In order to improve resource utilization in visible light communication(VLC) /WiFi heterogeneous networks, a dynamic proportional fairness(PF) algorithm is proposed. The algorithm optimizes the user's priority by introducing a compensation factor α. Firstly, the key performance indexes of VLC/WiFi heterogeneous network such as throughput, resource utilization and fairness index are theoretically derived. Then, a VLC/WiFi heterogeneous network system is constructed by using Matlab software for simulation. Finally, the performance of dynamic PF algorithm, PF algorithm and Max C/I algorithm in terms of throughput, resource utilization, fairness and access delay probability is compared. The simulation results show that the throughput of dynamic PF algorithm is higher than that of PF algorithm and much higher than that of Max C/I algorithm. The resource utilization rate of these three algorithms is higher than 0.96. The fairness index of dynamic PF algorithm is about 0.97, which is better than the other two algorithms, and the probability of access delay is lower.

In order to eliminate interference from multi-cell and multi-user signals in visible light communication systems, a two-dimensional power allocation algorithm based on orthogonal frequency division multiple access(OFDMA) for multi-cell collaborative visible light communication systems is proposed. Firstly, based on the channel situation, an OFDMA multi-cell collaborative resource allocation algorithm is adopted to allocate higher power to users with poor channel conditions. Then, based on the channel information, the transmission power ratio of multiple cooperative light emitting diode(LED) in the shared illumination area is calculated to obtain the optimal target superposition constellation. The simulation results show that compared with the average power allocation algorithm, the bit-error rate of one-dimensional power allocation algorithm is improved by 16% on average. Compared with the one-dimensional power allocation algorithm, the bit-error rate of the two-dimensional power allocation algorithm is improved by 10% on average.

Aiming at the low utilization of spectrum resources in optical network resource allocation, an improved deep Q network(DQN) reinforcement learning algorithm is proposed. Based on the ε-greedy strategy, the algorithm sets the loss function according to the difference between the action value function and the state value function, and constantly adjusts the ε value to change the exploration rate of the agent. In this way, the optimal action value function is realized, and the routing and spectrum allocation problems are solved well. In addition, different experience pool sampling methods are used to improve the convergence speed of iterative training. The simulation results show that the improved DQN reinforcement learning algorithm can not only make the elastic optical network training model converge quickly, but also improve the spectrum resource utilization by 10.09%, reduce the blocking rate by 12.41% and reduce the average access delay by 1.27 ms compared with DQN algorithm when the traffic volume is 300 Erlang.

In order to improve the performance of elastic optical networks, a combined time-frequency domain and blocking threshold triggered periodic defragmentation(PDT-TFDBT) strategy is proposed. In this strategy, the blocking delay triggering mechanism with improved fragmentation index is adopted to reduce the load of ineffective defragmentation in non-blocking conditions. In terms of sorting order, the time domain and frequency domain characteristics of services are comprehensively considered, and the reconfiguration order of services is reasonably allocated. At the same time, the heuristic deep learning-assisted routing and spectrum allocation(DKA-RSA) algorithm is introduced. The simulation results show that PDT-TFDBT strategy can not only reduce the number of reconstructed traffic and spectrum fragmentation, but also further reduce the traffic blocking rate and improve the spectrum utilization rate.

Aiming at the problem of low negotiation efficiency and throughput of traditional Cascade algorithm, a Cascade-Biconf information negotiation algorithm for codrive chaotic synchronous key distribution is proposed. Based on the principle of parity check and dichotomous error correction algorithm, the obtained key is scrambled, divided into blocks, and the check value is transmitted for dichotomous error correction. Different from the traditional Cascade algorithm, the Cascade-Biconf information negotiation algorithm reduces the interaction times and computation amount in the negotiation process by optimizing the block size. The simulation results show that when the initial bit error rate of the two communication parties is 0.1, the Cascade-Biconf information negotiation algorithm can increase the negotiation efficiency of the traditional Cascade algorithm from 0.76 to 0.84, and the total throughput can be increased by 27.9%.

In order to solve the crosstalk problem and reduce the blocking rate, an adaptive threshold and spectrum first 渊AT-SF冤algorithm is proposed. The fiber cores in each group are grouped so that they are not adjacent, and a typical 7-core fiber is dividedinto three groups. The priority level of the third group of fiber cores is variable during the service arrival process.. At the sametime, the AT-SF algorithm introduces the frequency slot渊FS冤threshold parameter, and allocates the services greater than the FSthreshold to the first, second or third group of fibers, while the services less than or equal to the FS threshold are only allocated tothe third group of fibers. The simulation experiments are carried out in NSFNET and USNET networks, and the performance offirst match渊FF冤, 3D resource allocation渊3D-RA冤, path first渊aW-PF冤and conjugate gradient spectrum first渊CG-SF冤algorithmsare compared. The simulation results show that the AT-SF algorithm can obtain better blocking rate and crosstalk performancethan other algorithms when the network is under high load.

In order to reduce the Doppler shift in the process of time-frequency transmission and improve the reliability of inter-satellite communication, a Doppler shift compensation technique is proposed. Firstly, the Doppler shift is analyzed theoretically and simulated, and the Doppler shift is obtained when the two satellites transmit the laser optical frequency before the communication is interrupted for 24 h. Then, a Doppler shift compensation system is designed for intersatellite laser Doppler shift compensation. The simulation results show that the system can complete Doppler shift compensation in about 15 control cycles, making the compensated Doppler shift less than 8 kHz, which is 6 orders of magnitude better than the uncompensated condition.

In order to make up for the shortage of microwave wireless communication in satellite network, satellite laser communication network has attracted much attention because of its advantages such as high data transmission rate, small size, light weight, low power consumption, small antenna size and strong anti-interference ability. This paper summarizes the research progress of satellite laser communication networks at home and abroad, and introduces the key technologies of satellite laser communication networks, satellite network structure, geometry of inter-satellite laser link, satellite network routing algorithm and other important content. Finally, the future development direction of satellite laser communication network is prospected.

In order to study the regulation of terahertz wave transmission characteristics by subwavelength structure of surface modified cosine grooves, a terahertz sub-wavelength metal Cu slit structure is proposed and the transmission spectrum characteristics of terahertz wave are simulated based on finite-difference time-domain method. The simulation results show that the transmission intensity and phase characteristics of terahertz waves can be controlled by changing the structure configuration, number, period and depth of the cosine grooving on the upper and lower surfaces of metal Cu slit structure.

In order to maintain the constant temperature of the constant temperature region of the directly tuned electrical/optical conversion module laser, a new type of thermal isolation high-frequency signal transmission structure is proposed. This structure uses a capacitor of a low thermal conductivity medium to transmit high-frequency signals, increasing the heat transfer resistance between the radio frequency and the laser chip in the electrical/optical conversion module effectively, and reducing the heat conduction from the radio frequency to the constant temperature region of the laser chip. The simulation results show that compared with traditional circuit chip overlap structure and gold wire cascade structure, the proposed transmission structure reduces the thermal load of the semiconductor cooler by 20.5% and 10% respectively, and reduces the current by 100 mAand 60 mA respectively. At the same time, the maximum return loss of this transmission structure in the frequency range of 2~18 GHz is only -16.7 dB, which has good radio frequency transmission performance.

In order to further improve the optical coupling efficiency and output power of the optical module, a 400G optical transceiver module is designed and optimized. A passive device with a lens is used as an important part of the optical coupling to improve the coupling efficiency. The input/output buffer information specification(IBIS) model is used to analyze and optimize the high frequency circuit of optical module. Finally, the optical design, link simulation and test of the optical module are carried out. The test results show that the designed optical transceiver module eye image is clear, the optical eye image jitter is about 2.3 ps, each channel has a good consistency, no error code within 202 s. After the signal is transmitted for 100 m, no packet loss occurs, and the system performance is stable.

A Figure-9 cavity mode-locked erbium-doped fiber femtosecond laser is designed to reduce the self-starting mode-locked threshold by introducing π/2 phase-shift bias into the cavity mode-locked oscillator to meet the need for high reliability, lightweight femtosecond lasers with output power greater than 100 mW and and repetition frequency greater than 50 MHz in special application fields. The experimental results show that the stable mode-locked pulse train output with average power of 12 mW, repetition frequency of 85.89 MHz and pulse width of 249 fs is obtained at the pump power of 130 mW. The output of femtosecond pulse sequences with a average power of 113 mW and a pulse width of 107 fs is obtained by using a stage of normal dispersion backward pumped erbium-doped fiber amplifer.

In order to reduce the power consumption of the directly modulated electrical/optical conversion module, based on the analysis of the heat dissipation of the module, a packaging method is proposed to reduce the power consumption of the directly modulated electrical/optical conversion module. The traditional Kovar alloy material is used in the area sensitive to module deformation, and the diamond copper material with high thermal conductivity is used in the area requiring fast heat dissipation at the bottom of the semiconductor cooler(TEC). The simulation and test results show that the proposed method can reduce the temperature difference between the hot TEC end face and the bottom of the component, the temperature difference between the cold and hot TEC ends, the current of TEC and the power consumption generated by TEC itself. At the operating temperature of 70 ℃, the single-channel power consumption of the electrical/optical conversion module is reduced from 2.875 W in the traditional package to 1.25 W, which is a 56.5% reduction in power consumption.

Aiming at the problem of low accuracy of traditional external modulation ranging method based on pilot signal, a phase laser ranging method based on switching circuit bias control module is proposed. Firstly, the reasons for the low accuracy of the traditional ranging method are analyzed. Then, a bias control module based on switching circuit is designed, which changes the traditional way of inputting coarse scale signal and fine scale signal at the radio frequency end at the same time, and inputting coarse scale signal and fine scale signal at the radio frequency end of Mach-Zender modulator(MZM) to realize bias control and laser distance measurement simultaneously. The experimental results show that the repeatability of phase difference is ±0.150 4° and the accuracy is ±0.026 1 cm when 500 measurements are made at a distance of 80 cm.

Aiming at the problems of low response speed and voltage control accuracy in traditional solid-state lidar, a solid-state lidar ranging system based on lens-assisted beam scanning(LABS) technology and field programmable logic array(FPGA) is proposed. The system adopts transceiver integrated structure. The LABS device in the system is composed of 1×16 optical switch chip, 4×4 optical fiber array and lens. According to the characteristics of the LABS scheme, only one optical switch is in working state at each level, and the beam is illuminated to different positions of the lens by selecting different emitters. The beam scanning is controlled by FPGA combined with an external elector, and the 4-stage Mach-Zehnder interferometer(MZI) optical switch is controlled by the output voltage to realize the fast switching of the beam. The experimental results show that the beam steering Angle step of the system is 0.35°, the maximum ranging range can reach 200 m, and the ranging error within 9.2 m is about 1 cm.

In order to solve the limitation of traditional time interval measurement method which is limited by clock resolution and high hardware cost, a dual-frequency time interval measurement method is proposed. In this method, phase perturbation or equivalent phase perturbation is performed on the received clock signal at the receiver, so that the received clock signal has a definite phase probability distribution in one period of the signal to be measured. Then, through multiple sampling and statistical calculation, the reading error caused by low clock resolution is eliminated. The simulation and experimental results show that the measurement error of the dual-frequency time interval measurement method is 0.06 ns, which reaches the measurement accuracy of sub-nanosecond and breaks the limit of clock resolution.

Aiming at the problems of complexity, high cost and lack of universal applicability of existing laser wavelength measurement methods, a laser wavelength measurement method based on 3×3 fiber coupler is proposed. The basic principle of the measurement method is deduced theoretically, and the experimental testing system of laser wavelength is built. The fiber delay line(FDL) is used to change delay difference of the interferometer measuring arm, and the half-period number of the interference signal fringe corresponding to the delay change is obtained by using field programmable gate array(FPGA). According to the relationship between half-period number and delay difference, the laser wavelength to be measured is calculated. The experimental results show that the laser wavelength measured by this method is 1310.84 nm and 1550.78 nm respectively, and the laser wavelength measured by spectrograph are 1 310.56 nm and 1 550.63 nm respectively, and the results of the two are almost the same.

When erbium-doped fiber amplifier(EDFA) amplifies the signal of the same wavelength and different frequency, the signal distortion is caused by the different frequency response. In order to improve the communication quality, an EDFA frequency response test method is introduced, and a data processing system based on Python is written. The experimental results show that this method can complete the frequency response test task, process the data quickly and accurately, realize the data visualization, and improve the test efficiency.

Reasonable planning and construction of 5G access network will help provide new generation network functions such as large bandwidth, low latency and multiple access. In order to improve the reliability of 5G access network, a high reliability wavelength division multiplexing passive optical networks(WDM-PON) architecture is designed for 5G access network by combining tree topology with dual fiber ring topology. Among them, the tree topology can ensure the transmission quality of the network in normal operation, the dual fiber ring topology can realize the flexible switching between nodes in case of failure, and provide fault protection. The signal transmission under normal operation and fault protection mode is analyzed. Through theoretical calculation and experiments on signal quality and delay, it is proved that the proposed network structure is more reliable than the conventional tree topology.

Millions of optical layer alarms may appear in optical transmission networks every month, which brings huge challenges to network operation, management and maintenance. Aiming at this problem, this paper proposes an optical network alarm compression method based on multi-dimensional attributes. First, the alarm transaction set is extracted according to the time when the alarm occurs and the network location. Then, the weight of the alarm attribute is obtained through the neural network to calculate the alarm similarity. Finally, the alarm information is compressed by alarm similarity. The actual network data verification results show that this method can effectively improve the alarm compression ratio by about 17%. By using association mining algorithm to mine association rules of compressed alarm set, the strong association alarm transaction increased by 55%, which is helpful to further analyze the correlation between alarm information and locate the fault source.

The common and secure services generally exist in optical networks. For the purpose of configuring the security service to the security links for transmission, avoiding the common services occupying the security spectrum, and reducing the optical connections set-up delay of the service requests, a low delay routing spectrum assignment algorithm with service security level awareness is proposed. The algorithm divides the network physical topology into security virtual topology and common virtual topology. The service security category and link difference are fully considered. By using the security virtual topology, the security service is configured to the security link for transmission, avoiding Eve eavesdropping the effective security service data. The parallel routing strategy of the algorithm reduces the average set-up delay of the optical connections and the blocking rate of the services, and the effectiveness is verified by experimental simulation.

Vibration signal analysis and recognition is one of the key technologies that affect the performance of perimeter system. This paper summarizes several algorithms commonly used in the field of optical fiber vibration signal recognition in recent years, including empirical mode decomposition recognition algorithm, neural network recognition algorithm, support vector machine recognition algorithm, extreme learning machine recognition algorithm, and stochastic configuration network recognition algorithm. The advantages and disadvantages of these algorithms are compared and the development trend of optical fiber vibration signal recognition algorithm is prospected.

A sliding mode control algorithm based on extended state observer(SMC-ESO) is proposed to solve the problem of tracking accuracy decrease due to the external environment of ship borne space optical communication and the influence of nonlinear model uncertainty inside the turntable. This algorithm uses the extended state observer to observe and estimate the internal disturbance of the turntable caused by the nonlinear disturbance moment and the external disturbance caused by the ship's swing, and takes the estimated disturbance value as the compensation of the sliding mode controller, so as to eliminate the influence of the disturbance on the controller and improve the control performance of the sliding mode controller. The simulation results of transmission mechanism show that the proposed SMC-ESO algorithm can improve the disturbance rejection performance by 44.22% compared with the proportional integral differential(PID) algorithm, and can meet the control requirements of coarse tracking application.

As the core device of the next-generation all-optical communication network, optical switching chip has a serious impact on the capacity of optical switching node. In order to evaluate the degradation degree of the optical switching chip to the communication system, a simulation model of optical switching integrated chip with 32×32 Benes structure is constructed using the silicon-based electro-optical Mach-Zehnder interferometers. The relationship between the crosstalk of the optical switching chip and the crosstalk of the optical switch unit is analyzed, and the influence of the optical switching chip on the optical fiber communication system is studied. The simulation results show that the degradation of receiver sensitivity due to optical switch crosstalk can be expressed by the rational function with a linear denominator for a single polarization 200 Gb/s 64QAM signal. When the crosstalk of each optical switch element is less than -46 dB, the sensitivity degradation of the optical receiver is no more than 0.5 dB.

The wavelength of fiber surface plasmon resonance(SPR) sensor is usually in the visible light band. In order to make the device work in the optical communication window, a wavelength controllable seawater refractive index sensor based on grating structure is proposed. The sensing area of the sensor is composed of side polished fiber coated with grating structure. The electric field intensity distribution of the cross section of the sensing area, the relationship between different grating structure parameters and SPR sensor wavelength, and the change trend of resonance wavelength with the change of refractive index in the range of seawater refractive index are studied. Finally, the relationship between resonance wavelength and refractive index is analyzed. The sensitivity of the sensor is 5650 nm / RIU, and the resolution can reach 1.77×10-6 RIU, which can realize fast and high sensitivity measurement of seawater refractive index.

In order to improve the detection sensitivity of glucose sensor in human environment (pH=7.4), a long period grating glucose sensor based on phenylboronic acid derivative is proposed. 4-vinylphenylboronic acid is polymerized on the surface of long period grating by C=C double bond recombination to measure glucose concentration. The experimental results show that the central wavelength of the glucose sensor can be reset to the original point after multiple repeated use, the sensitivity of the glucose sensor is 0.34 nm/(mg?mL-1) in the range of pH=7.4 and concentration of 0.1~3 mg/mL. Besides, the results show that the glucose sensor has good consistency and stability in repeated and long-term glucose determination.

In view of the problems that the existing fiber grating pH sensor can not be used for the detection of pH in the anchor cable due to its narrow response range, low sensitivity and limited application, this paper puts forward a method that fiber Bragg grating(FBG) pH sensor is embedded in the anchor cable to monitor the pH inside the cable. The working principle of FBG pH sensor is analyzed, and the FBG pH sensor is embedded in the steel strand for corresponding experimental research. The test results show that when the pH is 4-11, there is an optimum coating thickness in the FBG pH sensor embedded in the anchor cable, which makes the sensitivity of the sensor to the maximum value of pH response, and the FBG pH sensor embedded in the steel strand does not change the temperature characteristics of FBG.

In order to solve the problem of ultra-high speed space transmission and ensure the realization of our country's space-based broadband communication network, the necessity of applying digital coherence scheme to ultra-high speed space laser communication is analyzed in this paper, four problems in achieving this goal existed at current stage are summarized as follows: single-mode fiber coupling, low noise amplification of the erbium doped fiber amplifier(EDFA), algorithm of digital encoding and decoding, and hardware architecture. The solutions to these four problems are put forward one by one: fast steering mirror (FSM) tracking and adaptive nutation, segmented noise suppression, floating point parallel processing algorithm and integration. This paper analyzes the development trend of space laser communication technology, and finally gives the prediction of the future development direction of technology.

Aiming at the problem of poor signal detection performance and short communication distance of traditional direct-currnt biased orthogonal frequency division multiplexing(DCO-OFDM) underwater light-emitting diode(LED) visible light communi-cation system, a signal detection method of underwater LED visible light communication based on support vector machine(SVM) is proposed. Firstly, grid search and cross validation are performed on the signal data to obtain the optimal SVM model, then the received signal is judged by the model. The simulation results show that the system performance based on SVM is equivalent to that based on direct decision least mean square(DD-LMS) at the same communication distance and low signal to noise ratio(SNR), when the SNR is high, the former is better than the latter. Under the same bit error rate, when the communication distance is less than 3 m, the communication distance of the system based on SVM detection is slightly longer than that of DD-LMS detection.

In view of the problem that the outdoor visible light communication is influenced by the atmospheric turbulence, the influence of atmospheric turbulence and different weather conditions on the average time slot error rate(SER) of the visible light communication system under the Gamma-Gamma channel has been comprehensively analyzed by the generalized hypergeometric method. The simulation analysis shows that 16-order differential pulse position modulation(16DPPM) can effectively suppress the effects of turbulence, the average SER is significantly reduced by equal gain combined(EGC) reception merge receiving technique. Under the turbulence channel, the average SER of the 16DPPM based on single input multiple output system with three or four receiving antennas is significantly improved compared with the 16DPPM single input single output system when it is sunny, haze, mist and medium fog weather conditions.

Aiming at the problem that the light emitting diode (LED) limits the transmission rate in the existing visible light communication (VLC) system. According to the signal-to-noise ratio(SNR) and the set bit error rate(BER) of the receiver, this paper proposes a VLC system that adaptively selects the modulation mode, modulation order, channel estimation algorithm and coding mode, and carries out simulation research with Monte Carlo method. The simulation results show that the VLC system adopts adaptive modulation, channel estimation and coding, which can make full use of the spectrum resources and maximize the transmission rate and channel capacity under a certain SNR and BER.