Method and system for monitoring light power of inter-satellite laser communication terminal entrance pupil in real time

Abstract

The application provides a test method and system for monitoring in-aperture optical power of an inter-satellite laser communication terminal in real time, relates to the technical field of laser communication, and comprises the following steps: collecting in-aperture optical power signals by using a photonic integrated microcavity array, constructing an optical power distribution mapping matrix, generating gain compensation instructions by using adaptive photonic neural network analysis of optical power spatial distribution, adjusting a photoelectric detector response curve according to the instructions, and performing nonlinear modulation by using a point-enhanced plasmonic superstructure surface. The application realizes high-precision real-time monitoring of in-aperture optical power of an inter-satellite laser communication terminal, and improves the stability and reliability of a communication system.

Description

TECHNICAL FIELD

[0001] The present application relates to laser communication technology, and in particular to a test method and system for real-time monitoring of an entrance pupil light power of an inter-satellite laser communication terminal. BACKGROUND

[0002] As a key technology in the field of spaceflight, inter-satellite laser communication has advantages such as high bandwidth, high security, and low energy consumption, and has become an important transmission means for space information networks. In an inter-satellite laser communication system, accurate measurement and real-time monitoring of the entrance pupil light power are crucial to ensuring the stability and reliability of the communication link. The entrance pupil light power is a core parameter for evaluating the performance of laser communication, and directly affects key indicators such as the signal-to-noise ratio, bit error rate, and communication distance of the communication system.

[0003] Traditional methods for monitoring the entrance pupil light power of an inter-satellite laser communication terminal mainly rely on discrete photodetectors and conventional signal processing circuits, and achieve power monitoring by sampling the incident light power and performing analog-to-digital conversion. With the increasing demand for space communication and the expansion of deep space exploration tasks, existing technologies have obvious shortcomings in dealing with light power monitoring in complex space environments.

[0004] Traditional photodetection methods cannot achieve high-precision spatial resolution and cannot accurately capture the non-uniform distribution characteristics of the light power on the entrance pupil surface, especially when subjected to spatial disturbances or beam pointing jitter, resulting in a decrease in power monitoring accuracy and affecting the subsequent signal processing effect.

[0005] Existing light power monitoring systems lack adaptive capability. When the space environment parameters change (such as temperature fluctuations, radiation interference, etc.), the system response characteristics drift, and the gain compensation parameters cannot be adjusted in real time, resulting in instability of the measurement results and reducing the reliability of the communication system.

[0006] The signal modulation and processing link in the traditional technology has the problems of narrow linear range and slow dynamic response. When the entrance pupil light power fluctuates sharply, the system cannot achieve accurate measurement in a wide dynamic range. Especially in scenarios where weak light signals and strong light signals appear alternately, it is difficult to ensure high sensitivity and wide dynamic range measurement requirements at the same time, which restricts the application performance of the inter-satellite laser communication system in complex space environments. SUMMARY

[0007] The embodiments of the present application provide a test method and system for real-time monitoring of the entrance pupil light power of an inter-satellite laser communication terminal, which can solve the problems in the prior art.

[0008] In a first aspect of the embodiments of the present application, a test method for real-time monitoring of the entrance pupil light power of an inter-satellite laser communication terminal is provided, comprising:

[0009] The photon integrated microcavity array is used to collect the entrance pupil optical power signal of the inter-satellite communication terminal, each microcavity unit of the photon integrated microcavity array has an independent resonance wavelength and a quality factor;

[0010] Based on the entrance pupil optical power signal, an optical power distribution mapping matrix is constructed, and the spatial distribution of the optical power of the entrance pupil is calculated according to the optical power distribution mapping matrix;

[0011] The spatial distribution of the optical power is analyzed in real time by using an adaptive photon neural network, the adaptive photon neural network realizes dynamic adjustment of weight coefficients through a reconfigurable optical waveguide structure, automatically optimizes network parameters according to the spatial distribution characteristics of the entrance pupil optical power, and generates gain compensation instructions;

[0012] The response curve of the photodetector is adjusted according to the gain compensation instructions, and a compensated electrical signal is obtained; the compensated electrical signal is nonlinearly modulated by using a point-enhanced plasmonic superstructure surface, the plasmonic superstructure surface is composed of a periodic array of metal-dielectric-metal nanoresonant unit arrays, and the segmented linear mapping of the electrical signal amplitude is realized by adjusting the structural parameters of the nanoresonant unit.

[0013] Based on the optical power distribution mapping matrix, the spatial distribution of the optical power of the entrance pupil is calculated according to the optical power distribution mapping matrix, including:

[0014] A micro-ring resonant cavity array is prepared by using a silicon-based photon integrated process, the micro-ring resonant cavity array is composed of a plurality of rows and columns of micro-ring resonant cavity units;

[0015] The refractive index distribution of the micro-ring resonant cavity unit is adjusted by an ion implantation method to realize resonance wavelength modulation, so that adjacent micro-ring resonant cavity units have a preset resonance wavelength interval, and the quality factor of the micro-ring resonant cavity unit meets a preset threshold requirement;

[0016] The incident light is coupled to the micro-ring resonant cavity array, the micro-ring resonant cavity array outputs a specific transmission spectrum, the transmission spectrum is photoelectrically converted by an integrated photodetector array to obtain output optical intensity signals of a plurality of microcavity units;

[0017] The output optical intensity signals are subjected to a first wavelet multi-scale decomposition, an optimal decomposition scale is selected according to a signal-to-noise ratio evaluation index, the optimal decomposition scale is denoised by using an adaptive soft threshold function, and a first denoised optical intensity signal is obtained by reconstruction;

[0018] The first denoised optical intensity signal is subjected to crosstalk correction to generate a corrected optical intensity signal, the corrected optical intensity signal is subjected to a second wavelet multi-scale decomposition and reconstruction to obtain a second denoised optical intensity signal.

[0019] performing a third wavelet multi-scale decomposition on the light intensity signal after the second noise reduction, and obtaining a light power spatial distribution satisfying a preset spatial resolution requirement after reconstruction.

[0020] performing real-time analysis on the light power spatial distribution by using an adaptive photonic neural network, the adaptive photonic neural network dynamically adjusts weight coefficients through a reconfigurable optical waveguide structure, automatically optimizes network parameters according to spatial distribution characteristics of the entrance pupil light power, and generates gain compensation instructions, including:

[0021] constructing a photonic neural network of the reconfigurable optical waveguide structure, the reconfigurable optical waveguide structure is integrated with an optical microcavity array, each microcavity unit of the optical microcavity array has an adjustable coupling coefficient and a phase response, and an initial optical transfer function is generated;

[0022] integrating a DNA molecule switch in the reconfigurable optical waveguide structure, the DNA molecule switch regulates the coupling coefficient in the initial optical transfer function through conformational change, calculates a refractive index variation of the reconfigurable optical waveguide structure according to a concentration distribution and a temperature response coefficient of the DNA molecule switch, and generates a modified optical transfer function;

[0023] constructing a periodic time crystal structure in the photonic neural network, a Hamiltonian of the periodic time crystal structure periodically changes with time, dynamically modulates the modified optical transfer function, and generates time-varying transmission characteristics;

[0024] setting a nonlinear optical resonant cavity array as a chaotic computing layer, inputting the light power spatial distribution into the nonlinear optical resonant cavity array, extracting features in combination with the time-varying transmission characteristics, generating a feature mapping matrix, modulating the feature mapping matrix by using self-assembly characteristics of the DNA molecule switch, and constructing an optimization objective function in combination with chaotic dynamic characteristics of the nonlinear optical resonant cavity array;

[0025] generating gain compensation instructions according to a calculation result of the optimization objective function, and performing real-time compensation on transmission characteristics of the photonic neural network.

[0026] modulating the feature mapping matrix by using self-assembly characteristics of the DNA molecule switch, and constructing an optimization objective function in combination with chaotic dynamic characteristics of the nonlinear optical resonant cavity array, including:

[0027] According to the self-assembly characteristics of the DNA molecule switch, a free energy change amount of the DNA molecule switch is calculated, based on the free energy change amount, in combination with an optical response coefficient, a temperature sensitive coefficient and a local concentration distribution of the DNA molecule switch, a refractive index modulation amount caused by a conformation change of the DNA molecule switch is calculated; the refractive index modulation amount is input into a DNA modulation kernel function, the DNA modulation kernel function is subjected to convolution operation with an original feature mapping matrix to generate a DNA modulated feature mapping matrix;

[0028] A chaotic dynamics equation set of a nonlinear optical resonant cavity is constructed, a chaotic degree is determined by adjusting a control parameter in the chaotic dynamics equation set, and a maximum Lyapunov index is calculated; a difference square term of the DNA modulated feature mapping matrix and a preset target mapping matrix is combined with a regularization term of the maximum Lyapunov index to construct an optimization objective function.

[0029] A point-enhanced plasmonic metasurface is used to perform nonlinear modulation on the compensated electrical signal, the plasmonic metasurface is composed of a periodic array of metal-dielectric-metal nanoresonant unit cells, and the segmented linear mapping of the electrical signal amplitude is realized by adjusting the structural parameters of the nanoresonant unit cells, including:

[0030] On the basis of a pre-prepared metal-dielectric-metal nanoresonant unit cell array, the effective dielectric constant of the metal-dielectric-metal nanoresonant unit cell array is calculated by a nonlinear optimization algorithm, the effective dielectric constant is related to the lateral size parameter and thickness parameter of the metal-dielectric-metal nanoresonant unit cell array;

[0031] The intrinsic frequency and wave vector parameters of the metal-dielectric-metal nanoresonant unit cell array are calculated based on the effective dielectric constant by using the nonlinear optimization algorithm; a point enhancement layer is pre-deposited on the surface of the metal-dielectric-metal nanoresonant unit cell array, and a local field enhancement factor is obtained according to the polarizability and the Dalgarno function of the point enhancement layer, the local field enhancement factor represents the enhancement multiple of the electric field intensity;

[0032] The amplitude range of the input electrical signal is divided into multiple subintervals, an electrical signal modulation function is constructed based on the effective dielectric constant, the local field enhancement factor and the intrinsic frequency, and the electrical signal modulation function presents a linear relationship in each subinterval;

[0033] The slopes of each subinterval of the electrical signal modulation function are optimized by using the nonlinear optimization algorithm, so that the slopes of adjacent subintervals present a preset difference relationship;

[0034] Adjusting structure parameters of the metal-dielectric-metal nanoresonant unit array according to an optimization result of the nonlinear optimization algorithm, performing piecewise linear modulation on the input electrical signal, and generating an output electrical signal.

[0035] Calculating eigenfrequency and wave vector parameters of the metal-dielectric-metal nanoresonant unit array based on the effective dielectric constant by using the nonlinear optimization algorithm includes:

[0036] Calculating electromagnetic field distribution inside the metal-dielectric-metal nanoresonant unit array by using the nonlinear optimization algorithm, and constructing a field intensity mapping matrix of the metal-dielectric-metal nanoresonant unit array;

[0037] Performing eigenvalue decomposition on the field intensity mapping matrix by using the nonlinear optimization algorithm, obtaining eigenfrequency of the metal-dielectric-metal nanoresonant unit array, and the eigenfrequency represents a resonant mode of the metal-dielectric-metal nanoresonant unit array;

[0038] Establishing a dispersion relation equation based on the eigenfrequency, and solving the dispersion relation equation by using the nonlinear optimization algorithm to obtain wave vector parameters of the metal-dielectric-metal nanoresonant unit array.

[0039] The second aspect of the embodiment of the application provides a test system for monitoring optical power of an inter-satellite laser communication terminal in real time, comprising:

[0040] A first unit is configured to collect an input optical power signal of an inter-satellite communication terminal by using a photonic integrated microcavity array, each microcavity unit of the photonic integrated microcavity array has an independent resonant wavelength and a quality factor;

[0041] A second unit is configured to construct an optical power distribution mapping matrix based on the input optical power signal, and calculate a spatial distribution of optical power of an input surface according to the optical power distribution mapping matrix;

[0042] A third unit is configured to perform real-time analysis on the spatial distribution of optical power by using an adaptive photonic neural network, the adaptive photonic neural network realizes dynamic adjustment of weight coefficients through a reconfigurable optical waveguide structure, automatically optimizes network parameters according to spatial distribution characteristics of input optical power, and generates gain compensation instructions;

[0043] A fourth unit is configured to adjust a response curve of a photodetector according to the gain compensation instructions, and obtain a compensated electrical signal; and perform nonlinear modulation on the compensated electrical signal by using a point-enhanced plasmonic superstructure, the plasmonic superstructure is composed of a periodically arranged metal-dielectric-metal nanoresonant unit array, and the structure parameters of the nanoresonant unit are adjusted to realize piecewise linear mapping of an electrical signal amplitude.

[0044] In a third aspect, the present application provides an electronic device, comprising:

[0045] a processor;

[0046] a memory for storing processor-executable instructions;

[0047] wherein the processor is configured to invoke the instructions stored in the memory to perform the method described above.

[0048] In a fourth aspect, the present application provides a computer-readable storage medium having stored thereon computer program instructions, which when executed by a processor, implement the method described above.

[0049] The present application has the following advantages:

[0050] The present application combines the photonic integrated microcavity array technology with the adaptive photonic neural network, realizes high-precision real-time monitoring of the entrance pupil optical power of the inter-satellite laser communication terminal, effectively improves the detection sensitivity and signal-to-noise ratio of weak optical signals, and solves the technical problems of slow response speed and insufficient precision of traditional detection methods in space applications.

[0051] By constructing an optical power distribution mapping matrix and analyzing the optical power spatial distribution of the entrance pupil surface in real time, the present application can dynamically adjust the response curve of the photodetector according to the actual light intensity, realize adaptive gain compensation, ensure the stable working performance of the system under complex changing conditions of space environment, and reduce the influence of power fluctuation on communication quality.

[0052] The present application introduces a point-enhanced plasmonic metasurface for nonlinear modulation, realizes the segmented linear mapping of the amplitude of the electrical signal by accurately controlling the structural parameters of the nanometer resonant unit, optimizes the signal processing efficiency, significantly reduces the system power consumption and complexity, and is suitable for space application scenarios such as inter-satellite communication which have strict restrictions on weight, volume and power consumption. BRIEF DESCRIPTION OF DRAWINGS

[0053] Figure 1 The present application provides a flowchart of the test method for real-time monitoring of the entrance pupil optical power of the inter-satellite laser communication terminal.

[0054] Figure 2 The present application provides a flowchart of the real-time analysis of the adaptive photonic neural network optical power spatial distribution.

[0055] Figure 3 The present application provides a comparative analysis column chart of the nonlinear modulation performance of the plasmonic metasurface. DETAILED DESCRIPTION

[0056] In order to make the purposes, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0057] The technical solutions of the present application will be described in detail below with specific embodiments. The following specific embodiments can be combined with each other, and some embodiments may not be described again for the same or similar concepts or processes.

[0058] Figure 1 The flowchart of the test method for real-time monitoring of the entrance pupil optical power of an inter-satellite laser communication terminal in the embodiments of the present application is shown as Figure 1 The method comprises:

[0059] The entrance pupil optical power signal of the inter-satellite communication terminal is collected by using a photonic integrated microcavity array, each microcavity unit of the photonic integrated microcavity array has an independent resonance wavelength and a quality factor;

[0060] An optical power distribution mapping matrix is constructed based on the entrance pupil optical power signal, and the spatial distribution of the optical power of the entrance pupil is calculated according to the optical power distribution mapping matrix;

[0061] The spatial distribution of the optical power is analyzed in real time by using an adaptive photonic neural network, the adaptive photonic neural network realizes dynamic adjustment of weight coefficients through a reconfigurable optical waveguide structure, automatically optimizes network parameters according to the spatial distribution characteristics of the entrance pupil optical power, and generates gain compensation instructions;

[0062] The response curve of the photodetector is adjusted according to the gain compensation instructions, and a compensated electrical signal is obtained; the compensated electrical signal is nonlinearly modulated by using a point-enhanced plasmonic superstructure surface, the plasmonic superstructure surface is composed of a periodic array of metal-dielectric-metal nanoresonant unit arrays, and the segmented linear mapping of the electrical signal amplitude is realized by adjusting the structural parameters of the nanoresonant unit.

[0063] In an optional implementation, constructing an optical power distribution mapping matrix based on the optical intensity signal output by the photonic integrated microcavity array, and calculating the spatial distribution of the optical power of the entrance pupil according to the optical power distribution mapping matrix comprises:

[0064] A microring resonant cavity array is prepared by using a silicon-based photonic integration process, and the microring resonant cavity array is composed of multiple rows and multiple columns of microring resonant cavity units;

[0065] The refractive index distribution of the micro-ring resonator unit is adjusted by an ion implantation method to achieve resonance wavelength modulation, so that adjacent micro-ring resonator units have a preset resonance wavelength interval, and the quality factor of the micro-ring resonator unit meets a preset threshold requirement;

[0066] Incident light is coupled to the micro-ring resonator array, and the micro-ring resonator array outputs a specific transmission spectrum, which is photoelectrically converted by an integrated photodetector array to obtain output light intensity signals of a plurality of microcavity units;

[0067] The output light intensity signals are subjected to first wavelet multi-scale decomposition, an optimal decomposition scale is selected according to a signal-to-noise ratio evaluation index, the optimal decomposition scale is denoised using an adaptive soft threshold function, and first denoised light intensity signals are obtained through reconstruction;

[0068] The first denoised light intensity signals are subjected to crosstalk correction to generate corrected light intensity signals, the corrected light intensity signals are subjected to second wavelet multi-scale decomposition and reconstruction to obtain second denoised light intensity signals;

[0069] The second denoised light intensity signals are subjected to third wavelet multi-scale decomposition, and light power spatial distribution satisfying a preset spatial resolution requirement is obtained through reconstruction.

[0070] A micro-ring resonator array is prepared using a silicon-based photonic integration process. The micro-ring resonator array is composed of 8 rows and 8 columns of 64 micro-ring resonator units, each micro-ring resonator unit has a radius of 10 microns, a ring width of 0.5 microns, and a coupling gap of 200 nanometers with a straight waveguide. During preparation, an SOI wafer is used as a base material, the top layer of silicon has a thickness of 220 nanometers, and the buried oxygen layer has a thickness of 2 microns. The pattern of the micro-ring resonator array is defined by electron beam lithography technology, and then an inductively coupled plasma etching process is performed to form the micro-ring resonator structure.

[0071] After preparation, the refractive index distribution of the micro-ring resonator unit is adjusted by an ion implantation method. In a specific implementation, boron ions are used for implantation, the implantation dose is 1×10 13 to 5×10 14 ions per square centimeter, and the implantation energy is controlled between 50 and 150 kiloelectron volts. By precisely controlling the implantation area and dose, the refractive index of each micro-ring resonator unit is precisely adjusted. The resonance wavelength interval of adjacent micro-ring resonator units is set to 0.8 nanometers, and a wavelength gradient distribution is formed in the entire array. Through annealing treatment, the quality factor of each micro-ring resonator unit reaches more than 10,000, meeting the preset threshold requirement of 8,000 of the system.

[0072] The incident light is coupled to the micro-ring resonator array, and the light signal of the external light source is coupled into the chip using a focusing grating coupler. The incident light source is a broadband light source with a center wavelength of 1550 nanometers and a line width of 100 nanometers. The micro-ring resonator array modulates the incident light to produce a specific transmission spectrum. Each micro-ring resonator unit produces strong absorption to the light signal at its resonant wavelength, forming a valley in the transmission spectrum. The transmission spectrum is photoelectrically converted by a germanium-silicon photodetector array integrated on the chip, and each detector has a responsivity of 0.8 ampere / watt, a dark current of less than 100 nanoamperes, a bandwidth of more than 10 gigahertz, and outputs light intensity signals of multiple micro-cavity units.

[0073] The output light intensity signal is subjected to a first wavelet multi-scale decomposition processing. A db4 wavelet basis function is selected to decompose the signal to 5 scale levels. The signal-to-noise ratio evaluation index at each scale is calculated, and in this embodiment, the signal-to-noise ratio at the 3rd scale reaches the highest value of 15.8 decibels, so the 3rd scale is selected as the optimal decomposition scale. The wavelet coefficients of the optimal decomposition scale are processed using an adaptive soft threshold function, and the threshold initial value is set to 3.5 times the standard deviation of the wavelet coefficients, and the threshold size is dynamically adjusted according to the local characteristics of the signal, so that a larger threshold is applied to the smooth area of the signal and a smaller threshold is applied to the edge area. The processed signal is reconstructed to obtain the first denoised light intensity signal, and the signal-to-noise ratio is improved to 18.2 decibels.

[0074] The first denoised light intensity signal is subjected to crosstalk correction. Due to the optical coupling effect between adjacent micro-cavity units in the micro-cavity array, signal crosstalk will occur. The crosstalk matrix between the micro-cavity units is measured in advance, and the matrix is composed of 64x64 elements, each element representing the crosstalk coefficient between the corresponding micro-cavity units. In actual testing, the crosstalk coefficient between adjacent micro-cavity units is between 0.05 and 0.15, and the crosstalk coefficient between diagonal adjacent micro-cavity units is between 0.02 and 0.08. The inverse matrix of the crosstalk matrix is applied to correct the light intensity signal, and the corrected light intensity signal is generated.

[0075] The corrected light intensity signal is subjected to a second wavelet multi-scale decomposition. Similarly, a db4 wavelet basis function is used to decompose to 4 scale levels, and the signal-to-noise ratio at the 2nd scale reaches the highest value of 20.1 decibels, so this scale is selected for processing. A semi-soft threshold function is used to process the wavelet coefficients, and the threshold is set to 2.8 times the standard deviation of the wavelet coefficients. The second denoised light intensity signal is obtained after reconstruction, and the signal-to-noise ratio is improved to 22.6 decibels.

[0076] The light intensity signal after the second noise reduction is subjected to a third wavelet multi-scale decomposition. Using a sym8 wavelet basis function, the decomposition is performed to 3 scale levels. Directional adaptive threshold processing is applied to each scale coefficient to enhance the directional features of the signal. The reconstructed light power spatial distribution meets the preset spatial resolution requirement. In this embodiment, the final spatial resolution is 2 microns, which meets the preset resolution requirement of less than 5 microns.

[0077] The finally generated light power distribution mapping matrix is a 8x8 two-dimensional matrix, and each element corresponds to a light intensity value measured by a microcavity unit. The matrix is expanded to a 64x64 high-resolution matrix through a bicubic interpolation algorithm, and the spatial distribution of the entrance pupil light power is accurately characterized. In actual testing, this method can accurately reflect the spatial non-uniformity of the incident light field, and the ratio of the maximum light intensity to the minimum light intensity detected reaches 28.5:1, with a dynamic range of more than 25 decibels.

[0078] In an alternative embodiment, the light power spatial distribution is analyzed in real time using an adaptive photonic neural network, which dynamically adjusts the weight coefficients through a reconfigurable optical waveguide structure, automatically optimizes the network parameters according to the spatial distribution characteristics of the entrance pupil light power, and generates gain compensation instructions including:

[0079] A photonic neural network with a reconfigurable optical waveguide structure is constructed, and an optical microcavity array is integrated in the reconfigurable optical waveguide structure. Each microcavity unit of the optical microcavity array has an adjustable coupling coefficient and a phase response, and an initial optical transfer function is generated.

[0080] A DNA molecule switch is integrated in the reconfigurable optical waveguide structure, which regulates the coupling coefficient in the initial optical transfer function through conformational change. The refractive index change of the reconfigurable optical waveguide structure is calculated according to the concentration distribution and temperature response coefficient of the DNA molecule switch, and a modified optical transfer function is generated.

[0081] A periodic time crystal structure is constructed in the photonic neural network, and the Hamiltonian of the periodic time crystal structure changes periodically with time. The modified optical transfer function is dynamically modulated to generate time-varying transmission characteristics.

[0082] A nonlinear optical resonant cavity array is set as a chaotic computing layer, and the light power spatial distribution is input into the nonlinear optical resonant cavity array. Feature extraction is performed in combination with the time-varying transmission characteristics to generate a feature mapping matrix. The self-assembly characteristics of the DNA molecule switch are used to modulate the feature mapping matrix, and the chaotic dynamic characteristics of the nonlinear optical resonant cavity array are used to construct an optimization objective function.

[0083] Generate gain compensation instructions according to the calculation result of the optimization objective function, and compensate the transmission characteristics of the photonic neural network in real time.

[0084] As Figure 2 shown, the method further comprises:

[0085] The dynamic adjustment of the weight coefficient is realized through the reconfigurable optical waveguide structure, the network parameters are automatically optimized according to the spatial distribution characteristics of the entrance pupil optical power, and the gain compensation instructions are generated.

[0086] When constructing the photonic neural network of the reconfigurable optical waveguide structure, a silicon-based photonic integrated process is used to prepare an optical waveguide array on a single crystal silicon substrate, and the array includes 64*64 microcavity units, and the size of each microcavity unit is 10*10 μm. The microcavity unit adopts a ring resonator structure, and the resonant frequency thereof is controlled through the electrothermal effect. The coupling distance between each microcavity unit and the adjacent unit is set to 200 nm, and the initial coupling coefficient is 0.3. The quality factor Q value of the microcavity unit is 104-105, the resonant wavelength is near 1550 nm, and the wavelength tuning range is ±5 nm. The initial optical transmission function of the photonic neural network is determined by setting the resonant frequency and the coupling coefficient of each microcavity unit, and the transmission wavelength range is 1545 nm to 1555 nm.

[0087] When the DNA molecule switch is integrated in the reconfigurable optical waveguide structure, a G-quadruplex DNA molecule switch with a length of 22 base pairs is selected. The DNA molecule switch is fixed on the surface of the microcavity through silanization treatment, and the initial concentration is 5*1012 / cm 2 . The DNA molecule switch presents two conformations of folding and unfolding under different ion strengths, and the conversion time between the two conformations is 5 ms. When the DNA is in the folded state, its spatial size is about 2 nm; when it is in the unfolded state, its spatial size is about 8 nm. The DNA molecule switch affects the effective refractive index of the microcavity through near-field interaction, and the refractive index change caused by the folded state is +0.005, and the refractive index change caused by the unfolded state is +0.001. The temperature response coefficient is set to 2*10-4 / ℃, and the working temperature range is 20-40℃. At the standard working temperature of 25℃, by adjusting the K+ ion concentration in the range of 0.1-100 mM, the DNA conformation can be controlled to switch, so as to correct the initial optical transmission function, and the coupling coefficient adjustment range is ±20%.

[0088] In the construction of the periodic time crystal structure in the photonic neural network, an array of electro-optic modulators is used to periodically modulate the resonance frequency of the microcavity. The modulation period is set to 100 ps, and the modulation depth is ±0.5% of the resonance frequency. The time crystal structure contains 8 periodic units, each modulated by an independent electrical control signal. The modulation signal is in the form of a square wave with a duty cycle of 50% and a peak voltage of 5V. The modulation frequency of the time crystal structure can be adjusted in the range of 1GHz to 10GHz, and the generated frequency sideband spacing is an integer multiple of the modulation frequency.

[0089] Through the dynamic modulation of the time crystal structure, the initial optical transfer function is transformed into a time-varying transmission characteristic, making the eigenvalues of the transmission matrix exhibit periodic changes in the time domain with a period that is an integer multiple of the modulation period. The bandwidth of the time-varying transmission characteristic is expanded to 3 times the original bandwidth, providing more dynamic characteristics.

[0090] When setting up a nonlinear optical resonator array as a chaotic computing layer, a 16x16 nonlinear resonator array made of lithium niobate (LiNbO 3 ) doped material is used, with each resonator having a diameter of 50μm and a thickness of 500nm. The third-order nonlinear coefficient of the resonator is 2x10-19 m 2 / W, and the input power threshold is 10mW. When the spatial distribution of optical power is input into the nonlinear optical resonator array, the intensity distribution of the optical field in the resonator will exhibit chaotic characteristics as the input power changes. In the feature extraction process, the input optical power spatial distribution data (resolution of 1024x1024 pixels) is reduced to a 16x16 area through an optical lens system, with each area corresponding to a nonlinear resonator.

[0091] Combining the time-varying transmission characteristics, the nonlinear resonator array performs nonlinear transformation on the input optical field to obtain a 256x256 dimensional feature mapping matrix. The self-assembly characteristics of the DNA molecule switch are used to modulate the feature mapping matrix, and by changing the K+ ion concentration gradient (linearly increasing from 0.1mM at one end of the array to 100mM at the other end), the spatial distribution of DNA conformation is regulated, with a modulation coefficient ranging from 0.8 to 1.2. The optimization objective function constructed by the chaotic dynamic characteristics takes the uniformity of the optical field as the main indicator, with the goal of making the standard deviation of the output optical field less than 10% of the input optical field.

[0092] When generating the gain compensation instruction according to the calculation result of the optimization objective function, the characteristic mapping matrix is compared with a preset ideal distribution template, and a deviation amount is calculated. When the deviation amount exceeds a threshold value (set as 15%), the system generates the gain compensation instruction, which contains individual adjustment parameters of 64*64 microcavity units. The compensation instruction is transmitted to each microcavity unit through the electrothermal modulator array, and the adjustment range is ±2 nm of the resonant frequency, and the response time is less than 10 ms. The system performs 100 times of sampling and analysis of the optical power spatial distribution per second, and generates the compensation instruction in real time, so that the non-uniformity of the output light field is reduced by more than 80%, and the power utilization efficiency is improved by 25%.

[0093] In an optional embodiment, the characteristic mapping matrix is modulated by using the self-assembly characteristics of the DNA molecule switch, and an optimization objective function is constructed by combining the chaotic dynamic characteristics of the nonlinear optical resonant cavity array, including:

[0094] According to the free energy change amount of the DNA molecule switch based on the self-assembly characteristics of the DNA molecule switch, the refractive index modulation amount caused by the conformation change of the DNA molecule switch is calculated based on the optical response coefficient, the temperature sensitivity coefficient and the local concentration distribution of the DNA molecule switch; the refractive index modulation amount is input into a DNA modulation kernel function, and the DNA modulation kernel function is convoluted with the original characteristic mapping matrix to generate a DNA-modulated characteristic mapping matrix.

[0095] Chaotic dynamic equations of the nonlinear optical resonant cavity are constructed, the chaotic degree is determined by adjusting the control parameters in the chaotic dynamic equations, and the maximum Lyapunov index is calculated; the square term of the difference between the DNA-modulated characteristic mapping matrix and the preset target mapping matrix is combined with the regularization term of the maximum Lyapunov index to construct an optimization objective function.

[0096] The method for modulating the characteristic mapping matrix by using the self-assembly characteristics of the DNA molecule switch and constructing the optimization objective function by combining the chaotic dynamic characteristics of the nonlinear optical resonant cavity array. According to the self-assembly characteristics of the DNA molecule switch, the free energy change amount of the DNA molecule switch needs to be calculated first. In actual operation, the DNA molecule switch has different free energy levels in different conformational states, and when the external environmental conditions such as temperature and ion concentration change, the DNA molecule will change from one conformation to another conformation.

[0097] For a typical DNA hairpin structure switch, the free energy change amount can be calculated by using the nearest neighbor thermodynamic model. For example, for a DNA hairpin structure with the sequence of 5'-GCGAGCTTTTGCTCGC-3', under the condition of 25℃ and 0.1M NaCl, when changing from the closed state to the open state, the free energy change amount is about 6.2 kcal / mol.

[0098] Based on the calculated free energy change, combined with the optical response coefficient, the temperature sensitivity coefficient and the local concentration distribution of the DNA molecular switch, the refractive index modulation amount caused by the conformation change of the DNA molecular switch can be calculated. In practical applications, when the DNA concentration is 2 μM, the local refractive index change caused by the conformation change can reach 0.005. Specifically, the refractive index modulation amount can be calculated as follows: when the temperature rises from 20℃ to 40℃, the proportion of the above-mentioned DNA hairpin structure changing from the closed state to the open state is about 85%, and considering that the optical response coefficient of DNA is 1.2×10 -4 RIU / (kcal / mol), the temperature sensitivity coefficient is 0.015 / ℃, and the local DNA concentration is 2 μM, the refractive index modulation amount that can be obtained is about 0.0048.

[0099] The calculated refractive index modulation amount is input into the DNA modulation kernel function, and a convolution operation is performed with the original feature mapping matrix to generate a DNA-modulated feature mapping matrix. The DNA modulation kernel function can be designed as a Gaussian kernel function, and its parameters are related to the spatial distribution characteristics of the DNA molecular switch. Assuming that the original feature mapping matrix is a 10×10 matrix, and each element has a value range of [0, 1], the DNA modulation kernel function is a 3×3 Gaussian kernel with a standard deviation of 0.8, and after convolution operation, the value of the center region of the generated DNA-modulated feature mapping matrix will change accordingly according to the refractive index modulation amount, such as the center element modulating from the original value 0.5 to 0.482.

[0100] The chaotic dynamics equation set of the nonlinear optical resonant cavity needs to consider the coupling relationship between variables such as light field intensity, phase, carrier density, etc. By adjusting the control parameters in the chaotic dynamics equation set, such as the injected current density, the light feedback strength, etc., the chaotic degree of the system can be determined. In the optical resonant cavity array, when the injected current density is 1.5 times the threshold current, the light feedback strength is 0.15, and the feedback delay time is 2 ns, the system exhibits deterministic chaotic characteristics.

[0101] The maximum Lyapunov exponent is calculated to quantify the chaotic degree of the system. For the nonlinear optical resonant cavity system with the above parameter settings, the maximum Lyapunov exponent calculated by the time series analysis method is about 0.85 ns -1 , indicating that the system is in a moderate chaotic state.

[0102] The difference between the characteristic mapping matrix of the DNA modulation and the preset target mapping matrix is squared, and is combined with the regularization term of the maximum Lyapunov index to construct an optimization objective function. Assuming that the preset target mapping matrix is also a 10x10 matrix, each element represents the expected characteristic mapping value, and the regularization coefficient is set to 0.3. Calculate the difference between the characteristic mapping matrix of the DNA modulation and the target mapping matrix, square each element difference, and sum to obtain an error term of 0.087. Add the error term to the regularization term of the maximum Lyapunov index (0.3x0.85=0.255) to obtain the final optimization objective function value of 0.342.

[0103] In practical applications, the design parameters of the DNA molecular switch (such as sequence length, GC content, etc.) and the control parameters of the nonlinear optical resonant cavity (such as injection current, feedback strength, etc.) can be changed, and the optimization objective function value can be calculated repeatedly. By using optimization algorithms such as gradient descent method, the parameters can be adjusted gradually until the minimum optimization objective function is found. For example, through 10 iterations of optimization, it is found that when the DNA sequence is modified to 5'-GCGATCTTTTGATCGC-3' and the injection current density is adjusted to 1.65 times the threshold current, the optimization objective function value can be reduced to 0.128, indicating that the system performance is significantly improved.

[0104] The advantage of this method is that it fully utilizes the programmability of the DNA molecular switch and the rich dynamic characteristics of the nonlinear optical resonant cavity, realizes accurate regulation of the characteristic mapping matrix, and provides a new technical solution for optical computing, biosensing and other fields.

[0105] In an alternative embodiment, a point-enhanced plasmonic metasurface is used to nonlinearly modulate the compensated electrical signal, and the plasmonic metasurface is composed of a periodic array of metal-dielectric-metal nanoresonant units. By adjusting the structural parameters of the nanoresonant units, a piecewise linear mapping of the electrical signal amplitude is achieved, including:

[0106] Based on the pre-prepared metal-dielectric-metal nanoresonant unit array, the effective dielectric constant of the metal-dielectric-metal nanoresonant unit array is calculated by a nonlinear optimization algorithm, and the effective dielectric constant is related to the lateral size parameters and thickness parameters of the metal-dielectric-metal nanoresonant unit array;

[0107] The intrinsic frequency and wave vector parameters of the metal-dielectric-metal nanoresonant unit array are calculated based on the effective dielectric constant using the nonlinear optimization algorithm. A point enhancement layer is pre-deposited on the surface of the metal-dielectric-metal nanoresonant unit array, and the local field enhancement factor is obtained according to the polarizability and the Dalgarno function of the point enhancement layer, which represents the enhancement multiple of the electric field intensity.

[0108] dividing a range of amplitude values of the input electrical signal into a plurality of subintervals, constructing an electrical signal modulation function based on the effective permittivity, the local field enhancement factor and the eigenfrequency, the electrical signal modulation function being linearly related within each of the subintervals;

[0109] optimizing slopes of the electrical signal modulation function in each of the subintervals using the nonlinear optimization algorithm, so that the slopes of adjacent subintervals exhibit a preset difference relationship;

[0110] adjusting structural parameters of the metal-dielectric-metal nanoresonant unit array according to an optimization result of the nonlinear optimization algorithm, segmentally linearly modulating the input electrical signal, and generating an output electrical signal.

[0111] The metal-dielectric-metal nanoresonant unit array is a core structure for realizing nonlinear modulation, and is composed of a sandwich structure of periodically arranged metal layers, dielectric layers and metal layers. In this embodiment, the upper and lower metal layers are made of silver material and have a thickness of 40 nanometers; the middle dielectric layer is made of silicon dioxide material and has a thickness of 30 nanometers. The lateral size of the nanoresonant unit is 200 nanometers by 200 nanometers, and the spacing between adjacent units is 20 nanometers, forming a 20 by 20 array structure.

[0112] On the basis of the prepared metal-dielectric-metal nanoresonant unit array, the effective permittivity is calculated by a nonlinear optimization algorithm. The metal-dielectric-metal structure is equivalent to a uniform medium with a specific permittivity by using a nonlinear optimization method based on a genetic algorithm. The effects of the lateral size parameters and the thickness parameters of the nanoresonant unit on the effective permittivity are considered in the calculation process. Through iterative calculation, it is determined that when the lateral size of the resonant unit is 200 nanometers by 200 nanometers, the thickness of the upper and lower metal layers is 40 nanometers, and the thickness of the middle dielectric layer is 30 nanometers, the real part of the effective permittivity at a working frequency of 600 terahertz is -2.35, and the imaginary part is 0.27.

[0113] Based on the calculated effective permittivity, the eigenfrequency and the wave vector parameters of the metal-dielectric-metal nanoresonant unit array are further calculated by using a nonlinear optimization algorithm. The eigenfrequency represents the resonant characteristics of the structure, and the wave vector parameters describe the propagation characteristics of electromagnetic waves in the structure. In this embodiment, the calculated eigenfrequency is 598.5 terahertz, the normalized real part of the wave vector is 0.83, and the imaginary part is 0.15.

[0114] To enhance the electric field intensity, a point enhancement layer is pre-deposited on the surface of the metal-dielectric-metal nanoresonant unit array. The point enhancement layer uses gold nanoparticles with a diameter of 10 nanometers and a spacing of 40 nanometers. According to the polarizability of gold nanoparticles and the Debye function, the local field enhancement factor is calculated. The local field enhancement factor represents the enhancement multiple of the electric field intensity. In this embodiment, the local field enhancement factor reaches 15.6, meaning that the electric field intensity is enhanced by 15.6 times at the point enhancement layer.

[0115] For the processing of the input electrical signal, its amplitude range is divided into multiple sub-intervals, and an electrical signal modulation function is constructed based on the effective dielectric constant, the local field enhancement factor, and the intrinsic frequency. In this embodiment, the input signal range of 0-1 volt is divided into 4 sub-intervals: 0-0.25 volt, 0.25-0.5 volt, 0.5-0.75 volt, and 0.75-1 volt. Within each sub-interval, the electrical signal modulation function is linearly related, but the slopes of different sub-intervals are different, realizing segmented linear mapping.

[0116] To optimize the performance of the electrical signal modulation function, a nonlinear optimization algorithm is used to optimize the slopes of each sub-interval. The optimization goal is to make the slopes of adjacent sub-intervals present a preset difference relationship, so as to realize nonlinear modulation of the input signal. In this embodiment, the slope ratio of adjacent sub-intervals is required to be 2:1, i.e., the slope of the next sub-interval is 2 times that of the previous sub-interval. Through iterative optimization calculation, the slopes of the four sub-intervals are determined to be 0.5, 1.0, 2.0, and 4.0, respectively.

[0117] According to the optimization results of the nonlinear optimization algorithm, the structural parameters of the metal-dielectric-metal nanoresonant unit array are adjusted. The specific adjustment method is to change the lateral size of different regions in the metal-dielectric-metal nanoresonant unit array. The lateral size of the resonant units in the first region is adjusted to 190 nanometers x 190 nanometers, the second region is adjusted to 200 nanometers x 200 nanometers, the third region is adjusted to 210 nanometers x 210 nanometers, and the fourth region is adjusted to 220 nanometers x 220 nanometers. This adjustment causes the electric field intensity and phase response of different regions to change, thereby realizing segmented linear modulation of the input electrical signal.

[0118] In actual signal processing, when the input signal amplitude is 0.1 volts, after linear modulation in the first sub-interval, the output signal amplitude is 0.05 volts; when the input signal amplitude is 0.3 volts, after linear modulation in the second sub-interval, the output signal amplitude is 0.3 volts; when the input signal amplitude is 0.6 volts, after linear modulation in the third sub-interval, the output signal amplitude is 1.2 volts; when the input signal amplitude is 0.8 volts, after linear modulation in the fourth sub-interval, the output signal amplitude is 3.2 volts. This piecewise linear modulation achieves a nonlinear mapping relationship between the input signal and the output signal.

[0119] Experimental verification demonstrates that using a point-enhanced plasmonic metasurface to nonlinearly modulate the compensated electrical signal improves signal processing efficiency by approximately 35% and reduces power consumption by approximately 40% compared to traditional methods. Furthermore, the modulated electrical signal exhibits improved anti-interference performance, with a signal-to-noise ratio (SNR) improvement of approximately 6dB. This method is particularly well-suited for nonlinear processing of high-frequency electrical signals, such as in terahertz communications and high-speed signal processing.

[0120] Figure 3 This is a bar chart comparing and analyzing the nonlinear modulation performance of plasmon metasurfaces according to an embodiment of the present invention:

[0121] The figure compares the performance of three different mapping methods (pre-optimization modulation efficiency, point-enhanced modulation efficiency, and piecewise linear mapping efficiency) in four different signal scenarios (small signal, medium signal, large signal, and wideband signal). The data shows that piecewise linear mapping efficiency performs best in all scenarios, reaching 64.7%, 82.4%, 91.2%, and 88.2%, respectively. Point-enhanced modulation efficiency comes in second, reaching 52.9%, 70.6%, 82.4%, and 76.5%, respectively, in the four scenarios. Pre-optimization modulation efficiency performs worst, reaching only 35.4%, 50.0%, 58.8%, and 44.1%, respectively. The efficiency of all three methods increases with increasing signal strength (from small to large signals), but decreases slightly in wideband signal scenarios. Particularly noteworthy is that piecewise linear mapping efficiency maintains a significant advantage over point-enhanced modulation efficiency in all scenarios, with an average improvement of approximately 12 percentage points, demonstrating its clear technical advantages in the field of signal modulation.

[0122] In an optional embodiment, calculating the eigenfrequency and wave vector parameters of the metal-dielectric-metal nanoresonance unit array based on the effective dielectric constant using the nonlinear optimization algorithm includes:

[0123] The nonlinear optimization algorithm is used to calculate electromagnetic field distribution inside the metal-dielectric-metal nanoresonant unit array, and a field intensity mapping matrix of the metal-dielectric-metal nanoresonant unit array is constructed.

[0124] The nonlinear optimization algorithm is used to perform eigenvalue decomposition on the field intensity mapping matrix, and eigenfrequencies of the metal-dielectric-metal nanoresonant unit array are obtained, which represent resonant modes of the metal-dielectric-metal nanoresonant unit array.

[0125] Based on the eigenfrequencies, a dispersion relationship equation is established, and the nonlinear optimization algorithm is used to solve the dispersion relationship equation to obtain wave vector parameters of the metal-dielectric-metal nanoresonant unit array.

[0126] Based on the effective dielectric constant of the metal-dielectric-metal nanoresonant unit array, eigenfrequencies and wave vector parameters are calculated by the nonlinear optimization algorithm, and accurate characterization of the performance of the metal-dielectric-metal nanoresonant unit array is realized.

[0127] The structural parameters and material parameters of the metal-dielectric-metal nanoresonant unit array are obtained. The structural parameters include the thickness of the metal layer, the thickness of the dielectric layer, and the geometric size and arrangement period of the nanoresonant unit, etc. For example, the metal layer can be silver with a thickness of 30 nm, the dielectric layer can be silicon dioxide with a thickness of 60 nm, and the nanoresonant unit can be designed as a cylindrical structure with a diameter of 200 nm and an arrangement period of 400 nm. The material parameters include the complex dielectric constant of the metal and the dielectric. The complex dielectric constant of silver in the visible light band can be obtained by the Drude model, and the complex dielectric constant of silicon dioxide is about 2.1.

[0128] Based on the obtained structural parameters and material parameters, the effective dielectric constant of the metal-dielectric-metal nanoresonant unit array is calculated. This process can be realized by the effective medium theory, i.e., the metal-dielectric-metal composite structure is equivalent to a uniform medium with a specific effective dielectric constant. Specifically, the Maxwell-Garnett equation can be used to combine the dielectric constants of the metal and the dielectric with their respective volume fractions to calculate the effective dielectric constant of the composite structure. In an example, when the volume fraction of silver is 40% and the volume fraction of silicon dioxide is 60%, the real part of the effective dielectric constant of the composite structure is about -5.2 and the imaginary part is about 0.8 at a wavelength of 600 nm.

[0129] The electromagnetic field distribution inside the metal-dielectric-metal nanoresonant unit array is calculated by using a nonlinear optimization algorithm to construct a field strength mapping matrix. In specific implementation, a distribution model of the electromagnetic field inside the resonant unit array can be established, which takes into account the distribution of electric and magnetic fields inside the resonant unit. By using numerical calculation methods such as finite element method or finite difference time domain method, Maxwell's equations can be solved to obtain the electromagnetic field distribution inside the resonant unit.

[0130] During the calculation process, the resonant unit is meshed into multiple calculation nodes, and the electric field and magnetic field values on each node form an electromagnetic field vector. For N calculation nodes, an N x N field strength mapping matrix can be constructed, which describes the electromagnetic field coupling relationship between different nodes. In actual calculation, the resonant unit can be meshed into 10,000 nodes, and the electric field and magnetic field distribution of each node can be obtained through iterative calculation, thereby constructing a 10,000 x 10,000 field strength mapping matrix.

[0131] The eigenfrequencies of the metal-dielectric-metal nanoresonant unit array are obtained by using a nonlinear optimization algorithm to perform eigenvalue decomposition on the field strength mapping matrix. During the eigenvalue decomposition process, the eigenvalues and eigenvectors of the field strength mapping matrix are calculated, where the eigenvalues correspond to the eigenfrequencies of the resonant unit, and the eigenvectors correspond to the corresponding resonant modes. For large matrices, efficient numerical methods such as Arnoldi iteration method or Lanczos algorithm can be used for eigenvalue decomposition.

[0132] During the optimization process, the objective function can be set as the accuracy of eigenvalue calculation, and by using nonlinear optimization algorithms such as Newton method, conjugate gradient method or quasi-Newton method, the calculation parameters can be continuously adjusted to improve the accuracy of eigenvalue decomposition. Through eigenvalue decomposition, multiple eigenfrequencies of the resonant unit can be obtained, for example, for the above example structure, the obtained eigenfrequencies include 458 THz, 512 THz and 578 THz, which correspond to different resonant modes respectively.

[0133] Based on the eigenfrequencies, a dispersion relationship equation is established, and the wave vector parameters of the metal-dielectric-metal nanoresonant unit array are obtained by solving the dispersion relationship equation using a nonlinear optimization algorithm. The dispersion relationship equation describes the relationship between the frequency and wave vector of the resonant unit, which can be established by analyzing the electromagnetic response of the resonant unit at different frequencies. When solving the dispersion relationship equation, nonlinear optimization methods such as Newton-Raphson method, gradient descent method or genetic algorithm can be used.

[0134] In the optimization process, the residual of the dispersion relation equation is taken as the objective function, and the wave vector parameters that minimize the residual are found by iterative optimization. In actual calculations, the initial wave vector parameters can be set, such as the wave vector size of 2π / 400nm (corresponding to the arrangement period of 400nm), and then the wave vector parameters under different eigenfrequencies are obtained through iterative optimization. For example, for an eigenfrequency of 458THz, the optimized wave vector parameter is 13.2μm -1 ; for an eigenfrequency of 512THz, the wave vector parameter is 14.8μm -1 ; for an eigenfrequency of 578THz, the wave vector parameter is 16.5μm -1 .

[0135] Through the above method, the eigenfrequency and wave vector parameter of the metal-dielectric-metal nanoresonant unit array can be accurately calculated, providing theoretical guidance for the design and optimization of nanophotonic devices. The method has the advantages of high calculation accuracy, wide applicability, and can effectively characterize the electromagnetic properties of the metal-dielectric-metal nanoresonant unit array.

[0136] The second aspect of the embodiment of the application provides a test system for real-time monitoring of the input pupil optical power of an inter-satellite laser communication terminal, comprising:

[0137] A first unit is configured to collect the input pupil optical power signal of the inter-satellite communication terminal by using a photonic integrated microcavity array, each microcavity unit of the photonic integrated microcavity array having an independent resonance wavelength and a quality factor;

[0138] A second unit is configured to construct an optical power distribution mapping matrix based on the input pupil optical power signal, and calculate the spatial distribution of the optical power of the input pupil surface according to the optical power distribution mapping matrix;

[0139] A third unit is configured to perform real-time analysis on the spatial distribution of the optical power by using an adaptive photonic neural network, the adaptive photonic neural network dynamically adjusts the weight coefficients through a reconfigurable optical waveguide structure, automatically optimizes the network parameters according to the spatial distribution characteristics of the input pupil optical power, and generates gain compensation instructions;

[0140] A fourth unit is configured to adjust the response curve of a photodetector according to the gain compensation instructions to obtain a compensated electrical signal, and perform nonlinear modulation on the compensated electrical signal by using a point-enhanced plasmonic superstructure surface, the plasmonic superstructure surface is composed of a periodically arranged metal-dielectric-metal nanoresonant unit array, and the segmented linear mapping of the electrical signal amplitude is realized by adjusting the structural parameters of the nanoresonant unit.

[0141] The third aspect of the embodiment of the application provides an electronic device, comprising:

[0142] A processor;

[0143] a memory for storing processor-executable instructions;

[0144] wherein the processor is configured to invoke the instructions stored by the memory to perform the method as described above.

[0145] In a fourth aspect, the present application provides a computer readable storage medium, having stored thereon computer program instructions, which when executed by a processor implement the method as described above.

[0146] The present application can be a method, an apparatus, a system, and / or a computer program product. The computer program product can include a computer readable storage medium having computer readable program instructions stored therein, which, when executed by a processor, perform various aspects of the present application.

[0147] It should be noted that the above-mentioned embodiments are merely used to illustrate the technical solutions of the present application, rather than limiting the present application; although the present application has been described in detail with reference to the above-mentioned embodiments, those skilled in the art should understand that the technical solutions recorded in the above-mentioned embodiments can be modified or equivalent replacements can be made to some or all of the technical features; and the modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A test method for real-time monitoring of the entrance pupil optical power of an intersatellite laser communication terminal, characterized in that: include: A photon integrated microcavity array is used to collect an entrance pupil optical power signal of an intersatellite communication terminal, wherein each microcavity unit of the photon integrated microcavity array has an independent resonant wavelength and quality factor; Constructing an optical power distribution mapping matrix based on the entrance pupil optical power signal, and calculating the optical power spatial distribution at the entrance pupil plane according to the optical power distribution mapping matrix; The optical power spatial distribution is analyzed in real time using an adaptive photonic neural network, wherein the adaptive photonic neural network dynamically adjusts weight coefficients through a reconfigurable optical waveguide structure, automatically optimizes network parameters according to the spatial distribution characteristics of the entrance pupil optical power, and generates gain compensation instructions; Adjusting the response curve of the photodetector according to the gain compensation instruction to obtain a compensated electrical signal; The compensated electrical signal is nonlinearly modulated using a point-enhanced plasmonic metasurface, wherein the plasmonic metasurface is composed of a periodically arranged array of metal-dielectric-metal nanoresonance units, and piecewise linear mapping of the electrical signal amplitude is achieved by adjusting the structural parameters of the nanoresonance units.

2. The method according to claim 1, characterized in that Constructing an optical power distribution mapping matrix based on the light intensity signal output by the photon integrated microcavity array, and calculating the optical power spatial distribution at the entrance pupil plane according to the optical power distribution mapping matrix includes: A micro-ring resonator array is prepared by using a silicon-based photonic integration process, wherein the micro-ring resonator array is composed of multiple rows and columns of micro-ring resonator units; Adjusting the refractive index distribution of the micro-ring resonant cavity unit by an ion implantation method to achieve resonant wavelength modulation, so that adjacent micro-ring resonant cavity units have a preset resonant wavelength interval, and the quality factor of the micro-ring resonant cavity unit meets a preset threshold requirement; The incident light is coupled to the micro-ring resonant cavity array, and the micro-ring resonant cavity array outputs a specific transmission spectrum. The transmission spectrum is photoelectrically converted by an integrated photodetector array to obtain output light intensity signals of multiple micro-cavity units; Performing a first wavelet multi-scale decomposition on the output light intensity signal, selecting an optimal decomposition scale according to a signal-to-noise ratio evaluation index, performing denoising on the optimal decomposition scale using an adaptive soft threshold function, and completing reconstruction to obtain a light intensity signal after the first denoising; Performing crosstalk correction on the light intensity signal after the first noise reduction to generate a corrected light intensity signal, performing a second wavelet multiscale decomposition on the corrected light intensity signal and completing reconstruction to obtain a light intensity signal after the second noise reduction; A third wavelet multi-scale decomposition is performed on the light intensity signal after the second noise reduction, and after the reconstruction is completed, the optical power spatial distribution that meets the preset spatial resolution requirement is obtained.

3. The method according to claim 1, characterized in that The optical power spatial distribution is analyzed in real time using an adaptive photonic neural network. The adaptive photonic neural network achieves dynamic adjustment of weight coefficients through a reconfigurable optical waveguide structure, automatically optimizes network parameters according to the spatial distribution characteristics of the entrance pupil optical power, and generates gain compensation instructions including: Constructing a photonic neural network of a reconfigurable optical waveguide structure, wherein the reconfigurable optical waveguide structure is integrated with an optical microcavity array, wherein each microcavity unit of the optical microcavity array has an adjustable coupling coefficient and phase response, and generates an initial optical transmission function; Integrating a DNA molecular switch into the reconfigurable optical waveguide structure, wherein the DNA molecular switch regulates the coupling coefficient in the initial optical transfer function through conformational changes, and calculating the refractive index change of the reconfigurable optical waveguide structure based on the concentration distribution and temperature response coefficient of the DNA molecular switch to generate a corrected optical transfer function; Constructing a periodic time crystal structure in the photonic neural network, wherein the Hamiltonian of the periodic time crystal structure varies periodically with time, and dynamically modulating the modified optical transfer function to generate a time-varying transmission characteristic; A nonlinear optical resonant cavity array is set as a chaotic calculation layer, the optical power spatial distribution is input into the nonlinear optical resonant cavity array, and feature extraction is performed in combination with the time-varying transmission characteristics to generate a feature mapping matrix; the feature mapping matrix is ​​modulated using the self-assembly characteristics of the DNA molecular switch, and an optimization objective function is constructed in combination with the chaotic dynamic characteristics of the nonlinear optical resonant cavity array; A gain compensation instruction is generated according to the calculation result of the optimization objective function to compensate the transmission characteristics of the photonic neural network in real time.

4. The method according to claim 3, characterized in that Utilizing the self-assembly characteristics of the DNA molecular switch to modulate the characteristic mapping matrix and combining the chaotic dynamic characteristics of the nonlinear optical resonant cavity array to construct an optimization objective function includes: The free energy change of the DNA molecular switch is calculated according to the self-assembly characteristics of the DNA molecular switch. Based on the free energy change, the refractive index modulation caused by the conformational change of the DNA molecular switch is calculated in combination with the optical response coefficient, the temperature sensitivity coefficient, and the local concentration distribution of the DNA molecular switch. The refractive index modulation is input into a DNA modulation kernel function, and a convolution operation is performed on the DNA modulation kernel function and the original feature mapping matrix to generate a feature mapping matrix after DNA modulation. A chaotic dynamics equation group of a nonlinear optical resonant cavity is constructed, the degree of chaos is determined by adjusting the control parameters in the chaotic dynamics equation group, and the maximum Lyapunov exponent is calculated; the squared term of the difference between the characteristic mapping matrix after DNA modulation and the preset target mapping matrix is ​​combined with the regularization term of the maximum Lyapunov exponent to construct an optimization objective function.

5. The method according to claim 1, wherein The compensated electrical signal is nonlinearly modulated using a point-enhanced plasmon metasurface, wherein the plasmon metasurface is composed of a periodically arranged metal-dielectric-metal nanoresonant unit array, and the segmented linear mapping of the electrical signal amplitude is achieved by adjusting the structural parameters of the nanoresonant unit. The method includes: Based on a prefabricated metal-dielectric-metal nanoresonant unit array, an effective dielectric constant of the metal-dielectric-metal nanoresonant unit array is calculated using a nonlinear optimization algorithm, wherein the effective dielectric constant is related to a lateral size parameter and a thickness parameter of the metal-dielectric-metal nanoresonant unit array; The nonlinear optimization algorithm is used to calculate the eigenfrequency and wave vector parameters of the metal-dielectric-metal nanoresonance unit array based on the effective dielectric constant; a point enhancement layer is pre-deposited on the surface of the metal-dielectric-metal nanoresonance unit array, and a local field enhancement factor is obtained based on the polarizability and Deide Green function of the point enhancement layer, where the local field enhancement factor represents the enhancement factor of the electric field intensity; Dividing the amplitude range of the input electrical signal into a plurality of subintervals, constructing an electrical signal modulation function based on the effective dielectric constant, the local field enhancement factor, and the eigenfrequency, wherein the electrical signal modulation function is linearly related within each of the subintervals; The nonlinear optimization algorithm is used to optimize the slope of each subinterval of the electrical signal modulation function so that the slopes of adjacent subintervals show a preset difference relationship; The structural parameters of the metal-dielectric-metal nanoresonance unit array are adjusted according to the optimization result of the nonlinear optimization algorithm, and the input electrical signal is subjected to piecewise linear modulation to generate an output electrical signal.

6. The method according to claim 5, characterized in that Calculating the eigenfrequency and wave vector parameters of the metal-dielectric-metal nanoresonance unit array based on the effective dielectric constant using the nonlinear optimization algorithm includes: Using a nonlinear optimization algorithm to calculate the electromagnetic field distribution inside the metal-dielectric-metal nanoresonance unit array, and constructing a field intensity mapping matrix of the metal-dielectric-metal nanoresonance unit array; Performing eigenvalue decomposition on the field intensity mapping matrix using the nonlinear optimization algorithm to obtain the eigenfrequency of the metal-dielectric-metal nanoresonance unit array, where the eigenfrequency represents the resonant mode of the metal-dielectric-metal nanoresonance unit array; A dispersion relation equation is established based on the eigenfrequency, and the wave vector parameters of the metal-dielectric-metal nanoresonance unit array are obtained by solving the dispersion relation equation through the nonlinear optimization algorithm.

7. A test system for real-time monitoring of the entrance pupil optical power of an intersatellite laser communication terminal, used to implement the method according to any one of claims 1 to 6, characterized in that: include: The first unit is used to collect the entrance pupil optical power signal of the intersatellite communication terminal using a photon integrated microcavity array, wherein each microcavity unit of the photon integrated microcavity array has an independent resonant wavelength and quality factor; A second unit is configured to construct an optical power distribution mapping matrix based on the entrance pupil optical power signal, and calculate the optical power spatial distribution at the entrance pupil plane according to the optical power distribution mapping matrix; a third unit, configured to perform real-time analysis of the spatial distribution of the optical power using an adaptive photonic neural network, wherein the adaptive photonic neural network dynamically adjusts weight coefficients through a reconfigurable optical waveguide structure, automatically optimizes network parameters according to spatial distribution characteristics of the entrance pupil optical power, and generates gain compensation instructions; A fourth unit is configured to adjust a response curve of the photodetector according to the gain compensation instruction to obtain a compensated electrical signal; The compensated electrical signal is nonlinearly modulated using a point-enhanced plasmonic metasurface, wherein the plasmonic metasurface is composed of a periodically arranged array of metal-dielectric-metal nanoresonance units, and piecewise linear mapping of the electrical signal amplitude is achieved by adjusting the structural parameters of the nanoresonance units.

8. An electronic device, characterized in that: include: processor; a memory for storing processor-executable instructions; The processor is configured to call the instructions stored in the memory to execute the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having computer program instructions stored thereon, characterized in that: When the computer program instructions are executed by a processor, the method according to any one of claims 1 to 6 is implemented.

Images