KD in terms of Q-factor and equalization velocity. The experimental data showed that the Q-factor of the 1D-CNN increased by 1 dB after KD learning from the biGRU, and KD increased the RoP sensitivity of the 1D-CNN by 0.89 dB with the HD-FEC threshold of 1 × 10-3. At the same time, compared with the biGRU, the proposed 1D-CNN equalizer reduced the computational time consumption by 97% and the number of trainable parameters by 99.3%, with only a 0.5 dB Q-factor penalty. The results demonstrate that the proposed minimalist 1D-CNN equalizer holds significant promise for future practical deployments in optical wireless communication systems.

Laser-Based Mobile Visible Light Communication System.

Mobile visible light communication (VLC) is key for integrating lighting and communication applications in the 6G era, yet there exists a notable gap in experimental research on mobile VLC. In this study, we introduce a mobile VLC system and investigate the impact of mobility speed on communication performance. Leveraging a laser-based light transmitter with a wide coverage, we enable a light fidelity (LiFi) system with a mobile receiving end. The system is capable of supporting distances from 1 m to 4 m without a lens and could maintain a transmission rate of 500 Mbps. The transmission is stable at distances of 1 m and 2 m, but an increase in distance and speed introduces interference to the system, leading to a rise in the Bit Error Rate (BER). The mobile VLC experimental system provides a viable solution to the issue of mobile access in the integration of lighting and communication applications, establishing a solid practical foundation for future research.

Osteopontin mediation of disturbed flow-induced endothelial mesenchymal transition through CD44 is a novel mechanism of neointimal hyperplasia in arteriovenous fistulae for hemodialysis access.

In dysfunctional arteriovenous fistulae (AVF) for hemodialysis access, neointimal hyperplasia (NH) is prone to occur in the region exposed to disturbed flow. We hypothesized that disturbed flow contributes to NH in AVF by inducing endothelial mesenchymal transition (EndMT) through activation of the osteopontin/CD44 axis. In rats with aortocaval fistula, a rodent model of AVF, we demonstrated development of EndMT and expression of osteopontin and CD44 specifically in the vicinity of the arteriovenous junction using immunostaining. Duplex scan confirmed this region was exposed to a disturbed flow. A mixed ultrastructural phenotype of endothelium and smooth muscle cells was found in luminal endothelial cells of the arteriovenous junction by electron microscopy ascertaining the presence of EndMT. Endothelial lineage tracing using Cdh5-Cre/ERT2;ROSA26-tdTomato transgenic mice showed that EndMT was involved in NH of AVF since the early stage and that the endothelial-derived cells contributed to 24% of neointimal cells. In human umbilical vein endothelial cells (HUVECs) in culture, osteopontin treatment induced EndMT, which was suppressed by CD44 knockdown. Exposure to low oscillatory wall shear stress using a parallel-plate system induced EndMT in HUVECs, also suppressed by osteopontin or CD44 knockdown. In AVF of CD44 knockout mice, EndMT was mitigated and NH decreased by 35% compared to that in wild-type mice. In dysfunctional AVF of patients with uremia, expressions of osteopontin, CD44, and mesenchymal markers in endothelial cells overlying the neointima was also found by immunostaining. Thus, the osteopontin/CD44 axis regulates disturbed flow-induced EndMT, plays an important role in neointimal hyperplasia of AVF, and may act as a potential therapeutic target to prevent AVF dysfunction.

Physical prior inspired ensemble learning enables effective channel estimation of underwater visible light communication.

Deep neural networks have been applied to estimate the optical channel in communication systems. However, the underwater visible light channel is highly complex, making it challenging for a single network to accurately capture all its features. This paper presents a novel approach to underwater visible light channel estimation using a physical prior inspired network based on ensemble learning. A three-subnetwork architecture was developed to estimate the linear distortion from inter-symbol interference (ISI), quadratic distortion from signal-to-signal beat interference (SSBI), and higher-order distortion from the optoelectronic device. The superiority of the Ensemble estimator is demonstrated from both the time and frequency domains. In terms of mean square error performance, the Ensemble estimator outperforms the LMS estimator by 6.8 dB and the single network estimators by 15.4 dB. In terms of spectrum mismatch, the Ensemble estimator has the lowest average channel response error, which is 0.32 dB, compared to 0.81 dB for LMS estimator, 0.97 dB for the Linear estimator, and 0.76 dB for the ReLU estimator. Additionally, the Ensemble estimator was able to learn the V-shaped Vpp-BER curves of the channel, a task not achievable by single network estimators. Therefore, the proposed Ensemble estimator is a valuable tool for underwater visible light channel estimation, with potential applications in post-equalization, pre-equalization, and end-to-end communication.

Modulation format recognition in a UVLC system based on reservoir computing with coordinate transformation and folding algorithm.

Modulation format recognition (MFR) is one of the key technologies in adaptive optical systems and widely used in both commercial and civil applications. With the rapid development of deep learning, MFR algorithm based on neural networks (NN) has achieved impressive success. Due to the high complexity of underwater channels, to gain better performance of MFR tasks in underwater visible light communication (UVLC), the NN tend to be designed with a complex structure, which is costly in computation and hinders fast allocation and real-time processing. In this paper, we propose a lightweight and efficient method based on reservoir computing (RC), whose trainable parameters are only 0.3% of common NN-based methods. To improve the performance of RC in MFR tasks, we propose powerful feature extraction algorithms including coordinate transformation and folding algorithm. The proposed RC-based methods are implemented for six modulation formats, including OOK, 4QAM, 8QAM-DIA, 8QAM-CIR, 16APSK, and 16QAM. The experimental results show that our RC-based methods take only a few seconds for training process and under different pin voltages of LED, the accuracy for almost all exceeds 90%, and the highest is close to 100%. Analysis on how to design a well-performed RC to strike a balance between accuracy and time cost is also investigated, providing a useful guide for RC implementations in MFR.

Wavefront shaping for multi-user line-of-sight and non-line-of-sight visible light communication.

Visible light communication (VLC) has become a popular method for indoor communication, due to its high transmission speed and robustness against electromagnetic interference. Indoor VLC scenarios commonly consist of multiple users in line-of-sight (LOS) and non-line-of-sight (NLOS) paths. In NLOS, the light waves suffer from attenuation due to diffuse reflection from obstacles in the room, leading to significant attenuation in light intensity. This paper proposes a wavefront shaping method to enhance indoor VLC for multiple users, including both LOS and NLOS. By employing a spatial division scheme, we can simultaneously achieve a light intensity gain of 5.43 dB in NLOS through stepwise partitioning wavefront shaping and an opening angle range of 27° for two LOS users through computational holography. By employing bit-power-loading discrete multi-tone (DMT) modulation, we demonstrated VLC with transmission data rates of 3.082 Gbit/s and 3.052 Gbit/s for two LOS links and 2.235 Gbit/s for NLOS with 30.7% improvement compared with that without wavefront shaping, satisfying the 7% forward-error-correction (FEC) threshold.

Deep-learning-based multi-user framework for end-to-end fiber-MMW communications.

Fiber-wireless integration has been widely studied as a key technology to support radio access networks in sixth-generation wireless communication, empowered by artificial intelligence. In this study, we propose and demonstrate a deep-learning-based end-to-end (E2E) multi-user communication framework for a fiber-mmWave (MMW) integrated system, where artificial neural networks (ANN) are trained and optimized as transmitters, ANN-based channel models (ACM), and receivers. By connecting the computation graphs of multiple transmitters and receivers, we jointly optimize the transmission of multiple users in the E2E framework to support multi-user access in one fiber-MMW channel. To ensure that the framework matches the fiber-MMW channel, we employ a two-step transfer learning technique to train the ACM. In a 46.2 Gbit/s 10-km fiber-MMW transmission experiment, compared with the single-carrier QAM, the E2E framework achieves over 3.5 dB receiver sensitivity gain in the single-user case and 1.5 dB gain in the three-user case under the 7% hard-decision forward error correction threshold.

Spatial pilot-aided fast-adapted framework for stable image transmission over long multi-mode fiber.

Multi-mode fiber (MMF) has emerged as a promising platform for spatial information transmission attributed to its high capacity. However, the scattering characteristic and time-varying nature of MMF pose challenges for long-term stable transmission. In this study, we propose a spatial pilot-aided learning framework for MMF image transmission, which effectively addresses these challenges and maintains accurate performance in practical applications. By inserting a few reference image frames into the transmitting image sequence and leveraging a fast-adapt network training scheme, our framework adaptively accommodates to the physical channel variations and enables online model update for continuous transmission. Experimented on 100 m length unstable MMFs, we demonstrate transmission accuracy exceeding 92% over hours, with pilot frame overhead around 2%. Our fast-adapt learning scheme requires training of less than 2% of network parameters and reduces the computation time by 70% compared to conventional tuning approaches. Additionally, we propose two pilot-insertion strategies and elaborately compare their applicability to a wide range of scenarios including continuous transmission, burst transmission and transmission after fiber re-plugging. The proposed spatial pilot-aided fast-adapt framework opens up the possibility for MMF spatial transmission in practical complicated applications.

15.26Gb/s Si-substrate GaN high-speed visible light photodetector with super-lattice structure.

In this paper, we studied a series of high-speed photodetectors (PD) with different super-lattice interlayer periods and the scale of the effective area to examine their communication performance. The mini-PDs are designed with a single 1 mm × 1 mm effective area. The mini-PDs have three different super-lattice (SL) periods in the interlayer: 8, 15, and 32. The micro-PD sample has multiple 50um by 50um photosensitive areas that form a 4 × 4 receiver array, which shares a common N electrode. Its SL period is 26. The experiment shows that mini-PDs have the advantages such as better tolerance to beam spot deviation, larger field of view (FoV), higher responsibility, and wider peak width in spectral response. But micro-LED samples outperform the others in communication capacity and wavelength selectivity. The 8, 15, and 32 SL mini-PD samples achieve 6.6, 7.3, and 8.8 Gb/s data rates, respectively. The micro-PD gains the maximum data rate of 14.38Gb/s without applying waveform level post-equalization, and 15.26Gb/s after using an NN-based post-equalizer. This experiment shows that with proper DSP, GaN-based PD would be suitable for high-speed VLC systems, especially for the short wavelength spectrum in visible light.

11.2 Gbps 100-meter free-space visible light laser communication utilizing bidirectional reservoir computing equalizer.

In this paper, we introduce an innovative post-equalization technique leveraging bidirectional reservoir computing (BiRC), and apply it to waveform-to-symbol level equalization for visible light laser communication for the first time. This strategy is more resistant to nonlinearities compared to traditional equalizers like least mean square (LMS) equalizer, while requiring less training time and fewer parameters than neural network (NN) -based equalizers. Through this approach, we successfully conduct a 100-meter transmission of a 32-amplitude phase shift keying (32APSK) signal using a green laser operating at a wavelength of 520 nm. Remarkably, our system achieves a high data rate of 11.2 Gbps, all while maintaining a satisfying bit error rate (BER) below the 7% hard decision forward error correction (HD-FEC) threshold of 3.8E-3.

Inverse design of an ultra-compact dual-band wavelength demultiplexing power splitter with detailed analysis of hyperparameters.

Inverse design has been widely studied as an efficient method to reduce footprint and improve performance for integrated silicon photonic (SiP) devices. In this study, we have used inverse design to develop a series of ultra-compact dual-band wavelength demultiplexing power splitters (WDPSs) that can simultaneously perform both wavelength demultiplexing and 1:1 optical power splitting. These WDPSs could facilitate the potential coexistence of dual-band passive optical networks (PONs). The design is performed on a standard silicon-on-insulator (SOI) platform using, what we believe to be, a novel two-step direct binary search (TS-DBS) method and the impact of different hyperparameters related to the physical structure and the optimization algorithm is analyzed in detail. Our inverse-designed WDPS with a minimum feature size of 130 nm achieves a 12.77-times reduction in footprint and a slight increase in performance compared with the forward-designed WDPS. We utilize the optimal combination of hyperparameters to design another WDPS with a minimum feature size reduced to 65 nm, which achieves ultra-low insertion losses of 0.36 dB and 0.37 dB and crosstalk values of -19.91 dB and -17.02 dB at wavelength channels of 1310 nm and 1550 nm, respectively. To the best of our knowledge, the hyperparameters of optimization-based inverse design are systematically discussed for the first time. Our work demonstrates that appropriate setting of hyperparameters greatly improves device performance, throwing light on the manipulation of hyperparameters for future inverse design.

One-bit quantization delta-sigma modulation-based autoencoder for power-efficient free-space communication.

This Letter proposes a novel, to the best of our knowledge, approach utilizing a delta-sigma modulation (DSM)-based 1-bit autoencoder (AE) for efficient encoding and decoding in various channel conditions. Simulation analysis demonstrates the AE's ability to mitigate noise by reducing a peak-to-average power ratio (PAPR) and enhancing an in-band power of the signals, particularly under low signal-to-noise ratios (SNRs). The AE-DSM achieves theoretical transmission performance even at SNRs below 6 dB. In a 40-m free-space link experiment, the AE-DSM exhibits an 8.4-dB lower bit error rate (BER) compared to 64QAM-DSM, enabling a transmission rate of 1.31 Gbps. Furthermore, the 1-bit AE-DSM significantly reduces power consumption in the receiving analog-to-digital converter (ADC), facilitates transmission at low SNRs, and effectively mitigates nonlinear effects. Consequently, the DSM-based AE holds immense potential for future mobile fronthaul links.