In order to design a two-dimensional grating coupler(2D-GC) with high efficiency and low polarization coupling related loss, a plum-like 2D-GC based on silicon-on-insulator(SOI) is proposed. The 2D-GC is composed of several cells arranged periodically, and the pattern of each cell is like plum blossom. The cell in the structure is composed of two pairs of circular patterns. The finite difference time domain(FDTD) method is used to simulate and optimize the structure parameters of the 2D-GC. The simulation results show that the effective working band of the proposed 2D-GC is 1.525~1.6 μm, the optimal coupling efficiency is 56%, the corresponding coupling loss is -2.4 dB, and the polarization related loss is 0.18 dB when the parameter structure is optimized.

Aiming at the problems of low coupling efficiency, large polarization related loss and complex structure of two-dimensional grating couplers, a new two-dimensional grating coupler based on silicon on insulator(SOI) material applied in C band is proposed. The coupler is composed of four hole staggered rhombic etching cell unit cycle arrangement with a 45° incli-nation angle. The period, etching depth, relative position and radius of etching circular hole of the coupler are optimized by using inclined coupling mode, and conduct simulation verification on it. The simulation results show that the coupling loss of the coup-ler is as low as -2.4 dB at 1 540 nm, the polarization related loss at C band is less than 0.2 dB, and the 1 dB and 3 dB operating bandwidths are 36 dB and 58 nm respectively.

In order to easily achieve phase matching between interacting waves in waveguide and generate effective second harmonics, a new type of strip waveguide silica-lithium niobate-silicon dioxide(SiO2-LiNbO3-SiO2) is designed. The strip wave-guide is composed of SiO2 and LiNbO3 without etching z cutting. The dispersion of the waveguide is analyzed by adjusting the waveguide structure, and the influence of SiO2 with different sizes on the phase matching points of the fundamental wave and the second harmonic is studied. The feasibility of frequency doubling of the strip waveguide is analyzed, and the ultra-wideband continuous spectrum is realized by using the characteristics of the amplitude and width of the pulse to broaden the spectrum. The simulation results show that an ultra-short pulse with a pulse amplitude of 107 a.u.and a pulse width of 10 fs is used to obtain an ultra-wideband continuous spectrum with a bandwidth of 1 302.5 nm in a strip waveguide structure with a covering layer SiO2 width of 1 600 nm and a height of 400 nm.

A tunable metal/medium/metal type(MDM) near-infrared optical surface absorber based on Ge2Sb2Te5(GST) phase change material is proposed to realize active regulation of infrared absorption wavelength. Firstly, the absorption performance of the device is calculated and analyzed by using the finite difference time domain(FDTD) method. Then, the physical mechanism of plasma shock absorption of the absorber is analyzed according to the electric field distribution characteristics. The simulation results show that the absorption peak center position of the proposed absorber is shifted from 1.17 μm to 1.8 μm, and the deviation width is 630 nm. When the appropriate crystallization fraction of GST is selected, the absorber can achieve a high-low absorptivity conversion at a communication wavelength of 1.55 μm.

In order to achieve high integration and efficient output of optoelectronic information devices, and improve the optoelectronic information processing ability, a four-channel filter based on two-dimensional photonic crystals is designed. According to the coupling principle of four point defect microcavities and line defect waveguides, the reflection heterojunction is introduced and the distance between the micro-cavity and the reflection junction is adjusted to improve the output efficiency. The wavelength division multiplexing function of four-wavelength of 1 403, 1 426, 1 449 and 1 508 nm is simulated. The simulation results show that the device can achieve efficient transmission, the transmission of all four wavelengths is more than 95%, the insertion loss is less than 0.23 dB, and the cross-talk between channels is less than -8.7 dB.

Aiming at the problems of high cost and complex algorithm of conventional coherent receiver in optical access network, a low cost simplified coherent receiver(SCR) signal processing method is proposed. The transmitter uses four-level pulse amplitude modulation(PAM4) high order code to modulate the signal, and the receiver uses unbalance compensation algorithm and feedback full digital clock recovery algorithm to process the output signal, and calculates the bit error rate through frame synchronization and code element decision. The simulation and experimental results show that the proposed method has good performance in the system. When the optical power of the input signal is -9 dBm, the bit error rate of the system is 10-4, which is lower than the bit error rate decision of the hard decision forward error correction(HD-FEC) threshold.

In order to solve the problem of abnormal with reception and monitoring of the optical transceiver integrated module without optical input, a method of accurately selecting trans-impedance amplifier decoupling capacitors and using software calibration is proposed. First, the I2C bus waveform and receiving power noise waveform are tested in case of abnormal monitoring, and the impact of various types of trans-impedance amplifiers and different monitoring decoupling capacitor values on receiving monitoring are verified. It is confirmed that the abnormal receiving monitoring is mainly caused by two reasons: the trans-impeda-nce amplifier monitoring residual noise and monitoring capacitance decoupling. Then, select appropriate decoupling capacitor of the monitoring signal through experiments, and using the second-order equation for software calibration. The test results indicate that this method can completely solve the problem of receiving monitoring anomaly when there is no light inputting.

Aiming at the problems such as low positioning accuracy and serious signal interference in the large-area panel light arrangement scene of indoor visible light positioning(VLP), a code division multiple access(CDMA) modulation and demodulation VLP system based on received signal strength(RSS) algorithm is proposed, The principles of the RSS algorithm and CDMA modulation and demodulation VLP system are introduced, and the positioning accuracy of the system is verified by experiments. The experimental results show that the maximum error of the positioning of the fixed point is 0.3 m without considering the factors such as real-time movement and delay refresh. The algorithm solves the problem of low positioning accuracy of the edge corner region in the traditional positioning algorithm and improves the robustness of the VLP system.

The performance of the optical power detection circuit determines the quality of the received signal. This paper introduces the classification of semiconductor photodetectors in optical power detection circuits, as well as several commonly used transimpedance amplifiers connected to photodiodes, and compares their advantages, disadvantages and application ranges according to the performance indicators of transimpedance amplifiers. Finally, the prospects for optical power detection circuits development are briefly described.

In order to analyze the main factors that affect the mechanical reliability and lifespan of optical fiber, through the simulating test conditions, the causes of optical fiber fracture are analyzed. The surface morphology of the experimental fiber fracture was compared with the morphology characteristics of the optical fiber fracture source that appeared in actual working conditions, and the main reasons for the formation of optical fiber fracture were qualitatively analyzed. Analysis suggests that the presence of water or vapor in the working environment under external stress accelerates the growth and propagation of surface cracks on the optical fiber, ultimately leading to fiber fracture.

Aiming at the problems of low pattern recognition accuracy of high-order composite Laguerre-Gaussian(LG) beams with complex atmospheric turbulence interference and non-zero radial index, a convolutional neural network(CNN) based orbital angular momentum(OAM) pattern recognition method for high-order composite LG beams is proposed. An OAM pattern recognition model based on CNN is constructed to study the influence of different wavelength and transmission distance on the pattern recognition accuracy of radial high-order composite LG beam under complex atmospheric turbulence conditions. The simulation results show that, under the condition of complex atmospheric turbulence, when the transmission distance is 1 km and the CNN training is 100 times, the pattern recognition accuracy of the high-order composite LG beam with wavelength of 850, 1 310 and 1 550 nm is above 98.8%. When the wavelength is 1 550 nm and the transmission distance is 1, 2 and 3 km respectively, the pattern recognition accuracy of the model is above 86.8%.

In order to meet the rapid growth demand of space information transmission capacity, and in response to the development trend of high speed and large bandwidth of future satellites, a Beidou navigation network optimization method based on laser microwave hybrid inter satellite links is proposed. The ability of inter-satellite link ranging and routing convergence efficiency of hybrid link are analyzed, and simulation is carried out to verify the effectiveness of collaborative ranging and collaborative management methods in improving the efficiency of hybrid network. The simulation results show that the Ka laser cooperative ranging time is shorter than the traditional microwave network, and the maximum positioning accuracy factor(PDOP) decreases from 3.4 to 2.3. The routing convergence speed of hybrid network is faster than that of single laser network, which decreases from 41 s to about 15 s.

To solve the problem that the nonlinear suppression method requires prior statistical information to determine the optimal threshold and causes nonlinear distortion, an improved deep neural network(DNN) algorithm for power line impulse noise suppression and compensation is proposed. Firstly, the position of pulse noise is locked by DNN, and then the data of the position is cleared by zerosetting method. Finally, the nonlinear distortion is reconstructed and compensated for the processed signal, and the simulation and comparison experiments are carried out with other algorithms. The simulation results show that the proposed method improves the recognition rate of the DNN, reduces the bit error rate of the system, and has good robustness.

Due to the harsh natural environment and the change in the arc of the cable itself will affect the transmission efficiency of the all-medium self-supporting(ADSS) cable, and some environmental hazards such as construction or ultra-high vehicles and fireworks will seriously threaten its normal operation, an ADSS optical cable sag online monitoring and environmental hazard warning algorithm is proposed. Firstly, the angle measurement method is used to monitor the radian in real time. Then, the detection algorithm of YOLOv4 is improved, and the visual intelligent fusion terminal is designed. Finally, the proposed algorithm is compared with other algorithms. The experimental results show that the average accuracy of the proposed algorithm is 96.43%, which realizes real-time monitoring, data processing, fault analysis and intelligent hidden danger alarm of the communication optical cable running state, and integrates into the ADSS optical cable online monitoring visual intelligent terminal.

In order to solve the difficulty of reliable cooperative operation of five-prevention system in regional substations, a blockchain-based intelligent five-prevention control method for regional substations is proposed. Firstly, the communication channel between substations is established through virtual private network to realize the secure communication between substations. Then, the blockchain technology is used to establish the five-prevention trusted authentication mechanism of the equipment among the regional substations to achieve the trusted cooperative five-prevention operation of the regional substations. Finally, the method is applied in the simulation environment of a regional substation in a province to implement the intelligent five-prevention control of regional substation. The simulation results show that the correct execution rate of the proposed method is 99.92%. This method can realize the reliable authentication operation among the equipment of regional substation and improve the five-prevention management level of regional substation.

Aiming at the problem of limited transmission range and intermittent connection of nodes in underwater optical wireless sensor network(UOWSN), a network node localization algorithm is proposed by using multi-hop communication to expand the transmission range to enhance network connectivity. Firstly, the UOWSN is modeled as a three-dimensional(3D) random scaling model graph, and the connectivity probability expression of the network nodes under the model is derived according to the number of network nodes, communication range and optical divergence angle. Then, the received signal strength (RSS) location algorithm modified distance vector hop(DV-Hop) localization algorithm is used to solve the localization problem of network nodes, and compared with the existing schemes. The simulation results show that the performance of the proposed localization algorithm is significantly better than the standard DV-Hop algorithm and Hybrid DV-Hop algorithm.

In view of the situation that the wireless communication technology in coal mine can not meet the intelligent development needs of high speed, low delay and wide coverage, this paper introduces the working principle and technical advantages of light fidelity(Li-Fi), emphatically analyzes the overall architecture of Li-Fi network system in coal mine and the typical application scenarios of Li-Fi in coal mine, and finally looks forward to the future development of Li-Fi technology from six aspects research direction.

Aiming at the problems of hardware pre equalization technology and software pre-equalization technology based on finite impulse response(FIR) filter, a digital pre-emphasis software pre-equalization technology is proposed. The technology compensates the LED bandwidth according to the frequency response parameters of light-emitting diode(LED), and designs a direct-current biased optical orthogonal frequency division multiplexing(DOC-OFDM) underwater visible light communication system as an experimental test platform. The experimental results show that the modulation bandwidth of LED is increased from 7.8 MHz to 16 MHz by using digital pre-emphasis technology, and the implementation complexity of this technology is low, the flexibility is strong, and the signal spectrum after pre-equalization is more flat.

Aiming at the problem of poor bit error performance of space pulse position modulation(SPPM) in vertical link, concatenated channel coding is used to improve the bit error performance of the system. The bit error performance of low density parity check(LDPC) code and polar code under different code length and different code rate is compared and analyzed. The simulation results show that when the bit error rate is 10-4 and the bit rate is 0.5, polar code has a better performance than LDPC code about 2 dB. When the bit rate is 0.8, the bit error performance of the two systems is almost the same, and the performance gap is gradually narrowed when the bit rate increases. In SPPM system, concatenated channel coding can effectively improve the system error performance.

The performance of photoelectric transceiver module is very important to the whole communication system quality. An optoelectronic collaborative link simulation system including optoelectronic devices, packaging structure and other parameters is designed. Firstly, the electrical parameters of the package structure are obtained by optimizing the system package scheme, layout and signal integrity. Secondly, the photoelectric device parameters are optimized and an optoelectronic collaborative system is established which includes the influence of the electrical parameters of the packaging structure on the system performance simulation link, through the optoelectronic collaborative link simulation, the module is optimized and evaluated in advance to ensure that the optoelectronic hybrid integrated system meets the requirements. Finally, the module is assembled and tested. The simulation results show that under the input of 25 Gb/s signal source, the module carries out self receiving through 1 km optical fiber. Under the 213-1 code type, the error code of the module is less than 1E-15, and the total power is only 3.6 W.

In order to extend the signal transmission distance, facilitate the detection and noise reduction, and retain the signal energy as much as possible, the carrier suppressed single sideband(SSB-CS) signal output can be realized by phase shifting method. Using pilot method to control commercial lithium niobate double parallel (IQ) modulator can achieve the above goal. The experimental results show that the carrier to side ratio and side to side ratio of the output signal are greater than 20 dB in the range of 2-18 GHz. This method can effectively control the working state of the modulator without using the built-in photodiode (PD) of the IQ modulator, and the carrier side ratio and side side ratio can meet the expected SSB-CS signal.

Aiming at the high computational complexity of traditional selective mapping(C-SLM) algorithm, an improved grouping(G-SLM) algorithm is proposed. Without increasing the computational complexity of the system, the number of time-domain candidate signals is increased from M to M 24, and the optimal α value is determined by simulation. In view of the limited peak-to-average power ratio(PAPR) suppression performance of C-SLM algorithm, a μ-GSLM joint algorithm is proposed by combining G-SLM algorithm with μ-law compression expansion method, which further reduces the PAPR of the system. The simulation results show that when complementary cumulative distribution function(CCDF) is 10-3, the PAPR of μ-GSLM algorithm is reduced by 7.5 dB, 3.2 dB and 2.1 dB respectively compared with the original signal, G-SLM algorithm and μ-law compression expansion method.

Aiming at the problems of high computational complexity of traditional discrete Fourier transform(DFT) channel estimation algorithm and inefficient channel coefficients in cyclic prefix(CP) in direct-current biased optical orthogonal frequency division multiplexing system, an improved channel estimation algorithm based on Hermite symmetric DFT(H-DFT) threshold is proposed. Firstly, according to the property that the time domain response of the channel is transformed to the frequency domain to satisfy the Hermite symmetry, the frequency domain response of all the subcarriers can be obtained by only making Hermite symmetry for the frequency domain response of the effective subcarriers. Secondly, according to the average power of the noise outside the CP, the threshold value is set for the channel tap data inside the CP to filter the noise inside the CP. Compared with the traditional DFT channel estimation algorithm, the simulation results show that the H-DFT threshold improved channel estimation algorithm can effectively reduce the computational complexity, and improve the bit error rate and mean square error.

The traditional routing spectrum allocation(RSA) algorithm in elastic optical network(EON) mostly considers the hop number or spectrum resource occupation, and lacks the effective utilization of time domain and adjacent link information. This paper proposes a multi-dimension-aware RSA algorithm combined with prediction, which predicts the time information of future services through back propagation neural network for the known duration of services time, spectrum and adjacent link resource occupancy are considered. The simulation results show that compared with the traditional RSA algorithm, multi-dimension-aware RSA algorithm can effectively reduce the bandwidth blocking rate.

In order to solve the problems of low sensitivity, high loss and difficult control of nonlinear effect of traditional fiber, a novel elliptical core like photonic crystal fiber(PCF) structure with high birefringence, low loss and high nonlinearity is designed. Based on the full vector finite element method, the effects of the size and position of the air hole on the birefringence, confinement loss, mode field characteristics and nonlinearity of the fiber end face are analyzed by COMSOL. The simulation results show that the birefringence index of the PCF is 1.918×10-3 at the wavelength of 1.550 μm, the limiting loss of x and y polarization direction is Lcx=1.6×10-3 dB/km and Lcy=8.0×10-4 dB/km respectively, the nonlinear coefficient is 9.4 km-1W-1, and the single-mode transmission is satisfied. The high-quality and high-precision optical signal transmission and sensing are realized.

In optical time domain reflectometer(OTDR) test, the system link noise seriously affects the test accuracy, especially for the accurate detection of 0.05-0.2 dB fusion loss, it is necessary to improve the signal-to-noise ratio of the test curve. According to the characteristics of OTDR waveform, based on wavelet threshold denoising, an improved wavelet denoising method is proposed . The improved threshold makes use of the nonlinearity of logarithmic function and introduces noise variance, which is more conducive to filtering out noise; the improved threshold judgment function leaves original data details and filters out noise, and does not change the waveform attributes of reflection events and non-reflection events. The actual test results show that the relative signal-to-noise ratio of the test curve is increased by more than 10 dB, and the fusion event can be accurately detected after denoising with curve smoothing filter and constant false alarm rate(CFAR) detection.

Surface plasmon resonance(SPR) technology has been widely used in refractive index measurement in the external environment. In this paper, a symmetrical planar waveguide structure is designed. The relationship between SPR effect and the refractive index of the external environment is numerically simulated by using the finite difference time domain method. The selection of metal materials and the length of sensing area are optimized. The effects of modes of different orders on the sensitivity of the sensor are studied. The simulation results show that when the refractive index of the external environment is 1.38, compared with the sensitivity of 4100 nm/RIU under the base mode condition, the sensitivity of the third-order mode sensor is increased to 6209 nm/RIU, and the maximum sensitivity is increased by 51%. When the ambient refractive index is 1.34-1.38, the average sensor sensitivity increases from 2900 nm/RIU to 4025 nm/RIU.

In order to improve the performance of intensity modulation direct detection(IMDD) systems based on band-limited optoelectronic devices and direct modulation lasers(DML), solve the problem of high computational complexity of traditional equalizers, a simplified scheme for deep neural network(DNN) equalizers is proposed. Firstly, the adaptive momentum estimation (Adam) algorithm is used to update the weight coefficients of DNN, optimizing the iteration speed and convergence performance of traditional gradient descent algorithm. Then, based on this, discard layers and pruning operations are introduced to reduce the high computational complexity of DNN, reduce redundant connections in the network structure, and avoid overfitting phenomenon. Finally, the effectiveness and feasibility of the simplified scheme of DNN equalizer are verified in an 80 Gb/s bandlimited DML-IMDD simulation system.