“New infrastructure” provides more valuable opportunities for the further development of optoelectronic industry, thus the development of optoelectronic industry should strengthen the capability of basic research and development first. In this article, it focuses on the characteristics of the processes such as epitaxial growing, filming deposition, lithography, bonding, and etc., applied in the fabrication of compound semiconductor photo-electronic devices. And also analyzed are the capacity building trends of optoelectronics industry such as intelligent manufacturing, digital manufacturing, decentralized manufacturing and green manufacturing.

Due to the low near-infrared absorption coefficient of silicon, there is a conflict between response speed and detection efficiency for conventional surface-illuminated silicon detectors. Thus they are thought to be not suitable for short-range optical fiber communication. Micro-/nanostructures have been widely used in solar cells, near-infrared enhanced detectors and other fields for their abilities to improve the absorption of incident light in absorption layers and the quantum efficiency of optoelectronic devices by multiplying optical path efficiently. Recently, CMOS compatible high-speed silicon detectors with a data transmission rate of higher than 20Gb/s have been achieved by applying photon-trapping microstructure. In this paper, the optimal design and preparation methods of micro-nano structures and high-speed silicon detectors based on micro-nano structure are reviewed and analyzed.

In this paper, a kind of bidirectional electro-optic Mach-Zehnder intensity modulator was designed for microwave full-duplex transceiver system. According to the principle of modulation and superposition of electric signals to optical signals in the electro-optic modulator, analyzed is the deterioration of isolation caused by the distribution difference between the designed electric field and the actual electric field of the modulated electrode. By comparing the analysis, design and simulation results of different modulation electrode structures, it is concluded that the bidirectional electro-optic intensity modulator with the isolation better than -30dB above 3.5GHz is fabricated successfully, achieving an isolation better than -30dB in the range of 5~17GHz.

In the manufacturing process of all-fiber Mach-Zehnder interferometer, it needs to cut, weld and taper the photonic crystal fiber and single-mode fiber. First, by using the theory of contact mechanics, the internal stress of the photonic crystal fiber in the V-groove was analyzed, and the maximum stress point inside the photonic crystal fiber was deduced at the outermost air hole parallel to the contact surface. In conclusion, the stress distribution diagram and strain distribution diagram of the photonic crystal fiber were given, and the maximum load when the photonic crystal fiber is cut, fused and fixed was calculated. Then, the stress analysis of the fiber with air ball inside was carried out, and the maximum stress point inside the fiber was deduced from the edge of the air ball closest to the surface of the fiber. Finite element analysis was used to verify this conclusion, and the internal tension and stress distribution diagrams were given, providing a theoretical basis for the taper of the air ball in the optical fiber to the spindle shape.

Piezoelectric resonators, usually working at a resonant state with stable amplitude, are the core components of resonant sensors and actuators. As an approach to research the piezoelectric resonator driving circuit, a dual closed-loop control circuit is used to realize the resonant frequency tracking and the amplitude stabilizing through phase-locked loop (PLL) and automatic gain control (AGC). By optimizing the parameters of each component in the dual closed loop circuit, the piezoelectric resonator can work stably at the resonant frequency. The results of the open-loop experiment and the closed-loop experiment show that the dual closed-loop circuit can lock the resonator phase at 59°, and the error is less than 3°. The amplitude of output signal is controlled at 12.8V, and the error is less than 0.2V. The experimental results show that the circuit is capable of stabilizing the amplitude of the resonator output signal and tracking the resonant frequency of the resonator.

In order to realize the design of temperature sensor with lower cost, a new type of intensity modulated optical fiber temperature sensor is proposed. It mainly consists of light source with band-width spectrum, 3-port circulator, fiber coupler, polarization controller, polarization maintaining fiber (PMF), photodetector and signal processing unit. Under the condition of short PMF fiber, the temperature change will have a direct impact on the birefringence of PMF fiber, resulting in a certain phase shift change of the two coherent beams, thus they will have different transmission spectra after Sagnac interferometer, that is, wavelength drift occurs. By using photoelectric conversion and signal processing, the detected light signal is converted into a voltage signal, which is solved to complete the temperature detection. Therefore, without the need for a spectrum analyzer, the measured temperature can be calculated according to the temperature-dependent change function by measuring the transmission spectrum of the Sagnac interferometer. Experiments were carried out in the temperature range of 25~50℃. The test results show that the sensitivity of the optical fiber temperature sensor is 0.066mW/℃, and the resolution is 0.34℃. It has good application value for constructing temperature monitoring optical fiber sensor network.

The radiation in the space environment will degrade the performance of CMOS image sensors, even result in permanent damage. In this paper, the radiation harden techniques are proposed from the aspects of layout and circuit designs, and experiments of total-dose and signal-event were carried out on the samples. Experimental results show the CMOS image sensors fabricated with radiation-harden technology realize total-dose resilience up to 100krad(Si) and single-event effect resilience up to 1×107p/cm2, meeting the requirement on radiation hardened design.

The spectrum characteristics of gain and refractive index change of TE mode and TM mode of InGaAs/InGaAsP columnar quantum dot versus dot material components, barrier material components and aspect ratio are contrastively analyzed, and the physical mechanism are dissected. The effects of changes of dot components and barrier components on balancing gain and refractive index change and polarization are further jointly investigated, then a multi-parameter adjustment method is proposed, and the In0.97Ga0.03As/In0.76Ga0.24As0.52P0.48 quantum dot with low-polarization that balances gain and refractive index change within 1550nm communication band(1540~1560nm) is designed. Finally, the appropriate carrier concentration is selected through analysis. When the carrier concentration is 0.6×1024m-3, the overlap region area of 3dB spectrum width of TE mode and TM mode is 8.66×103nm/cm and 7.55×103nm/cm respectively, and the polarization dependence of gain and refractive index change is smaller than 3% and 10%, respectively. The research result is helpful for the optimization design of some key devices in the future all-optical network.

High-purity chiral single-walled carbon nanotubes are of great significance for the development of next generation of carbon-based electronic devices. In this paper, PFO-BPy, PFO, and PFO-BT polymers were used to separate the (6,5), (7,5), (10,5) chiral single-walled carbon nanotubes (s-SWCNTs) with high purity and high concentration of dispersant and more than 99% of the residue was removed. High-density and high-uniformity carbon thin films were obtained by using the above deposition liquid. And then chiral s-SWCNTs field effect transistor (FET) arrays were fabricated. Experimental results indicate chiral s-SWCNTs FETs with large diameter (10,5) have better electrical properties, the mobility is up to 16cm2?V-1?s-1 and the Ion/Ioff ratio can reach 107.

C-doped TiO2 thin films were prepared by magnetron sputtering and the influence of nitrogen on the films during the sputtering process was studied. The microstructure, composition, optical properties and morphology of TiO2 thin films prepared under different nitrogen flow rates were investigated by X-ray diffraction, Raman spectrometer, X-ray photoelectron spectroscopy, spectrophotometer and atomic force microscopy. The results showed that all the deposited films were mainly amorphous and some weak anatase phase was found in Raman spectra. Moreover, with the increase of nitrogen flow rate, the characteristic peak intensity of anatase decreased, which meant grain refinement. When the nitrogen flow rate was up to 4cm3/min, the N content in the C-doped TiO2 films was 3.54% with its optical band gap changing from 3.29 to 3.55eV. At the same time, the transmittance of C-doped TiO2 films increased obviously. It can be concluded that nitrogen flow rate can control the optical band gap and optical absorption properties of the C-doped TiO2 films effectively.

Paraxial transmission characteristic of a finite Olver-Gaussian beam (FOGB) in a LiNbO3 electro-optical crystal is investigated based on the generalized Huygens-Fresnel integral equation. The exact expression for the first-order FOGB passing through this electro-optical crystal orthogonal to the optical axis is derived. The contour graph of the FOGB intensity distribution on some transversal cross sections of the crystal, and the side view of this beam propagating evolution are discussed, respectively. The relation between external electric field and the central beam location, central relative beam intensity are also explored when wave propagating in the electro-optic crystal. Furthermore, the transmission characteristics of the FOGB in three kinds of traditional uniaxial crystal orthogonal to the optical axis are discussed as well.

With the continuous increase of signal frequency and device integration density in electronic system, serious coupling noise problem in three-dimensional integration system becomes the bottleneck of system performance. Aiming at the three-dimensional interconnection widely used in integrated technology, which is composed of trough silicone via (TSV) and redistribution layer (RDL), an optimized design scheme for three-dimensional interconnection is proposed based on segmental transmission line (STL). In this scheme, the three-dimensional interconnection is divided into several segments under the STL requirement, and the genetic algorithm (GA) is used to screen the segments transmission characteristics to optimize superposition effect of reflected waves, which can compensate the signal loss in the transmission process. The simulation results show that the scheme proposed in this paper reduce the coupling noise caused by wave reflection, and effectively improve the transmission performance in the three-dimensional interconnect system.

A new type of photonic crystal fiber structure is designed in this paper. The fiber cladding contains six axisymmetric triangular lattices composed of air holes with different shapes, and a row of elliptical air holes with different sizes were introduced along the x-axis in the middle to achieve high birefringence, ultra-low confinement loss and large negative dispersion. The model of the photonic crystal fiber was constructed by the finite element method, based on which, the effects of the rotation angle of the two central air holes on the mode field distribution, birefringence coefficient, confinement loss and dispersion coefficient of the LP01 mode and LP11 mode were studied. The results show that the fiber has the optimal performance when the rotation angle α=90°, that is, at the wavelength of 1550nm, the LP01 mode can obtain the birefringence, confinement loss and dispersion as 3.618×10-2, 1.999×10-14dB/m and -764ps/(nm?km), respectively.

The Sn prefabricated layer was evaporated by electron beam and then selenized by Se powder, then SnSe films were successfully prepared on the glass substrate by adjusting the selenation temperature and annealing time. The phase, microstructure and optical properties of SnSe films were studied by X-ray diffraction, Raman spectroscopy, scanning electron microscopy and UV-Vis-NIR spectrophotometer. The results show that the pure phase polycrystalline SnSe films with a band gap of 0.93eV can be prepared by annealing selenide for 60min at 450℃. Under the irradiation of a 980nm laser with a power of 200mW/cm2, the photoelectric response characteristics of SnSe film were tested, and the response time and recovery time of the prepared film were 62ms and 80ms, respectively, obtaining through curve simulation.

Back thinning is an important process for the fabrication of InP-based optoelectronic chips. When the wafer loses the structural support after thinning, its warping degree will increase due to the severe deformation caused by stress. Serious warping will reduce the chip reliability or even make it invalid, thus the warpage of the wafer should be controlled and corrected. In this paper, based on the theory of ‘damage layer-warping degree’, the influences of wafer thickness, adhesive mode, grinding pressure, grinding speed and abrasive particle size on warping degree were analyzed. And the process parameters were optimized according to the test results, and the warping degree of the wafer was reduced by about 20% after optimization. Then wet corrosion was used to remove the damaged layer and correct the warping, which reduced the warping degree of the wafer by about 90%. Using the optimized thinning process to reduce the damage stress and wet-corrosion to remove the damage layer are the applicable methods for controlling and correcting the warping degree of the wafer, which can reduce the warping degree to be less than 10% of that before the back-thinning process.

The thermoelectric coupling characteristics of a kind of n-type piezoelectric semiconductor nanowire (ZnO) were studied with finite element method. The effects of external temperature on internal distribution of mechanical, electric and current fields of the ZnO nanowire were analyzed. And effects of the linearization of the constitutive equation on electric parameters were discussed. The results show that the external temperature greatly affects the electric field, carrier concentration and electric current density. The electric field, carrier concentration and current density derived from linear and nonlinear constitutive relation differ greatly, the maximum difference are 24%, 32% and 68%, respectively. A large error will appear when one adopts linear constitutive relation to analyze electric properties in piezoelectric semiconductors.

Process control monitor (PCM) is a key quality monitoring technique reflecting the process status of the production line. In this paper, it focuses on the factors affecting the test results of PCM in the fabrication process of charge coupled device (CCD), and the test results of PCM is statistically analyzed. The results show that, the sheet resistance is 18Ω/□ when the deposition temperature of LPCVD is 700℃ and the film thickness is 580nm. The contact resistance is 7Ω when the hole process adopts dry etching with CF4 flow rate of 15cm3/min and CHF3 flow rate of 45cm3/min. The threshold voltage of the MOS transistor is -8.5V when the energy and dose of phosphorus ion implanted into the buried channel under the gate are 250keV and 2.5×1012atoms/cm2, respectively. It can effectively avoid the metal leakage caused by residues when the secondary aluminum etching adopts Cl2 flow rate of 90cm3/min, BCl3 flow rate of 45cm3/min and N2 flow rate of 30cm3/min.

The uniformity of the uniform light system is the key for the parameter test of deep ultraviolet CMOS image sensors. Based on the Fourier optics theory and combining the characteristics of the ArF excimer laser output spot, the initial structure of the compound-eye array uniform light system was designed, and a uniform light system model was established in ZEMAX non-sequential mode. According to the characteristics of the randomness of the light sampling in the ZEMAX light source and the uniformity requirements of the uniform light system, the number of tracking rays and the number of lens elements in the compound eye array were optimized. By setting the lens element size of 1mm, applying 100million rays and being averaged for 30 times, a uniform illumination spot with a uniformity of 0.986 is obtained in the 12mm×12mm spot range, which is better than the requirement on the spot uniformity of 0.97 for CMOS image sensors.

Aiming at the problem of low SNR of phase-sensitive optical time domain reflectometer (Ф-OTDR), a de-noising method based on improved variational mode decomposition (VMD) and independent component analysis (ICA) is proposed. First, the simulated annealing method (SA) was used to optimize the VMD. Then, SA-VMD was used to decompose the pre-processed Ф-OTDR signal into a series of intrinsic mode function (IMF) components, and IMF components were selected for virtual noise reconstruction according to relevant criteria. Finally, the original signal and virtual noise were used as the input of ICA to remove noise and improve the SNR of the signal. Experiments were carried out to verify the proposed method on self-designed coherent Ф-OTDR system, and the results show that the method can effectively remove noise. Compared with the SA-VMD and EMD-ICA methods, the SNR is improved by 4dB, which is of great significance to the practical application of the system.

In distributed optical fiber sensing based on Brillouin scattering, Brillouin frequency shift (BFS) is linear to temperature and stain in the optical fiber. In order to improve the measurement accuracy of temperature and strain, an improved quadratic polynomial fitting algorithm is proposed. In the algorithm, the median filtering algorithm was proposed to preprocess the noisy Brillouin spectra, so as to improve the accuracy of gain peak location; then spectra within one linewidth and symmetrical about the peak gain were intercepted to extract the BFS precisely using the quadratic polynomial fitting algorithm. Firstly, after systematic comparison according to BFS error and error in frequency corresponding to peak value gain, the Brillouin gain of the same frequency corresponding to all spatial points was selected as the input signal. Subsequently, the effect of the proposed algorithm under different frequency intervals, signal to noise ratios (SNRs) and different filter window sizes was studied, meanwhile the optimal window size selection problem was investigated. The results show that the BFS error decreases first and then increases as the window size increases, and the general optimal window size ranges from 53 to 163.

Sign language recognition is widely used in communication between deaf-mute and ordinary people. In adequate extraction of spatial-temporal features in sign language recognition task is likely to result in low recognition rate. In this paper, proposed is a novel sign language recognition model based on spatial-temporal attention which can learn more discriminative spatial-temporal features. Specially, a new spatial attention module based on residual 3D convolutional neural network (Res3DCNN) is proposed, which automatically focus on the salient areas in the spatial region. Then, to measure the importance of video frames, a new temporal attention module based on convolutional long short-term memory (ConvLSTM) is introduced. The crucial purpose of the proposed model is to focus on the salient areas spatially and pay attention to the key video frames temporally. Lastly, experimental results demonstrate the efficiency of the proposed method on the Chinese sign language (CSL) dataset.

The adaptive optical system of a large ground-based telescope needs to use a high-speed and high-sensitivity camera to measure the wavefront error in real time to obtain a target image close to the diffraction limit. The multi-tap electron-multiplying charge coupled device (EMCCD) camera is one of the best choices for adaptive optical wavefront detection. Based on the 8-tap CCD220, a high-speed (2000f/s), high-precision and multi-channel synchronous timing generator is designed. A variety of pixel combinations have been implemented on the device, thus the camera frame rate is increased to 3500f/s (2×2 merge) and 5700f/s (4×4 merge), and the area of interest of the camera can be controlled. The stepping accuracy of the timing generator can reach 2.5ns, and the phase jitter of each driving signal output can reach below 200ps.

Affected by the aero-optical effect, the wavefront of the light wave from the target will cause dynamic disturbance, which will cause blurring in imaging. A common correction method is to perform deconvolution processing based on measuring the wavefront so as to restore the image. The traditional wavefront sensor can only measure the central field of view, thus the image area that can be restored is too small due to non-equal halo problems. As a new type of wavefront sensor, the light field camera wavefront sensor has the advantages of large field of view and large dynamic range. It can simultaneously detect the point spread functions of different areas of a blurred image, so as to restore the full image at one time. The large-field wavefront detection characteristics of the light-field camera were numerically simulated by using MATLAB software, and the blurred images caused by aero-optical effects were clearly processed, and compared with the simulation results of the Shack-Hartman sensor. The results show that the wavefront sensor of the light field camera can detect the wavefront disturbance caused by the pneumatic optical effect effectively and the image of the whole field of view can be clearly detected at one time, and the field of view is several times of that of the traditional wavefront sensor.

In order to improve the performance of visible light communication (VLC) systems, a novel construction method of quasi-cyclic low-density parity-check(QC-LDPC) codes is proposed to solve the problem that the minimum distance between code words of QC-LDPC codes is not large enough, which leads to the degradation of the error correction performance. In the proposed construction method, the information bit of QC-LDPC codes is constructed by combining the greatest common divisor (GCD) algorithm with Lucas sequence. At the same time, in order to reduce the encoding complexity, the check bit is designed in the form of the quasi-double diagonal, which can ensure the large girth length and realize the fast encoding of QC-LDPC codes. Then the GL-QC-LDPC (2650,1325) code with a code rate of 0.5 is constructed with the proposed construction method, and the simulation performance is analyzed in the established VLC system simulation model. Simulation results show that at the bit error rate(BER) of 10-6, the net coding gain(NCG) of the GL-QC-LDPC(2650,1325) code constructed by this method is respectively 0.10, 0.14 and 0.25dB more than those of AC-QC-LDPC(2652,1326) code based on the array codes(AC), GM-QC-LDPC (2650, 1325) code constructed directly using the GCD algorithm and the modification technology and GDD-QC-LDPC (2652, 1326) code based on group divisible designs (GDD).

Due to the high resonance frequency of the multi-ring resonant micro-mechanical gyro, the traditional digital control circuit can not accurately balance the accuracy and tracking speed of the gyro amplitude signal. Based on the traditional hemispherical gyro digital control loop, a frequency tracking circuit is proposed for multi-ring resonant micro-mechanical gyroscopes, and it is applied to multi-ring resonant micro-mechanical gyros for the first time. The circuit is based on a high-speed A/D conversion circuit. The frequency and phase information is calculated by high-speed sampling of the amplitude signal, and the output signal is generated by the CORDIC algorithm. Test results show that the circuit makes the frequency tracking accuracy of the multi-ring resonant micro-mechanical gyro amplitude signal reach 0.78Hz, and the frequency tracking time is less than 40μs, which greatly improves the performance of the control circuit.

In this paper, demonstrated is a super-resolution polarization-modulated 3D imaging framework based on electron multiplying charge coupled device (EMCCD). Due to its low bandwidth characteristics, the electro-optic modulatoris applied as a fast shutter with sub-nanosecond-level to implement temporal and range resolution of 3D imaging system simultaneously. The depth information can be reconstructed from the gray image with only a frame of raw data, so as to achieve dynamic imaging performance. Based on the system structure, a method of multi-dimensional data fusion is proposed to calculate target position and attitude. The feasibility of the method is verified by experiments. Experimental results show thatfor the target 1km away from the lidar system, the displacement measurement accuracy is less than 3cm and the pose angle error is less than 3°.