Scientific Rrsearch Direction-Shanghai Jiaotong University

Intelligent Optical Fiber Communication System

In 2021, Huawei released the "Ten Mathematical Challenges for the Information Industry in the Post-Shannon Era". Among them, the challenge involving optical fiber communication is "Optical Communication Nonlinear Channel Compensation Problem", which is specifically broken down into fast and accurate modeling of optical fiber channels and Nonlinear Compensation Problem. Realizing low-complexity nonlinear accurate compensation through fast and accurate modeling of fiber channels, it is expected to approach the Shannon capacity limit of linear channels, break through the "capacity crisis" of fiber channels, and support the continuous growth demand of communication capacity. LIFE conducts continuous in-depth research on this key issue, introduces AI technology, and designs AI architecture and algorithms suitable for optical fiber communication systems, aiming to break through the key challenge of optical fiber nonlinear compensation.

Intelligent Modeling and Simulation Platform for Optical Fiber Communication System

Both the modeling of the fiber channel and the design of the nonlinear compensation algorithm rely on the fiber, the communication system simulation tool, the academia and the industry widely use matlab as the simulation tool, in addition to the controlled risk, there are also low operating efficiency and long calculation time , Difficult to be compatible with AI algorithms and other issues. In order to solve the above problems, LIFE has developed a complete fiber optic communication system simulation tool based on the open source Python language, including the shaping modulation pre-equalization algorithm at the transmitting end, the demodulation equalization compensation algorithm at the receiving end, fiber channel modeling, and signal visualization analysis. Due to the natural AI compatibility of the Python language, all algorithms at the transceiver end and fiber channel modeling can be implemented based on AI. LIFE researches AI architecture and algorithms suitable for fiber channel modeling to achieve fast and accurate modeling of multi-channel high-speed long-distance fiber channels. On this basis, the global optimization of algorithms at the transceiver end can be realized.

The optical fiber communication system intelligent modeling and simulation platform developed by LIFE is independent, internationally leading, and integrated and open. While supporting its own research, it plans to open source all codes and serve the entire optical communication industry. Based on this open platform, we will jointly carry out research on the innovative architecture and algorithms of optical communication systems with the industry, and jointly break through the core challenge of optical fiber nonlinear channel compensation.

Nonlinear Compensation and Global Optimization of Optical Fiber Communication System

On the basis of realizing fast and accurate modeling of optical fiber channel, the construction of intelligent simulation and experiment platform for optical fiber transmission is completed. The intelligent optical fiber transmission experiment platform adopts the same algorithm library as the simulation platform, and the simulation and experiment platform realize data alignment and cross-validation. All experimental equipment can be controlled remotely, and the end-to-end nonlinear compensation and global optimization experiments of the optical fiber communication system can be performed by remotely calling the algorithm library. The AI algorithm library includes toolkits such as various general neural networks, reinforcement learning, mutual information estimation network (MINE), fiber nonlinear estimation network (FINE), end-to-end learning, and training strategies. Equalization, nonlinear estimation and compensation at the receiving end, and nonlinear performance optimization combined with the global optimization of the receiving end and the channel, in order to minimize the impact of fiber nonlinearity and achieve the goal of approaching the Shannon capacity limit of the linear channel.

High-speed direct inspection optical access system

The optical access network is close to the user side, and the cost is limited. It is necessary to use low-bandwidth optoelectronic devices combined with low-cost direct adjustment and direct detection technology to achieve high-speed and high-power budget optical access. Therefore, high-performance optical channel equalization technology is particularly critical. LIFE has been adhering to the technical route of high-speed direct detection optical access channel equalization for the past ten years, and proposed methods such as optical time domain equalization, optical frequency domain equalization, photoelectric hybrid equalization, neural network equalization, etc., to achieve a single-wavelength access rate from 10Gb/s to 10Gb/s Continuous improvement to 100Gb/s. By comprehensively comparing the performance and complexity of different equalization algorithms under the same standard system, the equalization algorithm most suitable for high-speed direct detection optical access systems is located. Further use reinforcement learning to automatically optimize the equalizer structure, reduce complexity under the premise of ensuring performance, and finally deploy it in FPGA in real time to achieve low power consumption, high performance, high-speed direct detection optical access.

High-speed coherent optical access system

The point-to-multipoint (P2MP) optical access system architecture based on coherent detection is an important research direction of optical access networks in recent years, and it is expected to break through the high-speed burst detection of uplink signals in traditional time-division multiplexing optical access networks (TDM-PON) problem. The central office uses high-speed coherent detection technology in the field of optical transmission to aggregate and receive uplink signals from different users, and the user side uses low-cost coherent detection technology to select and receive its own channel. Wavelength locking on the user side, low-voltage phase modulation, and low-cost polarization-independent coherent detection are the main challenges faced by this scheme.

Smart Fiber Laser System

Fiber lasers are divided into pulsed fiber lasers and continuous fiber lasers. LIFE mainly conducts research on the scenarios of femtosecond ultrafast fiber lasers and ultra-high power continuous fiber lasers. Femtosecond ultrafast fiber lasers mainly use fiber nonlinearity to achieve ultrashort and ultrastable pulses The ultra-high power continuous fiber laser mainly suppresses the nonlinearity of the fiber to achieve ultra-high power output. LIFE takes the lead in introducing intelligent control technology into the field of fiber laser, embedding intelligent algorithms into hardware platforms to realize real-time intelligent control, and achieving the above two main goals.

Femtosecond-scale pulses correspond to many ultrafast processes in rich material systems such as atoms, molecules, materials, biological proteins, and chemical reactions, and have important applications in the field of scientific research; on the other hand, femtosecond pulses are The peak power is widely used in the field of material fine processing. Ultrafast fiber lasers need to comprehensively optimize multiple parameters such as ultrafast seed source output pulse width, spectral shape, chirp amount, ultrafast amplifier pulse stretching/compression, pump power, etc. to output the laser with the narrowest pulse width and maximum energy femtosecond pulse. However, the traditional multi-parameter optimization process is mainly based on manual adjustment, with heavy workload, low efficiency, poor stability and repeatability, and it is difficult to guarantee that the obtained laser output index is the optimal result.

In response to the above problems, LIFE has expanded the intelligent control technology to the whole process of ultrafast pulse generation, amplification, and transmission on the basis of intelligent mode-locking (received the important progress of global optics in 2019 and the top ten progress of China's optics). AI is used to quickly and accurately model the entire process of ultrafast pulse generation, amplification and transmission, and reverse design of ultrafast fiber laser parameters. The optical + AI architecture is proposed for global intelligent recognition of femtosecond pulses, combined with reinforcement learning, global intelligent regulation of multiple parameters of ultrafast lasers from seed source to compressor, to achieve the optimal output of ultrafast lasers, and break through ultrafast lasers A major challenge for single-pulse output stability.

Ultra-high power narrow linewidth continuous fiber laser has important application value in scientific research, directed energy laser, modern industrial processing and other fields. In the process of power amplification, the narrow linewidth continuous light seed source is affected by nonlinear effects such as the stimulated Brillouin scattering (SBS) effect, which greatly limits the output power of the amplifier. The master oscillating power amplification (MOPA) structure using the spectral broadening of the seed source can effectively increase the output power, but it is currently facing challenges such as limited spectrum control and the self-pulsation effect caused by random signal modulation, which seriously threatens system security. LIFE focuses on the intelligent spectrum control of seed sources and the improvement of laser power. Through high-precision measurement of spectra and arbitrary control of modulation waveforms, the intelligent and precise control of seed source spectra with wide-range tunable bandwidth and arbitrarily configurable spectrum types is realized, and compact structure is developed. The spectrum intelligent control module is used to explore the output power limit of the MOPA structure laser amplification system.

Transformation of scientific and technological achievements

The research direction of LIFE laboratory originates from optical fiber communication, starting from solving the nonlinear compensation of optical fiber channel to improve the capacity of optical fiber communication system, introducing artificial intelligence technology, and carrying out research on intelligent optical fiber communication system. Further expand intelligent technology to the field of fiber laser, and carry out research on intelligent fiber laser systems. In the future, it is planned to rely on the developed intelligent ultrafast fiber laser to cooperate in the research of laser fine processing. Based on the three-dimensional direct writing fine processing capability of ultrafast fiber lasers, it can process space-division multiple***ng photonic chips and photonic signal processing chips, which can be applied to optical fiber communication scenarios to further improve system capacity and break through the "capacity crisis" of communication systems. From optical fiber communication to optical fiber communication, a closed-loop research ecology is formed, which matches the English name of the laboratory "Laboratory of Intelligent Fiber Ecosystem".

Vision

The research direction of LIFE laboratory originates from optical fiber communication, starting from solving the nonlinear compensation of optical fiber channel to improve the capacity of optical fiber communication system, introducing artificial intelligence technology, and carrying out research on intelligent optical fiber communication system. Further expand intelligent technology to the field of fiber laser, and carry out research on intelligent fiber laser systems. In the future, it is planned to rely on the developed intelligent ultrafast fiber laser to cooperate in the research of laser fine processing. Based on the three-dimensional direct writing fine processing capability of ultrafast fiber lasers, it can process space-division multiple***ng photonic chips and photonic signal processing chips, which can be applied to optical fiber communication scenarios to further improve system capacity and break through the "capacity crisis" of communication systems. From optical fiber communication to optical fiber communication, a closed-loop research ecology is formed, which matches the English name of the laboratory "Laboratory of Intelligent Fiber Ecosystem".

Scientific Research Project

Presided over the National Key R & D Program, Key Technologies and Experimental Verification of Ultra-wideband and Ultra-high-speed Single-mode Optical Fiber Communication, with the project number of 2023YFB2905400, the starting and ending time of 2023.12-2016.11, and the amount of 36 million yuan.
Presided over the national major scientific research instrument development project, large-energy few-cycle femtosecond fiber laser based on intelligent regulation and control of time-frequency space domain, project No.62227821, starting and ending time: 2023.1-2027.12, amount: 7.746 million yuan.
Presided over the general fund of China Postdoctoral Fund, femtosecond pulse global single-frame intelligent measurement, project number 22Z020704850, starting and ending time 2022.7-2024.7, amount 80,000.(Pu Guoqing)
Presided over National Science Fund for Distinguished Young Scholars, "Optical Communication Intelligent Information Processing", Program No.: 62025503, starting and ending time: 2021.1-2025.12, amount: 4 million.
Presided over Huawei Technologies Co., Ltd., optical communication system modeling and equalization technology and new optical access system and algorithm research, project number YBN2019075059, starting and ending time 2020.12-2021.12, amount 2.0497 million yuan.
Presided over China Academy of Engineering Physics, deep learning method feasibility verification and module development, project number 20200380YZL, starting and ending time 2020.6-2021.6, amount of 245,000 yuan.
Presided over Huawei Technologies Co., Ltd., high-speed modulation channel compensation algorithm for optical wavelength injection, project No. YBN2020035172, starting and ending time 2020.5-2021.5, amount 590,000 yuan.
Presided over the key research and development project of the Ministry of Science and Technology, Research on Key Technologies of Artificial Intelligence in Network Layer and Physical Layer of High-speed Optical Access Network, project number 2019YFB1803803, starting and ending time 2020.1-2022.12, amount of 5.37 million yuan.
Participated in the key research and development project of the Ministry of Science and Technology, multi-dimensional parallel multiplexing optical transmission channel damage compensation and crosstalk suppression, project number 2018YFB1800904, starting and ending time 2019.7-2023.6, amount of 1.87 million yuan.
Presided over Huawei Technologies Co., Ltd., iTOF chaotic lidar signal processing technology, project No.: YBN2019045006, starting and ending time: 2019.8-2020.8, amount: 841,925 yuan.
Participated in the key research and development plan of Guangdong Province, the key technology of distributed 5G system of photonic-wireless integration, project number 2018B010114002, starting and ending time 2019.1-2022.1, amount 2.01 million.
Presided over the Youth Fund Project of National Natural Science Foundation of China in 2022, Research on High Nonlinear Threshold Fiber Laser Based on Fine Control of Spectrum
Presided over the Youth Fund Project of National Natural Science Foundation of China in 2022, Femtosecond Pulse Single Frame Global Measurement and Global Intelligent Control
Participated in the key research and development project of the Ministry of Science and Technology, "Multidimensional Parallel Multiplexing Optical Transmission Channel Impairment Compensation and Crosstalk Suppression"

Scientific Rrsearch Direction

Location：Home >Research direct...