This PDF file contains the front matter associated with SPIE Proceedings Volume 12966, including the Title Page, Copyright information, Table of Contents, and Conference Committee information.

Phase unwrapping is a vital task applied in fringe projection profilometry (FPP). In FPP, fast-speed and high-accuracy three-dimensional (3D) imaging has been the goal. One prominent approach to achieving this objective is the dualfrequency temporal phase unwrapping method (DF-TPU). However, the highest period number for the DF-TPU approach is usually constrained to no more than 16 or 32 by inevitable phase errors, thereby limiting reconstruction precision. For single-camera FPP systems, existing deep learning-based methods capable of unwrapping high-frequency phase maps require accurate labels. This paper proposes a novel deep-learning-based phase unwrapping method for single-camera FPP systems. The inaccurate unit-period phase map is used as the weakly supervised label to guide the convergence of the unwrapping of the high-frequency phase map. The trained network can unwrap the phase map of 64 periods. The proposed approach has been validated in multiple real-world scenarios, including motion blur, isolated objects, non-uniform reflectivity, and phase discontinuity.

We propose a high-order microring filter structure based on the Silicon-On-Insulator (SOI) platform for flattening filtering. This filter structure can achieve a box-like spectral response with large free-spectral-range (FSR) and high out-of-band extinction ratio (ER), which can be widely used in various fields, especially in the wavelength division multiplexing. The filter consists of two or more rings which are connected in series, with Ti electrodes added on them for thermal tuning. According to the proposed requirements of spectral bandwidth and FSR, the physical dimensions of the high-order microring filter structure can be calculated based on theoretical formulas, and we optimize them for better performance. The relationship of the coupling coefficients of each coupling zone is calculated based on the Butterworth spectral responses. Result shows that the higher-order filter has steeper rise-fall edges and a flatter top, and can be designed to achieve a larger FSR. The FSR of the fourth-order microring filter can be designed to 78.4 nm, with ER of 12.6 dB, and the insertion loss of 1.1 dB. The fluctuations at the top of the box-like spectra caused by the fabrication deviations can be compensated by applying voltages to Ti electrodes, which can change the effective refractive index of the waveguide.

The growing demand for global communication services necessitates high bandwidth interconnects. With a gradual maturity in the spectral efficiency (SE) growth, extending optical bandwidth beyond the C + L-band is crucial for future optical network upgrades. Silicon photonics (SiPh) technology has already been practically implemented in short-reach optical interconnects and coherent communication links, due to the advantages in size, cost, and power consumption. Despite its commercial success, achieving a high extinction ratio (ER) with wide optical bandwidth in the volume production poses difficulties. In this paper, a power compensation method is proposed and investigated theoretically and experimentally. Compared to the convention method, our algorithm is still robust in a large range of extinction ratio from 10 dB to 45 dB, when the other child Mach–Zehnder modulator (MZM) is non-ideal and other noise exists. The new ER compensation method will be very useful in wide optical bandwidth transceivers.

We demonstrated the inverse design of a multifunctional silicon photonic device named wavelength demultiplexing power splitter (WDPS) which can realize dual-band (1310/1550 nm) demultiplexing and 1:1 power splitting simultaneously. We proposed a novel two-step hybrid binary-analog optimization (TH-BAO) method that combines two distinct optimization techniques: direct binary search (DBS) for binary pixel-state optimization and particle swarm optimization (PSO) for analog pixel-position optimization. Compared with the traditional DBS method, the TH-BAO method achieves comparable optimization performance with a reduction of the total simulation runs by 29.2%. The inverse-designed WDPS achieves insertion losses of 0.76 dB and 1.19 dB, as well as channel crosstalks of -17.96 dB and -11.20 dB at 1310 nm and 1550 nm, respectively. Furthermore, the dual-band functionality of our device can efficiently support the development of next-generation passive optical networks.

Integrated silicon micro-ring resonator (MRR) has been widely used as on-chip single photon source. In this paper, we introduce an approach to generate frequency-degenerate photons by bidirectionally pumping an add-drop micro-ring resonator (MRR) within a Sagnac loop configuration. This scheme facilitates the concurrent generation of photon pairs from both clockwise (CW) and counterclockwise (CCW) directions, transforming a single MRR into a dual-source system. Through CMOS fabrication techniques, we realized the proposed device and conducted measurements of key parameters including the single side count rate (SSCR), coincidence count rate (CCR), and coincidence-to-accidental coincidence ratio (CAR) for both directional outputs. Notably, the CCRs exhibit remarkable similarity between the two sources, reaching approximately 550 Hz at an waveguide power of 0.59 mW. Furthermore, we observed a differential CAR, with the CCW direction yielding a lower value compared to the CW direction, with estimated maxima of 161 and 387, respectively. These findings underscore the viability of utilizing a singular MRR as a dual-source entity. Additionally, we outline the design of an on-chip structure for heralded Hong-Ou-Mandel (HOM) interference, necessitating 4-fold coincidences.Our work holds the promise of advancing the realm of large-scale integrated quantum photonic chips.

In recent years, lithium niobate (LiNbO3) has been widely used in optical fiber communication, quantum communication, fiber optic gyroscopes and microwave photonics as an important photonic material. Lithium niobate on insulator (LNOI) has attracted much attention as an emerging photonic integrated material. Compared with traditional lithium niobate crystal materials, LNOI materials have the ability to realize miniaturized photonic devices with higher efficiency and lower energy consumption, thus showing great potential in the design and manufacture of photonic devices. However, due to the high hardness and inactive chemical properties of LNOI materials, the traditional semiconductor process cannot process its nanostructures, which limits the optimization of the key performance indicators of the device, thus hindering the further development of high-quality and miniaturized LNOI optoelectronic functional devices. In the preparation process of the LNOI waveguide, the sidewall of the ridge waveguide formed by etching is often not flat enough, which may lead to an increase in light scattering loss. To solve this problem, surface polishing technology, especially chemical mechanical polishing (CMP), has become an important method. Polishing the surface of the LNOI device by CMP can reduce the roughness of the waveguide sidewall after etching, improve the performance and power transmission efficiency of the device, and ensure that the sidewall surface is flat and smooth to achieve the minimum optical coupling loss and maximum power transmission.

Understanding and dealing with complex, large-scale, nonlinear complex systems such as those studied in the fields of bionic computing, climate change, molecular modelling, economic development, and language processing are important. In recent decades, artificial intelligence (AI) technology has played a prominent role in dealing with complex systems and problems, and the continuous improvement of the performance of various AI computing equipment is one of the decisive factors for its rapid development.

All-photon memory, holding significant potential for applications in optical communication systems and neural network computing, and developing an all-optical dual-channel fiber storage platform that achieves integrated storage and computation is challenging. In this paper, a non-volatile, high-contrast, and highly repeatable bipolar memory is demonstrated, achieved by integrating two tapered fibers with a fiber microsphere containing phase change materials(PCMs). Employing an external laser modulation technique, repeatable or randomly accessible 6-level data storage is enabled by altering the state of Ge₂Sb₂Te₅ (GST). Multi-stage writing is accomplished using a 532 nm pump laser witha10 ns pulse width and laser energy ranging from 0.423 mJ to 1.206 mJ, while a 793 nm continuous wave (CW) laser with an average power of 4 mW to 11 mW is utilized for the multi-stage reset process. Exhibiting a write response time of 75ns, a reset response time of 180 ns, and a contrast of 18 dB, the bipolar memory preliminarily realizes the synaptic weight update mechanism in the synapse of neural network systems

This article mainly discusses the applications of artificial intelligence technology in the field of space optical telescopes and the challenges it faces. With the continuous development of artificial intelligence technology, its application in the field of space optical telescope is becoming more and more extensive. Artificial intelligence technology can help space optical telescope achieve more efficient and accurate imaging and improve image quality. For example, through the analysis of image data, artificial intelligence algorithms can automatically adjust camera parameters to achieve better imaging effects; through the processing of image data, artificial intelligence algorithms can improve image quality and eliminate problems such as noise and blur. However, the application of artificial intelligence technology in the field of space optical cameras also faces some challenges. Firstly, the huge amount of data obtained by space optical cameras poses high requirements on the processing ability of artificial intelligence algorithms. Secondly, when artificial intelligence algorithms process space optical data, the computational complexity is very high, which also brings some difficulties to algorithm optimization. In view of these challenges, this article summarizes some existing solutions. For example, algorithm structure can be optimized to improve algorithm running efficiency; distributed computing and other technologies can be used to improve computing power. In addition, by introducing new hardware devices and computing architectures, we can further enhance the application effect of artificial intelligence technology in the field of space optical cameras. In summary, artificial intelligence technology has broad prospects for application in the field of space optical cameras, but it also faces some challenges. Through continuous efforts and innovation, we are confident that we can overcome these challenges and achieve the extensive application of artificial intelligence technology in the field of space optical cameras.

Due to the defects such as low resolution, lack of hierarchy, and blurred visual effects in infrared images, the accuracy of detecting infrared images is much lower than that of detecting visible images. In this study, we propose a super-resolution enhancement method for single-frame infrared images based on improved Real-ESRGAN to resolve the problem of low resolution and lack of detailed texture of infrared images so that we can improve the accuracy of object detection. We add attention mechanism based on the original network to improve the network’s ability to extract the detailed texture. In addition, we also adjust the training epochs of the generator and discriminator to accelerate the generator update and prevent the discriminator from converging too fast. We also use pooling layers for downsampling to remain the important detailed texture features and make it easier for convolution layers to extract detailed features. The experimental results show that compared with original Real-ESRGAN, when improving the resolution twice as before, our improved network reach an increment of 6.23% of PSNR and 13.9% of SSIM, and when improving the resolution four times as before, the increment is 0.95% of PSNR and 4.42% of SSIM. In object detection, we use YOLOv5 to detect super-resolution infrared images generated by our improved network and the original infrared images and reach an increment of 2.95% of mAP. These promising results confirm that our network works effectively as a method of infrared image super-resolution enhancement and improves object detection accuracy.

The rapid development of radar technology requires RF transmitters with high working frequency, broad tuning bandwidth, flexible reconfigurability, and the ability of generating large time-bandwidth product signals. In this paper, we reviewed the main architecture and research status of high-frequency, broadband, multi-format radar waveform generation technology based on optical frequency combs (OFCs) in the context of microwave photonics, and we highlighted the technical route and the main problems that exist for the purpose of practical engineering applications. The main challenges and development trends of microwave photonic radar RF front-end based on optical frequency comb in future applications are also presented.

A novel scheme with simultaneous measurement of Angle-of Arrival (AOA) and Doppler-frequency-shift (DFS) of microwave signal without direction ambiguity is proposed and demonstrated. At remote antenna unit (RAU), two received high-frequency microwave signals and a reference signal are modulated by a dual-polarization dual-drive Mach-Zehnder modulator (DPol-DDMZM). After transmission over a segment of fiber link, two signals in low-frequency bands are generated at central station (CS) through a frequency down-conversion. The DFS (including values and direction) and the non-ambiguous AOA in the range of 180° can be simultaneously calculated by monitoring the power and frequency of the two low-frequency electrical signals. The proposed structure not only improves the concealment and security of the CS but also can be extended to have multiple antenna elements in remote locations to realize multi-target detection.

In this paper, a dispersion-assisted multi-functional scheme with compact structure and high cost-effectiveness based on microwave photonics is proposed. By virtue of the amplitude regulation mechanism of intermediate frequency signal induced by fiber dispersion, combined with the optical high-precision time delay matching, the image interference and the self-interference can be eliminated simultaneously. It avoids applying electrical tuning devices with lower precision and electrical couplers with limited bandwidth or extra optical filters which is beneficial to improve the system performance and compactness. It is also compatible with optical fiber transmission, which can be combined with radio over fiber technology to bring out the advantages of high spectrum utilization, distributed configuration and low loss transmission of in-band full-duplex radio over fiber systems.

In this paper, we demonstrate an 8-bit silicon optical true-time delay line with a high resolution of 1.78 ps based on the cascaded optical switches. The total delay time is larger than 400 ps. The crosstalk of the optical switch is below -25 dB and -21 dB in the C band under cross-state and bar-state, respectively. The calculated delay loss is about 2.4 dB/ns.

We studied the influence of frequency and phase noise of semiconductor laser on the performance of photonics assisted terahertz wave system. The laser spectrum is Lorentzian intrinsic shape when only Gaussian white noise exists. The phase noise of the laser increases with the increase of the Lorentzian spectral linewidth of the laser. When considering 1 f noise, the laser frequency will be superimposed with 1 f noise. The 1 f noise in the low-frequency band will make the reconstructed laser spectrum of the whole frequency noise tend to be non-Lorentzian shape, which has a serious impact on system performance. When the system is affected by the resonance frequency, the side lobes appearing on both sides of the main peak of the laser spectrum also have a certain impact on the system performance. We simulated dual-polarization (DP) 16-ary quadrature amplitude modulation (16QAM) signal transmission up to 60Gbaud on a 50km standard single mode fiber (SSMF) at 300GHz.

The grayscale mapping of infrared images is an important research direction in the field of infrared image visualization. The grayscale mapping method of infrared images directly determines important visualization indicators such as detail preservation and overall perception of the original infrared images and can be considered as the foundation and guarantee for detail enhancement. Although the current mainstream grayscale mapping methods for infrared images can achieve good mapping results, there is still room for improvement in terms of preserving image details and enhancing image contrast. In this paper, we propose a grayscale mapping method for infrared images based on generative adversarial networks. Firstly, our discriminator adopts a unique global-local structure, which allows the network to consider both global and local losses when calculating the loss, effectively improving the image quality in local regions of the mapped image. Secondly, we introduce perceptual loss in the loss function, which ensures that the generated image and the target image have consistent features as much as possible. We conducted subjective and objective evaluations on the mapping results of our method and eight mainstream methods. The evaluation results show that our method is superior in terms of preserving image details and enhancing image contrast. Further comparison with a parameter-free tone mapping operator using generative adversarial network (TMO-Net) indicates that our method avoids problems such as target edge blur and artifacts in the mapped images, resulting in higher visual quality of the mapped images.

We demonstrate the 4-stage traveling wave photodetector (TWPD) with monolithically integrated bias circuitry network based on a silicon photonics process. A bias circuitry network comprised of inductors is integrated at the input terminal to provide the bias voltage for device while prevent the leak of the RF signal into the voltage circuitry. Experimentally, the maximum RF powers of load terminal are 8 dB higher than input end at high frequencies, validated the effectiveness of RF-choke.

Machine vision is not a mere upgrade of the specification of the current imaging devices, but rather a form of visual perception technology that involves intelligent modules in the processes of measurement, processing, and decision- making. Given the novel functionalities and features of machine vision-based intelligent detection devices, the traditional evaluation methods based on testing the physical parameters of imaging devices need further refinement and development. Taking the electroluminescence (EL) imaging in photovoltaic (PV) tests as an example, we investigate the influence of changes in dataset characteristics on the performance of object detection by combining digital image processing and deep learning methods. Features regarding to the crack-type defect datasets, such as the grayscale, contrast, shape and resolution, are controlled and adjusted based on new generated datasets from the original datasets. From the numerical experiments, some new aspects for evaluating the intelligent detection.

Radio over fiber (RoF) has become a promising alternative for 5G mobile fronthaul due to its inherent advantages of large bandwidth, low susceptibility to radio frequency interference, and high flexibility. In traditional RoF system, analog signals are transmitted through optical fiber, which is susceptible to nonlinearity and dispersion. However, digitized radio over fiber (DRoF) can effectively mitigate these defects by digitizing the analog signals. This paper explores FPGA-based real-time DRoF transceiver technology, which utilizes varying numbers of quantization bits for real-time vector quantization across different modulation formats. This approach helps prevent the reduction of system spectral efficiency caused by excessive quantization bits. The simulation results demonstrate that, after real-time vector quantization and transmission with 8 and 10 quantization bits, the EVM performance of QPSK and 16-QAM signals stabilizes at 11.8% and 6.7%, respectively, thereby confirming the effectiveness of the proposed scheme.

In recent years, there has been a surge of interest in the development of invisibility cloaks that can manipulate electromagnetic waves. This trend stems from the rapid growth of transformation optics and metamaterials technology, which offers exciting opportunities for controlling light propagation across a broad range of frequencies. One particular type of invisibility cloak that has gained significant attention is the carpet cloak. However, previous implementations of carpet cloaks have encountered limitations that hinder their performance in certain contexts. For example, quasi-conformal mapping carpet cloaks are known to produce a lateral shift in the reflected light ray, leading to undesirable effects in imaging applications. Similarly, birefringent carpet cloaks exhibit polarization dependence, which poses major challenges in situations where polarization may vary. To address these drawbacks, we propose a novel approach using an isotropic non-resonant medium and a judiciously designed conformal mapping to develop a carpet cloak. This design approach overcomes previous limitations by enabling the realization of an invisibility cloak for near-IR. Analytical calculations and numerical simulations are conducted to confirm the polarization-robust performance of this near-IR design, which offers superior performance over prior designs. Furthermore, a microwave carpet cloak design are proposed and verified to prove the universality of this approach. By developing this new approach, we hope to contribute to the advancement of cloaking technology and foster its practical applications in various fields.

In this paper, we experimentally demonstrated the transmission of 112 Gb/s PAM-4 and PAM-6 signals over standard single-mode fibers for 2 km and 10 km based on a 30GHz-EML. We have used three algorithms, of which MLSE has best BER performance. In consideration of algorithm complexity, FFE is cascaded before MLSE. For PAM-4 signal, the sensitivity is improved by 1 dB when using MLSE compared to FFE.

Increasing the layer number can improve the model performance of on-chip optical neural networks (ONNs). However, this results in larger integrated photonic chip areas due to the successive cascading of network hidden layers. We introduce a novel architecture for optical computing based on neural ordinary differential equations (ODEs) that employing optical ODE solvers to parameterize the continuous dynamics of hidden layers. The architecture comprises ONNs followed by a photonic integrator and an optical feedback loop, which can be configured to represent residual neural networks (ResNets) and implement the function of recurrent neural networks with effectively reduced chip area occupancy. For the interference-based optoelectronic nonlinear hidden layer, we demonstrate that the single hidden layer architecture can achieve approximately the same accuracy as the two-layer optical ResNets in image classification tasks. Furthermore, the architecture improves the model classification accuracy for the diffraction-based all-optical linear hidden layer. We also utilize the time-dependent dynamics property of architecture for trajectory prediction with high accuracy.

An efficient and tolerant chip-to-chip optical coupling approach employing silicon nitride grating couplers was proposed and investigated. The integration of bottom mirrors and the strategic extension of grating length led to notable results, with a peak coupling loss of -0.28 dB and a 1-dB alignment tolerance along the x-axis reaching 25.4 μm when using a grating length of 50 μm.

The waveguide branch plays an important role in integrated photonic circuits by dividing input light into two or more output lights, thereby facilitating optical power distribution and mode selection. Ordinary optical waveguides used in waveguide branches suffer from excessive optical loss and narrow branch angles, limiting their effectiveness in mode selection among other problems. Photonic crystals are constructed by arranging macroscopically homogeneous dielectric (or metallic) materials into periodic arrays, with carefully designed internal defects that provide them with frequency-selective and spatial properties. In this study, a silicon-based wide-angle waveguide branch composed of two-dimensional photonic crystals has been successfully created. The branch is capable of separating two wavelengths of light, namely 850 nm and 950 nm, by adjusting the positions of silicon cylinders in the two-dimensional photonic crystal with the purpose of optimizing optical power at different wavelengths. The silicon-based wide-angle waveguide branch is expected to be employed in multimode optical communication systems. Its utilization will contribute towards the reduction in size and complexity of integrated optical communication systems, while enhancing system reliability.

Photonic wire bonding (PWB) is an enabling technology that allows the combination of the complementary strengths of different optical integration platforms in advanced photonic multichip modules. By utilizing PWB technology, compact designs with great design flexibility can be achieved while still maintaining high performance. The principle of PWB technology relies on highly precise direct-write 3D laser lithography for the printing of freeform single-mode waveguides. The PWB single-mode waveguides are printed between the optical dies of separate optical components, enabling interconnectivity between different material platforms. An advantage of the technology is that it offers a path towards fully automated mass production of interconnects without the need for active alignment. The flexible freeform geometry of the PWB and the principle of its manufacturing can compensate for a range of challenges inherent in the coupling of optical components on a submount. PWB technology can hereby offer solutions for the typical challenges of photonic multi-chip assemblies such as offsetting the pitch error of arrays, correcting for mode field mismatches between different optical devices and amending the misalignment of optical components either via manual or pick and place errors. This publication studies a solution to realize optical interconnects between edge-emitting Indium Phosphide based laser diodes and single-mode optical fiber arrays (SMF-FA). Here, the interconnects were achieved by using the PWB technology in a Vanguard Automation Symphony 1000 suite (comprising of Vanguard’s Sonata 1000 and Reprise 1000 systems). The Vanguard Automation Symphony 1000 suite can compensate for lateral as well as vertical misalignments of up to 20µm. This enables the laser diodes as well as the SMF-FA to be assembled on a common sub-mount either manually or with standard pick and place tooling equipment. The precise locations of the coupling interfaces of laser diode waveguides and the SMF are determined with an accuracy of well below 100 nm by the detection mechanisms of the Vanguard Sonata 1000 system. Based on the detected positions of the laser and SMF interfaces, the Vanguard Brightwire3D software calculates on-the-fly the optimized low-loss trajectory of the PWB. At the same time the software also calculates the taper structures on both ends of the assembly which match the different mode fields of the laser diode waveguides and the SMF. Afterwards the freeform PWB is printed by a 2-photon-polymerization process with a femtosecond laser along the trajectory pre calculated by the Vanguard Brightwire3D software. 3D printing of the PWB is achieved in a relatively short time frame and can be fully automated with Vanguard Composer software. In this publication coupling losses of less than 2dB are demonstrated for the optical interconnects of laser diodes and SMF-FA. In general, PWB can achieve low insertion losses and have shown good reliability, e.g. meeting the GR-468-CORE standard requirements of more than 2000 hours at temperatures of 85°C and 85% relative humidity, and more than 500 thermal cycles between temperatures of -40°C to 85°C. This makes the PWB technology suitable for a wide range of applications from telecom and datacom, 3D sensing like lidar and quantum applications.

As the complexity of optoelectronic integrated circuits (OEICs) develop, the need for an accurate and efficient compatible simulation environment that supports both photonics and electronics becomes increasingly critical. This paper addresses the demand by proposing an approach that leverages Verilog-A language to build equivalent circuit models and compact models for photonic devices. Passive components, including couplers and waveguides, are modeled using compact models. Active components, such as CW lasers, are realized through the adoption of equivalent circuit models. Additionally, a depletion-type phase shifter is separated in two parts: the electrical part for parasitic parameters and the p-n junction are presented with RC components, while the optical characteristics, influenced by electrical modulation, are achieved through the use of compact models. The proposed compatible system design scheme, which consists of Verilog-A models, can be analyzed in the frequency-domain using EDA software. The simulation results demonstrate a mean absolute percentage error (MAPE) of less than 0.003% when compared to those obtained from commercial interoperable design software. Therefore, this study effectively addresses the challenge of incompatible design and simulation for OEIC, and providing strong evidence that OEIC design can be achieved in a unified EDA platform.

Dual-station cross localization method is another promising technical route to realize cooperative localization in GNSSdenied environment, requiring only two anchor node UAVs instead of at least three nodes if inter-node distance measurement based localization method is used. Traditional Angle-of Arrival (AOA) estimation techniques cannot obtain accurate and highly available angle measurements used for position calculation, for example, optical measuring equipment can provide precise angle measurements but is restricted by weather condition, array antenna can be used in all situations but its measurement error would be too large. In order to acquire precise azimuth and pitch estimations, an optical phase scanning based AOA estimation method is utilized in this work. At the remote antenna unit (RAU), one received microwave signal is applied to a phase modulator (PM), and another received microwave signal coupled with a low-frequency large-voltage sawtooth-wave signal is applied to another PM, this sawtooth-wave signal is utilized to scan the phase of the optical sideband from 0 to 2π. By transmission through a segment of fiber link, the AOA value can be measured by processing the obtained low-frequency electrical signals at the central office (CO). Experimental results demonstrate that the AOA estimation precision could be less than 2.27° when the distance between array elements is half-wavelength of the microwave (d = λ/2 0.015 @10GHz), and smaller than 0.017° if the array elements spacing is bigger than 2m for medium-sized or large-sized UAV. Then the experimental result is applied into dual station cross localization simulation structure of UAV swarm, then localization precision distribution is evaluated in different scenarios, corresponding outcomes indicate that optical phase scanning based high precision AOA estimations are beneficial to cooperative localization, and in order to acquire more accurate cooperative localization results, the positions of anchor node UAVs need to be properly adjusted.

Periodically poled lithium niobate (PPLN) is a promising platform for realizing high-speed active polarization mode conversion. Especially, the development of thin-film PPLN techniques drives related devices to lower power consumption, higher performance and more integration. However, the wavelength shifting with the temperature variation is still a problem that brings instability and impedes modulation efficiency. In this paper, we first analyzed the temperature characteristics of a well-designed z-cut polarization mode converter based on thin-film PPLN. The simulated modulation voltage is smaller than 5V. Then a temperature-insensitive device was proposed with different coating materials of negative thermo-optic coefficients. Compared to the structure without coating, the wavelength shifting decreases from 0.25nm/°C to 0.07nm/°C, in the meantime, the modulation voltage can still be kept smaller than 5V or even be reduced slightly.

The HgCdTe Photodiode is the most basic and important unit of HgCdTe IRFPA (Infra-red focal plane array) detectors, which have been widely used in the fields of security, fire protection, remote sensing and deep space detection. For HgCdTe IRFPA, the trapped charges of the HgCdTe material and the ionic charges introduced during the preparation process are the factors, other than environmental stress, that have the greatest impact on IRFPA performance. The trapped charges come from the trap energy level in the HgCdTe material, which exist during the crystal growth process and can be improved by adjusting the growth conditions, but it cannot be completely avoided. The ionic charges introduced during the process are generally concentrated at the interface and surface of the HgCdTe material, which can be reduced by process improvement, but cannot be completely avoided. In order to analyze the mechanism of multiple charges affecting the HgCdTe detector performance, a type of n+ -on-p HgCdTe Photodiode is selected as the object of this work, and the effects of the concentration and distribution of charges on the carrier distribution and energy band structure of the n+ -on-p HgCdTe are analyzed in detail. The introduction of additional net charge relative to an ideal n+ -on-p HgCdTe Photodiode leads to the aggregation or scavenging of local carriers and affects the energy band structure near the charge, creating additional potential barriers or potential wells, which is likely to cause device degradation. On this basis, the optoelectronic properties of the HgCdTe Photodiode have been investigated under infrared radiation at a wavelength of 9.5 μm, as the light I–V characteristics, the dynamic resistance–voltage characteristics, band structure and carrier density distribution. According to the results of this work, the quasi-fixed charges introduced by defects or contamination will directly affect the generation rate of photogenerated carriers and affect the I–V and R–V characteristics of the HgCdTe Photodiode, leading to phenomena such as rising dark currents, decreasing spectral response, and decreasing quantum efficiency.

1310nm long-wavelength vertical-cavity surface-emitting lasers (VCSELs) have a wide application prospect in optical data transmission over long distances, in particular for hybrid integration with silicon photonics. With the wide application of 1310nm VCSEL, the reliability requirement is becoming more and more high. In this paper, the degradation mechanism of 1310nm VCSEL is studied by accelerated stress aging experiment. The device accelerates aging for 4000 hours at 8 mA, the maximum output power decreases by 0.04 mW, and the power saturation current and V-I curve remain basically unchanged. Leakage current of the device increases and reverse bias breakdown voltage decreases. Current noise power spectral density of the device is an order of magnitude higher than before aging. In addition, the device with degraded performance is characterized by optical emission microscopy. When the device is forward biased, dark spot defect is found on the edge of the light-emitting hole of the device after burnin. The internal topography of the device is characterized by FIB-SEM, and the oxide layer warpage is found. This is due to the increase of heat inside the device and the increase of stress in oxide layer, resulting in degradation of device’s performance.

We propose a precision linewidths measurement using short-delay self-heterodyne interferometers and multiple peak-to-valley differences (MPVD). The method of MPVD of the coherent envelope to determine the laser linewidth is proved to be stable. Based on the relationship within MPVD values, the delay length and laser linewidth were calculated theoretically and via simulations. We also eliminate the effect of the broadened spectrum induced by the 1/f frequency noise and the influence of noise floor on the measurement using short-delay self-heterodyne techniques, providing MPVD that can satisfy high-precision measurements without being affected by noise. The results showed that this new method is capable of significantly improving the measurement accuracy of narrow linewidth.

We report a scheme for optical coupling between the chip and the laser off the chip by means of combining lens coupling and edge coupling. Two lenses with aspherical surfaces are employed in the lens coupling, which the first lens collimates the laser source light and the second lens focus the light. And the edge coupling is accomplished by a triple-tip spot-size converter (SSC). The simulation studies of proposed method are described in detail. Simulation results show that perfect lens coupling can be generated, and the maximum coupling efficiency is 87.7%. For the edge coupling, the triple-tip SSC contains one central taper and two side waveguides that locate at both sides of the central taper equidistantly. After calculating by the 3-D finite-difference time-domain (FDTD), low coupling loss of 0.53 dB can be realized. The proposed scheme with simple structure, low cost, and high coupling efficiency has potential applications on coherent communication systems, the high-speed optical module, and integrated optical communication systems.

A method of dual-broadband signal generation based on the same optoelectronic oscillator with different tuning mechanisms is proposed in this paper. The structure includes a compatible dual-passband microwave photonic filter based on stimulated Brillouin Scattering effect (SBS) and phase-shifted Bragg fiber (PS-FBG). By implementing the Fourier-domain mode-locked mechanism, the proposed optoelectronic oscillator can simultaneously generate signals in different frequency bands with adjustable center frequency and bandwidth. The effectiveness of the proposed method is verified by experiments. Oscillating signals with bandwidth of 600 MHz and center frequencies at 5 GHz and 6 GHz are generated.

Due to the difficulty in simulation for massive array microwave photonics systems, a distributed cross-domain parallel simulation method is proposed in this paper. Firstly, the link parallel computation across system structural domains is achieved based on the independent transmission between channels for each array element in the microwave photonics system. Secondly, the data parallel computing across time domains is achieved by utilizing the relative independence between the pre and post processing times. Furthermore, assisting by a static load balancing strategy to allocate computational resources, these two approaches are effectively combined to achieve high efficient simulation of the microwave photonics system, which addresses the issue of long simulation time caused by large amounts of data and models. For a microwave photonics system with a 64-array and more than 400 models, this technique reduces the simulation time from 39 hours to 23 minutes, resulting in a simulation efficiency improvement of two orders of magnitude. This advancement holds the potential to significantly shorten the development cycle of microwave photonics engineering prototypes.

An optimization method to improve the spurious free dynamic range (SFDR) of photonic sampling without sacrificing the signal-to-noise and distortion ratio (SINAD) is proposed and experimentally demonstrated. It is realized by managing the chirp in the generated ultrashort optical pulse train by simply changing the group velocity dispersion (GVD) of the dispersion compensation module (DCM) in the cavity-less ultra-short optical pulse source. In the simulation, the SFDRs of the photonic sampling for the input signals in the frequency range of 0.1 GHz to 40.1 GHz are significantly improved with residual linear chirp in the optical pulse train compared with the situation that the chirp is completely compensated. In the experiment, a 10.1 GHz single-tone microwave signal is sampled and the SFDR is improved by 10.95 dB owing to the residual chirp in the optical pulse train. In addition, the SINAD is improved by 2.76 dB even though the power of the fundamental frequency signal is slightly reduced. The proposed scheme can also be applied to photonic sampling ADCs based on other optical pulse sources, which is favorable for alleviating the limitation from the nonlinearity of the electro-optic amplitude modulator.

The Whispering Gallery Modes Resonators(WGMR) are specific electromagnetic modes confined in circular resonators, where light beams are guided along the circumferences of the resonators with very small losses by total internal reflections. WGMR is the core component of optical filters, ultra-narrow linewidth laser, microwave photoelectric oscillator, optical frequency comb, rubidium atomic clock etc and high-precision sensors. Q factor is defined as the ratio of the total energy of photons in WGMR to the loss lost in one propagation cycle. The energy loss of WGMR is discussed. The cutting dynamics model diagram was established based the cutting force model for turning. The key technology affecting WGMR’s Q factor was the fabrication process. The manufacture of ultra-high Q factor WGMR are realized through rough machining, ultra-precision turning and precision polishing. The processing of ultra-precision turning and precision polishing is reported in the paper. The important fabrication processing of WGMR is ultra-precision turning. The test results show that the Q value of WGMR is 1.5×10 9@1550nm by Q value measurement system. The surface roughness and shape error of WGMR are 0.6nm (Ra value) and 5.3nm (PV value) respectively, measured by white light interferometer.

Performance improvement of microwave photonic frequency down-conversion link is investigated both in simulation and experiment. To achieve maximum signal-to-noise ratio, a dual-drive Mach-Zehnder modulator (DDMZM) biased at the minimum transmission point and an erbium-doped fiber amplifier (EDFA) are employed to boost the power of the first-order modulation sidebands. The power spectrum density of the output noise is limited by the thermal noise and relative intensity noise (RIN) with low and high gain of EDFA, respectively. Correspondingly, the output noise figure (NF) decreases rapidly in the case of a thermal-noise-dominant noise floor and remains stable in RIN-dominant noise floor situations. When the local oscillation (LO) and the radio frequency (RF) signal powers are set to be 10 dBm and the input optical power before photodetector (PD) is amplified to be 20 dBm, an intermediate frequency signal with a high power of 10.8 dBm is obtained. When LO/RF power is set to be 0 dBm and the amplified optical power entering the PD is 16 dBm, the NF of the proposed link is measured as 31.5 dB, which is 14 dB lower than the link without EDFA, while the conversion gain has improved to 12 dB with an increase of 57.2 dB. Finally, the spurious-free dynamic range of the link is measured as 113.3 dB·Hz ^2/3 when the power of the LO signal is 0 dBm and the gain of EDFA is 18 dB.

We proposed a compact and efficient TE-pass polarizer in silicon-on-insulator(SOI) platform, which consist of polarization beam splitters (PBS), polarization rotators (PR) and polarization splitter-rotators (PSR) . The transmission of transverse magnetic(TM) and transverse electric(TE) modes can reach 88.9% and 98.6% at the wavelength of 1550 nm, and the loss is less than 1.6 dB. The PR shows a high polarization conversion efficiency (PCE) of 98.2% within a total device length of about 7.2 um. The device is much smaller than the general level of polarization devices, which can fulfill the polarization control in the integrated optical gyroscope (IOG).

A multi-beam optical beamforming network with low loss is proposed based on the integrated arrayed waveguide grating (AWG). By using the diffraction effect of AWG, the optical multiple beamforming architecture is reduced significantly since the true time delays for different wavelength can be realized simply by employing only one module. The 3 channels of 88 AWG for dense wavelength division multiplexing (WDM) with low loss is fabricated successfully, and then the integrated optical multiple beamformer for eight-element is demonstrated. Experimental results show that the insertion loss for each beamformer is reduced to 5.64 dB. Furthermore, the number of simultaneous beamformer achieves to be 4, which could cover the airspace from -36° to 18°.

In recent years, quantum communication has received extensive attention due to its high security of transmitted information. Quantum key distribution (QKD), an important branch of quantum information, is developing rapidly and has been gradually moving toward practicality and networking. The use of phase coding through fiber optic channels is the basis for the implementation of QKD systems. In QKD systems, electro-optical modulation techniques are mainly used to change the photon phase through phase modulators to realize the phase coding scheme. Among them, lithium niobate is a common material for making phase modulators in QKD systems. Lithium niobate (LN) crystals are an optical material with excellent acousto-optical and electro-optical properties. It has good physical and chemical stability, a wide optical low-loss window, a large electro-optical coefficient and an excellent second-order nonlinear effect. It has a wide range of applications in high-speed electro-optical tuning, holographic storage, nonlinear frequency conversion, etc. Thin-film lithium niobate (LNOI), as a new integrated optical material, can well combine the excellent electro-optical, acousto-optical and nonlinear properties of the material with a compact optical waveguide. It also has the advantages of a small waveguide cross-section size, high electric field density, strong nonlinear effect, low half-wave voltage length product, and small size. It has significant advantages in the integration of optoelectronic devices. In the phase-coded QKD system, the coding object of the information is the phase of the optical signal. The polarization state of the optical signal can have a serious impact on the system. The phase-encoded QKD system based on the Faraday-Michaelson interference loop is able to self-compensate for the polarization variations in the system to remove the relevant effects of polarization variations on the QKD system. The application of a phase modulator based on thin film lithium niobate preparation in quantum key distribution can effectively enhance the rate of the quantum key distribution system. However, there is still a need to study the transmission and modulation characteristics of LNOI waveguides on polarized optical signals. In this paper, we develop a phase modulator based on thin-film lithium niobate for high-speed QKD systems. Simulation and analysis of the polarization mode of the optical signal transmitted in the optical waveguide. Test and study the transmission loss and modulation efficiency difference of a thin-film lithium niobate optical waveguide for TE and TM polarization state optical signals. To build a test system for application to the measurement and modulation test of the polarization state of optical signals in a high-speed phase-encoded QKD system.

We propose and analyze a novel compact modulator on a silicon-on-insulator (SOI) waveguide with the modulation mechanisms of free-carrier plasma dispersion effect. The free carriers are produced by two-photon absorption (TPA) effect, and to understand the change process of free carriers in the silicon, we carried out theoretical simulation and calculation on it, which based on the coupling wave propagation equation and boundary conditions. Free carriers density of 2.5×10¹⁷/cm3 could be obtained under pump power of 50 mW and the refractive index change Δn of 9.24×10^-4 are achieved. The result indicates that TPA-induced free carriers could be effectively change the effective refractive index of the silicon and further realize the modulation application in silicon modulators.

InP-based uni photodiode(UTC-PD) array consisting of four photodiodes, power combiner and a monolithically integrated bias circuit using a 1/4-wavelength microstrip is presented. To increase the upper limit of power output, four identical UTC-PDs were monolithically integrated along with T-junctions to combine the power from the four PDs. Each single photodiode exhibits at least -6dBm at 110GHz, and the array was designed to produce at least 1mW in the terahertz frequency band with photocurrent of around 25mA per PD and bias voltage of -3V. The circuit has been fabricated on a 12µm-thick InP substrate, and is flipped on a 50μm-thick AlN-based coplanar waveguide circuit for test.

Optical phased array (OPA) has been widely employed across various applications, including light detection and ranging. Nevertheless, OPA faces significant limitations, such as excessive power consumption, complex control systems, and challenging packaging formats, which hinder its further development. Focal plane arrays (FPAs) have garnered increasing attention due to their absence of these drawbacks. However, FPAs currently face a dilemma as their ranging performance fails to meet application requirements. To address this issue, this paper presents a novel structure featuring small-scale receiving array and high directional antenna design. Utilizing this chip, we showcase a scanning range of 5.98° and a coherent detection capability of 6 meters.

A filling material based on a similar refractive index with SiN is designed as the mode converter for thin film lithium niobate (TFLN). Such a design can realize an output mode field compatible with different sizes ranging from 3.5 um-9.2 um. The double-layer mode converter core with SiN has a similar height as the ridge waveguide of TFLN, which is helpful to increase the conversion efficiency. An overall coupling loss of less than 0.6 dB was achieved theoretically at 1310 nm for both modes. The proposed scheme avoids the disadvantage of high reflection when the inclined TFLN section result from dry-etching is directly used as the coupling end face and can improve the performance of integrated TFLN electro-optic modulation on the chip level. Three-dimensional simulation results show that the designed structure is insensitive to fabrication tolerance, which provides a feasible solution for reducing the volume of integrated devices, increasing overall performance and high-density integration.

Dielectric metasurfaces designed based on bound states in the continuum (BICs) exhibit extremely high quality factors, high sensitivity, and low loss, making them suitable for surface-enhanced infrared absorption and refractive index sensing. It has been demonstrated in some literatures that the symmetric property of the elliptical unit cell of the BICs mode can be broken by increasing the rotation angle, thereby achieving high-quality factor resonant structures. By varying the rotation angle, major and minor axis lengths of the ellipse, and the period of the bilayer elliptical unit, unique resonant properties can be achieved even without considering the existence of BIC. In this work, we have designed a deep learning-based transmission spectrum prediction network by combining the parameters of the elliptical unit. This network can replace traditional electromagnetic simulation calculations to quickly obtain the transmission spectrum of the target structure. While simulating 3000 sets of elliptical unit using the finite-difference time-domain method on an i7- 12700H processor requires 14 hours, using our neural network yields a transmission spectrum with prediction accuracy better than 10^-3 in less than 6.9 seconds, significantly improving the design efficiency. The network takes the image of metasurface unit as input data and couples the scaling factor that affects the period of the metasurface into the image data, making it possible to train and predict the spectra with different structures and ratios, effectively improving the generalization ability of the network.

In order to avoid an unacceptable in-band S21 flatness in the multi-channel broadband microwave photonic system design, this paper established a relationship between the S21 flatness of a muti-channel broadband link and amplitude and delay errors. According to the deduction, the effect of amplitude error on S21 flatness is much less than that of delay error. And the maximum in-band S21 difference is expressed into a function of the muti-channel delay inconsistency. According to the derivation, for ensuring the in-band S21 flatness of a multi-channel microwave photonic system is less than 3dB, the delay error must be less than 13.89ps. Experimental results and simulation results demonstrate the reliability of the derivation relationship between the in-band S21 flatness and amplitude and delay errors.

A frontside-illuminated single photon avalanche diode based on 110 nm CMOS process is proposed in this paper. The P-well/Deep N well multiplication junction is adopted to enhance the near infrared absorption. To reduce the dark count rate, only the well guard ring is used in the photosensitive region to avoid the defects caused by shallow trench isolation process. The SPAD achieves a DCR of 3.2 cps/μm2 at room temperature, a peak photon detection probability over 40% at 500 nm, and over 3.6% at 905 nm with 2.0 V excess bias voltage. The device we designed is very suitable for LiDAR applications.

In view of the demand for multi-format modulated signals from multifunctional radar, a reconfigurable pulse compression signal generation scheme based on dual-parallel Quadrature Phase Shift Keying modulator (DP-QPSK) is proposed. In this scheme, the phase coded signal and linear frequency modulation signal can be switched by changing the encoding format of the input electrical signal without altering the link structure. To generate pulse signals with flexible and tunable carrier frequencies, optical filtering is not employed in this scheme. Simulation results demonstrate that the proposed scheme successfully achieves the switching between phase-encoded signals and linear frequency modulation signals, and the Peak-to-side Ratios (PSR) and Pulse Compression Ratios (PCR) of the generated signals closely approximate.

To overcome the limitation of the phased array (PA), the frequency diverse array (FDA) was introduced that can provide precise scanning in distance and angle domains. Utilizing microwave photonics (MWP), a FDA signal generation method based on dual optical frequency combs (OFCs) is proposed in this paper. The scheme successfully generates a five-channel FDA signal with a center frequency of 8 GHz and a frequency offset of 1 MHz. With a power flatness below 2.5 dB and a spurious suppression ratio exceed 27.2 dB, the method demonstrates ability to generate high-quality FDA signal and realize beam forming and scanning. Additionally, the radiation pattern shows a distinct “S” shape, influenced by distance and angle.

A microwave photonic (MWP) pulse radar system for high-resolution target detection is proposed and experimentally demonstrated in this article. In the transmitter, a pulsed linearly-frequency-modulated (LFM) wave is generated based on optical frequency operation module (OFOM), which can generate LFM waves with ultra-flexibly tunable center frequency. In the receiver, optical-domain down-conversion is employed to convert the incoming echo to an intermediate frequency signal by a microwave photonic frequency mixer, which can free the receiver from high-speed ADC and provide an excellent wideband processing. Experimentally, a Ku-band pulsed LFM wave with a bandwidth of 840 MHz is generated and received through self-closed-loop and target detection test by the constructed system. The performance verifies that the proposed pulsed MWP radar has the potential of supporting high-resolution detection and recognition of distant targets.

A compact silicon-based four-port coarse wavelength-division multiplexer (CWDM) with a footprint of 200×200 μm2 and an insertion loss of ~2dB is demonstrated. This configuration can support each laser power of over 100mW without inducing silicon nonlinear effects. The design eliminates the need for interference in multiplexing different wavelength channels, resulting in significant fabrication tolerance and eliminating the requirement for phase shifters. The crucial components, such as power splitters/combiners and crossings, are designed and optimized using genetic-algorithm-based deep neural network (GDNN) inverse design methodologies to achieve minimal loss and broad bandwidth.

The phase shifter is the key component for phased array systems. However, the previously proposed microwave photonic phase shifter always has a complex structure and is difficult to operate. In this paper, a simple microwave photonic phase shifter based on waveplate rotation is proposed. It is composed of a cascaded quarter-wave plate (QWP) and a half-wave plate (HWP), the QWP is used to control the amplitude and HWP is adopted to control the phase of the microwave signal. By fixing the rotating angle of QWP, the phase of the microwave signal will be linearly changed with the rotation angle of the HWP. Compared with other schemes, the proposed method is easy to use. A proof-of-concept experiment is performed. Experimental result shows that a wide phase shift range of 542.38° can be realized and the power fluctuation can be within 0.47 dB. For a 360° phase shift, the power fluctuation can be maintained within 0.33 dB.

With the integration and design complexity of silicon photonics chips raises, the cost of manual routing based on design experience has significantly increased, which brings urgent demand for auto routing methods. This article proposed a global auto routing method for single waveguide connection in photonic integrated circuits. The method firstly partitions all routable areas into non-uniform grids based on the layout placement information and converts the grids into an undirected graph with respect to the path length and congestion of waveguides between adjacent grids. Next, an improved A* algorithm runs in the graph to find a best path between two optical ports, with additional penalties on every bend in the path, beyond cumulative lengths and congestion penalties. Heuristic terms have also been added to increase the convergence speed of the algorithm. The optimized path in graph is further used to perform multiple adhesive lines for waveguide control points generation. Then, three types of 90-degree bends and two types of S-bend according to specific spacing and angle relationships between control points are applied to generate waveguide trace. This method can avoid collisions between target waveguide and other devices on the layout, optimize the global path length, minimize the number of waveguide’s bends, prevent waveguide congestion in specific spaces from exceeding the set threshold, and at the same time generate the optimal waveguide and insert it into layout.

A novel approach to suppressing the time-delayed signature (TDS) in chaotic signals from the semiconductor laser (SCL) is proposed and experimentally demonstrated based on the optoelectronic hybrid feedback. Through combining the distributed feedback in the chirped fiber Bragg grating (CFBG) with the nonlinear optoelectronic feedback provided by the microwave photonic link, a low-TDS chaotic oscillation is successfully built up in the SCI cavity. Thanks to the employment of the nonlinear optoelectronic feedback, the proposed scheme can generate low TDS chaotic signals by using a CFBG with a much smaller grating dispersion coefficient of about 22.3 ps/nm compared with the scheme relying solely on the distributed feedback (i.e., 2000 ps/nm). In addition, different from the broadband optoelectronic oscillator, there is almost no stringent requirement of the extremely high net gain of the microwave photonic link. In the proof-of-concept experiment, the chaotic signal generated by the proposed scheme has a much lower TDS of about 0.05, a higher PE of 0.9978, and a better spectral flatness of 8 dB, compared with the conventional distributed feedback laser.

A monolithic integrated injection-locked distributed feedback laser based on reconstruction equivalent chirp technique is proposed. It is proved that the optical injection-locked enhances the quality of optical frequency comb, which can realize the wide range adjustment of different comb spacing within 2GHz-10GHz. The 10-db spectral width of the OFC with different comb spacing is between 66-119GHz. CNR up to 45dB.

A microwave photonic image rejection mixer (IRM) with large operation bandwidth and high image rejection ratio (IRR) is proposed and demonstrated based on a single dual-drive Mach-Zehnder modulator (DDMZM) and a wavelength division multiplexer (WDM). In the proposed scheme, the two radio-frequency (RF) ports of the DDMZM is driven by a RF signal and a local oscillator (LO) signal, respectively. Two intermediate-frequency (IF) signals are generated by beating the upper sideband and the lower sideband of the DDMZM, respectively. Then, the two IF signals are injected into an electrical 90° hybrid coupler (HC) to realize image rejection. In the experiment, a stable IRR above 60 dB in the RF signal frequency range of 10 GHz to 40 GHz and with a 1- GHz fixed IF signal frequency is achieved. When the IF signal frequency is changed from 1 GHz to 6 GHz and the LO signal is with a fixed frequency of 29 GHz, the IRR can also large than 60 dB.

In this paper, we develop and demonstrate a proof-of-principle OEO, which features ultra-low phase noise in a Ka frequency band. The prototype of the whole OEO is in a cylindrical form. The optical fibers are wound on the outside, while all the optoelectronic devices are in the center. The fiber transmission noise is suppressed via phase modulation for the power redistribution. The spur-level improvement and steady state operation is guaranteed by dual-loop structure with 8.7 km and 11.6 km fiber spools, respectively. The optical power loss is reduced by the dual-output electro-optical intensity modulator (DEOM) instead of another 50:50 optical coupler. The noise floor for the fiber link from laser intensity and phase noises is suppressed by the balanced photodetector (PD) with specialized working conditions. Performance is investigated in detail. The OEO operates at the frequency of 30 GHz with the spur suppression of 74.6 dBc. The phase noise of -130.7 dBc/Hz (-149.1 dBc/Hz) @1 kHz (10 kHz), respectively, are achieved. The spectral purity is much higher than the current commercial signal source and equipment. Further, the developed OEO is applied to the frequency conversion. The RF signal, to be converted with a frequency of 7 GHz, is coupled into the OEO. Each beat results with OEO are observed clearly. All these results show that OEO has broad prospects in high precision infrastructure and projects.

Carrier leakage in the transmission spectrum can significantly impair the performance of coherent optical communication systems. Though a common occurrence, the underlying causes remain ambiguous and not clearly analyzed. This paper embarks on a systematic exploration of the sources of carrier leakage, proposing resolutions across different dimensions. These insights aim to pave the way for enhancing system performance, thereby contributing to the burgeoning field of coherent optical communications.

The insertion loss of the optical true time delay line is mainly composed of optical waveguide transmission loss, optical switch loss and input-output coupling loss. When the insertion loss is large, the optical power needs to be compensated by the optical amplifier. In order to realize bidirectional optical amplification of true time delay line and ensure high consistency of bidirectional link performance, a new scheme of optical delay line based on unidirectional optical amplifier multiplexing is proposed. Different from the traditional bidirectional delay line that requires one optical amplifier in each of the uplink and downlink, the scheme proposed in this paper reuses the same optical amplifier in the two-way links. As a consequence, the number of devices employed is halved, and the pressures on circuit layout, power consumption and heat dissipation are also reduced. Furthermore, it solves the problem of large performance differences in insertion loss, gain, and optical noise, etc. caused by the opposite experience of optical bidirectional transmission in traditional links. In addition, the scheme in this paper also avoids the problem of performance degradation that may be caused by the use of bidirectional optical amplifiers and other devices with high complexity. In a typical circuit of two-stage cascaded optical delay line, the optical amplifier is placed between the two stages. Through the scheme of unidirectional optical amplifier multiplexing, the bidirectional differences of insertion loss, gain of the optical amplifier and the optical noise are reduced by about 4dB.

In response to the problem of high noise figure in microwave photonics links, a high-efficiency, low-noise T-shaped matching resonant enhanced microwave photonics receiving chip is established. Compared with traditional broadband resistor matching, this chip can achieve a 9 dB increase in power between 8-10 GHz, and the noise figure of the microwave photonic links is reduced from 35.2 dB@9GHz to 31.6 dB@9GHz. The noise figure has been optimized by 3.6 dB, thereby achieving efficient conversion between light wave and microwave energy.

Photonic RF direct sampling with multiple electronic analog-to-digital converters (ADCs) for time-interleaved quantization is an effective technique for digitally receiving RF signals. Since band-pass sampling is inevitable for RF direct sampling, the required total sampling rate of a photonic ADC system could be remarkably higher than the low-pass Nyquist sampling rate given by twice the bandwidth of input signal. In order to reduce the sampling rate and thus boost the efficiency of the photonic ADC hardware, we propose a method of introducing time errors between time-interleaved channels with post correcting algorithm in frequency domain. Simulation is performed based on a photonic ADC consisting of multiple 2-GSa/s sub-ADCs. A wideband signal covering the frequency band from ~4 GHz to ~9 GHz is employed as the input RF signal for direct sampling in the simulation. Reduction of the total sampling rate from the band-pass Nyquist rate of 18 GSa/s to the low-pass Nyquist rate of 10 GSa/s is achieved, through which the advantage of the proposed method is verified.

An active mode-locking optoelectronic oscillator (OEO) based on an electric mixer is proposed and experimentally demonstrated. In this scheme, the external drive signal is injected into the intermediate frequency (IF) port of the electric mixer to achieve the periodic loss modulation of the OEO cavity. Once the frequency of the drive signal is set to be equal to an integer multiple of the free spectral range (FSR) of the OEO, the phase between the longitudinal modes can be locked to generate the stable multi-tone microwave combs in the OEO cavity, which are coherently superimposed in the time domain to form the short microwave pulse signal with a repetition frequency equal to the frequency of the drive signal. In the experiment, the fundamental mode-locking and 50th -order mode-locking are realized in the proposed active mode-locking OEO, where the microwave pulse signals with the carrier frequency of 10 GHz, repetition rates of 98 kHz and 4.9 MHz, are generated. The phase noise at a frequency offset of 100 Hz is measured to be -94 dBc/Hz and -103 dBc/Hz for those two cases. Compared to the free-running OEO, the phase noise at 100 Hz frequency offset is reduced by 11 dB and 20 dB, respectively.

Applying various functional materials to silicon to enhance the functionality of silicon photonics is a potential solution for silicon photonics platform under the requirement of CMOS compatibility. In this paper, two LN heterogeneous integration platforms have been proposed. One is the integration of LN film with a 220 nm top silicon SOI platform, in which the simulated results demonstrate that the designed modulator has a low half wave-voltage length product of 2.27 V·cm. And the other is the integration of LN film with a 400 nm top silicon nitride on insulator platform, in which the the proposed device achieves a VpiL of 2.58 V·cm and a 3-dB bandwidth of ~130 GHz with 7-mm long modulation region is verified by simulation.

High-quality factor(Q) waveguide ring resonators (WRRs) are essential components of the resonator-integrated optical gyroscope (RIOG). The single-mode and multimode silicon oxynitride (SION) WRRs are investigated in this paper. The optical, resonance, and polarization properties are characterized. The intrinsic Q of single-mode WRR is 8.25×105 . And the multimode WRR achieves an intrinsic Q of 1.02×10⁶ , corresponding to the waveguide propagation loss of 0.3 dB/cm. The designed multimode SION WRRs can be used as a critical sensing element to enhance RIOG performance further. Keywords: High Q ring resonators, Multi-mode SION waveguide, Integrated optics.

We propose a two-dimensional beam scanner based on hitless microring switch array. By changing the resonance state of the microring interference between different wavelength selection switches is avoided. The operation complexity of the beam steering FPA chip based on the microring switch array is reduced.

Demonstration of LFMCW radar system by hardware-in-the-loop simulation based on tunable microwave photonics generated radar signal and electric receiving link is presented. The seed signal is given by Direct Digital Synthesis (DDS) with tunable signal wave forms, which can afford the system with different time-width and bandwidth. The microwave photonics electro-optic modulation and photoelectric transformation system turns the seed signal with low-frequency and narrow-bandwidth into the signal by 4 times at frequency carried on the laser through the dualparallel electro-optic modulator, and obtain the high-frequency and broadband radar signal after the photoelectric detector, and then transmit the radar signal into the input port of the hardware-in-the-loop simulator with time delay function. The radar signal is set as different delay corresponding to different transmission distance, such as 1km, 2km, and other distances. After the hardware-in-the-loop simulator with some distance delay, the signal is transmitted from the output ort of the hardware-in-the-loop simulator into the receiver. In the receiving link, the electric de-chirping method is carried out to down convert the radar echo signal. After electric ADC, the ranging data is processed. Two typical wave forms, such as the wave form with 1GHz bandwidth and 2ms pulse width, and the other wave form with 2GHz bandwidth and 8ms are operated through the system respectively. The demonstration by hardware-in-the-loop simulation has been given, and the experimental results show that the range of frequency modulated continuous wave radar based on this system can reach 5 km.

A 16-channel optical transmitter chip with a digital transmission capacity up to 1.6 Tb/s has been demonstrated. In this chip, a 16-wavelength III–V DFB laser array (MLA), a silicon Mach-Zehnder interferometer (MZI) modulator array and a 16-channel fiber array are hybrid integrated by photonic wire bonding (PWB) technique. The MLA based on reconstruction-equivalent-chirp (REC) technique proves a good wavelength spacing uniformity of all wavelengths. Each unit laser with 1.2 mm cavity length in the MLA exhibits good single-longitudinal-mode operation with the output power over 18 dBm at an injection current of 300 mA. Spectral measurements show the channels coincide well with the designed 200 GHz spacing, with wavelength deviations within a range of ±0.2 nm. Based on PWB technique, three chips mentioned above are integrated optically on one Wu-Cu substrate as a 16-channel optical transmitter. The largest output power of optical transmitter is 1.5 mW and all channels still keep good single mode outputs after PWB integration. The tested modulation speed of each channel is up to 100 Gb/s, which implies the total transmission capacity of this device is 1.6 Tb/s.

We build three different schemes for 800G ICR (integrated coherent receiver) S21 testing. By analyzing the accuracy, repeatability, efficiency, dynamic range etc., we can get their advantages and disadvantages, which will guide our choice for proper 800G ICR testing.

High-accuracy spatial distance measurement is essential to scientific research and equipment manufacturing, and industrial production control. To meet the demand of high-performance measurement, a distance measurement approach based on microwave photonic technology is proposed, in which two electro-optic modulators are connected in series. The first one is employed for performing transmission signal and the second is used for modulating echo signal, forming a vector superposition of microwave signal. By microwave frequency sweeping, the measured distance can be resolved from the microwave amplitude spectrum. To verify the performance of the proposed approach, a proof-of-concept experiment is carried out. The measurement results show that an accuracy of ±0.5 mm is obtained.

We propose a complex-valued matrix-vector multiplication method in this work, which make full use of the amplitude and phase information of input signal light and weight matrix. We demonstrate our computing method theoretically and experimentally. In order to verify our theory, a photonic chip is designed and used to setup a photonic neural network system, and the corresponding accuracies reach more than 90%.

In this paper, two kinds of ultra-low temperature-sensitivity optical cables are proposed and demonstrated. Negative expansion coefficient material of liquid crystal polymer (LCP) is tightly integrated with conventional optical fibers through precise extrusion process. Performance investigations were conducted in detail. The temperature delay coefficients for the proposed optical cable products are within the range of 2.2～11.2 ps/km/℃ and -4.1～2.0 ps/km/℃ for the ambient temperature varied from -40 ℃ to +55 ℃. Compared with the conventional one, the attained performance improvement of an order of magnitude is quite attractive. To further verify the phase consistence in large-scale microwave photonics distribution, a microwave frequency distribution system is established. The phase difference is below 0.25° for a 10 GHz frequency signal in a normal room condition, while the absolute phase for each unit fluctuates over 1.6°. The proposed ultra-low temperature-sensitivity optical cables will promote diverse advances in large-scale distributed scientific facilities with high performance and simple system complexity.

Convolutional neural network (CNN) has attracted widespread attention in image feature extraction and speech recognition owing to greatly reducing the complexity of model parameters and the number of weights, but it cannot be separated from the support of hardware accelerator. The limitations of electronic devices in terms of power, speed, and size make it difficult for current electron accelerators to meet the computational power requirements of future large-scale convolution operations. Here, we proposed a photonic vector architecture. This structure combines time, space and wavelength, and the non-volatile phase change material and the integrated microcomb form an optical matrix multiplier to realize memory calculation, thus reducing the energy consumption of reading weight data. The tooth spacing of the integrated microcomb is more than 100 GHz, and the microcomb coverage is from 1510 nm to 1610 nm. Finally, we replace the weight values in the CNN with the optimal weight values that the optics can achieve. The final recognition accuracy reached 97.04%, which is comparable to the efficiency of the first electronic equipment. Our results could be helpful for the development of non-volatile and ultra-fast optical neural network (ONN) with feathers of low energy consumption and high integration.

Deep reinforcement learning (DRL) has been introduced in routing, modulation and spectrum assignment (RMSA) of the elastic optical networks. Since the DRL agent’s learning is based on the state it observes and the reward it receives, key information should be embedded in the state and the reward. In previous studies, the observed and feedback information is limited. In this paper, we propose a busyness level-based DRL method for the RMSA of the elastic optical networks. Since the busyness of the links or transmission paths highly affects the performance, we believe the busyness information should be perceived by the agent to learn a good RMSA policy. Specifically, we define two indicators to quantify busyness level, and then combine these two indicators into the design of reward and state. Simulation results show that our approach works better than the case that busyness is not

Reservoir computing (RC) is a computational framework for information processing based on neural network. It can be implemented with different physical platforms, principally, electronic architectures and photonic architectures. Photonic RC shows potential path to ultra-fast and efficient processing beyond the traditional Turing-von Neumann computer architecture. Typical photonics RC consider specifically a semiconductor laser (SL) with delayed feedback as reservoir substrate. Basically, the SL is a kind of type B laser, needing enough long delay feedback for the high dimensional chaos generation and for the RC mapping. But on the other hand, long delay feedback leads to the setup big size, being nonconductive of integration implement and stable operation performance in real world. To solve the problem of a huge size, we propose a new photonics RC scheme that using chaotic SL hybrid with Si3N4 micro-resonator, which works as the storage layer and feedback loop. The Si3N4 micro-resonator could help SL producing high-dimensional chaos and reaching high-complexity RC. Meanwhile, the size of Si3N4 micro-resonator is highly compressed at the level of ten micrometers, thereby realizing a size compression of over ten times than that of typical photonics RC setup. In our experiment, we make the free spectrum range (FSR) of micro-resonator is 35GHz, reaching the nonlinear frequency of SL. Then, with careful operation, two-mode mixing chaos can be realized, being very conductive for the photonics RC applications. These results are conducive for the development of on-chip photonic RC.

We propose and demonstrate a sub-gigahertz bandwidth photonic differentiator employing the self-induced optical modulation effect in a silicon-on-insulator micro-ring resonator. The all-passive DIFF is controlled through an all-optical pump-probe scheme. Input Gaussian-like pulses with 7.5ns pulse width are differentiated at high processing accuracy. A semi-analytical model that agrees with the experimental results is also derived. The DIFF’s energy efficiency is higher than 45%, far surpassing all previously reported schemes for sub-gigahertz bandwidth pulses. Our scheme expands the application potential of photonic DIFFs.

In order to meet the demand of broadband signal generating, this paper proposed a time-domain sythisis method with low sampling Digithal-to-Analog Converter. This method takes the anlog signal generated by low sampling DACs to the optical sampling pulses by Mach-Zender moduator with optical carriers of different vwavelength. And then, the multichannel optical pulse with modudlated data are delayed to forms into data sequences with setted interval. Last the optical pulses are combined into a sequence by WDM multiplexer and generated broadband microwave by a photodetectors. An experiment has conducted to verify the effectiveness of the method.

The edge coupler holds paramount importance as a link bridging optical signals between fibers and silicon-based optoelectronic chips. It surpasses the grating coupler in terms of elevated coupling efficiency, diminished polarization sensitivity, and an expanded bandwidth. However, designing a low-loss silicon edge coupler with a broader minimum width, especially for the O-band, presents substantial obstacles. To surmount these challenges, a silicon nitride (SiN)- assisted double-etching silicon structure with a 180 nm minimum width is adopted in this work. This innovation capitalizes on the double-etching silicon taper, propelling the SiN layer's height to 1.6 μm relative to the bottom of silicon waveguide, resulting in pronounced reduction of silicon leakage loss. By meticulously implementing coupled mode theory, an exceptional coupling efficiency exceeding 0.95 is achieved for both TE and TM polarizations at 1310 nm, facilitating the seamless transition of light from SiN to the thinner silicon waveguide. Further enhancements in curbing silicon leakage loss and shortening device length are achieved through mode analysis-driven designs for both the SiN and silicon taper. Ultimately, these intricate designs culminate in an edge coupler boasting a 180 nm minimum width, with minimal losses of approximately 0.7/1.5 dB for TE/TM polarization and a 0.5-dB bandwidth of around 100 nm within the O band, as demonstrated through simulation while interfacing with standard single-mode fibers.

A photonic-assisted approach for instantaneous frequency measurement (IFM) based on frequency-power mapping is proposed and demonstrated. The principle of the technique is based on frequency-to-optical power mapping. In this scheme, in order to obtain the amplitude comparison function (ACF), two counter-propagation optical channels are constructed by four intensity modulators (IM) and two RF time delay lines. The RF signal-under-test (SUT) is divided evenly into four parts: two of them are time-delayed compared to other two. For the clockwise channel, the optical carrier is firstly intensity modulated by SUT in a manner of double sideband signal with suppression central carrier (CS-DSB). When traveling to the next IM, the CS-DSB optical light is further CS-DSB modulated by a time-delayed replica of SUT. After output from the two cascaded IMs, the RF modulated optical signal is launched into the last two IMs used in a reverse direction and filtered by an optical band-pass filter (OBPF) before received by an optical power meter. Here, it should note that the IM worked in a reverse direction has weak intensity modulation on the transmitted optical wave and can be regarded as a transmission waveguide. Similarly, the optical power traveled in anticlockwise channel can also be detected. For the anticlockwise direction, the optical carrier is CS-DSB modulated by another SUT and time-delayed SUT in sequence. Hence, the ACF can be established to convert the frequency into optical power ratio between two optical signals propagated in opposite directions. Experimental results show that RF signals varying from 0 to 14 GHz can be measured with an acceptable error.

Ohmic contact is an important expression of the energy band change after metal-semiconductor contact. This paper summarizes the prerequisites and methods for achieving ohmic contact in the experimental process. This paper sums up in detail the several important factors affecting the formation of ohmic contact during the experimental process, including annealing temperature and time, epitaxial method, type of semiconductor contact layer and doping concentration, type and thickness of the metal contact layer, and size of the metal/semiconductor contact area, etc. We propose a simple and effective method to find the appropriate annealing temperature and time by determining the under-annealing and over annealing states of ohmic contact during the annealing process through a Ring Transmission Line Model. Through extensive experimental verification, the theoretical reasoning matches well with the experimental results. This has a tremendous role in Improving experimental efficiency.

A multi-functional microwave photonic circuit with meshed architecture is designed and demonstrated on thin film lithium niobate platform. Taking the advantages of the fast response of the Pockels effect and optimized device design, the operation bandwidth of the chip exceeds 60GHz. By controlling the transmission paths of the photon at each node, the on-chip device resources are configurated as a variety of microwave links, corresponding to different signal processing functions. The capabilities of signal generation, down-conversion mixing with high dynamic range and self-interference cancellation with high suppression ratio are experimentally demonstrated. For signal generation, the chip can be regarded as a frequency doubler, and both linear and nonlinear frequency modulated waveforms are obtained with a time-bandwidth product of 2×10⁵ and an in-band spurious suppression ratio higher than 40dB. When configurated as a mixer, the chip achieves a spurious free dynamic range of 105 dB/Hz2/3 and a down-conversion efficiency of -7.4dB. The lithium niobate avoids the nonlinear carrier transportation and absorption existing in traditional silicon photonics, breaking the limitation of linearity and efficiency. As self-interference cancellation mode is set, the interference is suppressed by 50dB over 1.1GHz span. The uniformity of microfabrication in combination with the precise adjustment of the amplitude and phase of the optical field guarantees the high cancellation ratio. To the best of our knowledge, this photonic chip possesses the largest bandwidth and excellent comprehensive performance in terms of active signal processing among integrated multifunctional microwave photonic circuits.

The spectral purity of transmitted signal plays the key role in the application of modern airborne radar. Microwave photonic signal sources based on photoelectric oscillators (OEO) can directly generate high frequency local oscillator signal, and its phase noise level is significantly better than that of traditional microwave local oscillators. Currently, there are limited researches on the application of the photoelectric oscillator in the airborne practical environment, while the operational environment of the airborne radar is extremely harsh. This will hinder the practicality of the photoelectric oscillator in the airborne radar. In this paper, the key role of phase noise in airborne radar scene detection is analyzed first. The influence of different signal forms on the phase noise of the transmitted excitation signal in radar is calculated, and the principle of reducing phase noise to improve the detection performance of airborne radar is summarized. Based on the design framework of the optical oscillator, the phase noise level of the optical oscillator in the airborne vibration environment is analytically analyzed and numerically simulated. The physical model and empirical formula of the influence of the phase noise of the optical oscillator in vibrational environment are established. Finally, the validity of the theory of phase noise deterioration in the vibrational environment is verified by the vibration-test-platform experiment. Combined with the test results, some suggestions for improving the design of the photoelectric oscillator for airborne radar in the actual combat environment are summarized. This work lays a good foundation for the application of photoelectric oscillator in airborne radar.

In this paper, we reviewed and presented our latest progress of linearization methods for microwave photonic systems based on programmable photonic circuits, including the linearization in microwave photonic transmission links and programmable functional circuits.

A silicon nitride narrow-band multimode waveguide Bragg grating (NBMWBG) optical filter with high SLSR and low insertion loss based on the inverse design of MWBG is demonstrated. The complementary lateral-misalignment modulation apodization is used to create physically grating structure of NBMWBG. The insertion loss, 3-dB bandwidth and SLSR of proposed NBMWBG are measured to be 0.9 dB, 0.8 nm and 32 dB, respectively. Keywords: Silicon nitride, Bragg grating, Narrow-band,