Test method and system for monitoring entrance pupil optical power of inter-satellite laser communication terminal in real time

Abstract

The invention provides a test method and system for monitoring entrance pupil optical power of an inter-satellite laser communication terminal in real time, and relates to the technical field of laser communication, and the method comprises the steps: employing a photon integrated microcavity array to collect entrance pupil optical power signals, constructing an optical power distribution mapping matrix, and obtaining an optical power distribution mapping matrix; a self-adaptive photon neural network is used for analyzing optical power space distribution to generate a gain compensation instruction, a response curve of the photoelectric detector is adjusted according to the instruction, and nonlinear modulation is carried out by adopting a point-enhanced plasmon meta-structure surface. According to the invention, high-precision real-time monitoring of the entrance pupil optical power of the inter-satellite laser communication terminal is realized, and the stability and reliability of a communication system are improved.

Description

Technical Field

[0001] The present invention relates to laser communication technology, and in particular to a testing method and system for real-time monitoring of the entrance pupil optical power of an inter-satellite laser communication terminal. Background Art

[0002] As a key technology in the aerospace field, intersatellite laser communication, with its advantages of high bandwidth, high security, and low energy consumption, has become a vital transmission method for space information networks. In intersatellite laser communication systems, accurate measurement and real-time monitoring of entrance pupil optical power are crucial for ensuring the stability and reliability of communication links. Entrance pupil optical power is a core parameter for evaluating laser communication performance, directly affecting key indicators such as the communication system's signal-to-noise ratio, bit error rate, and communication distance.

[0003] Traditional methods for monitoring the entrance pupil optical power of intersatellite laser communication terminals rely primarily on discrete photodetectors and conventional signal processing circuits, sampling the incident optical power and performing analog-to-digital conversion to achieve power monitoring. With the growing demand for space communications and the expansion of deep space exploration missions, existing technologies are becoming increasingly inadequate for optical power monitoring in complex space environments.

[0004] Traditional photoelectric detection methods have difficulty achieving high-precision spatial resolution and cannot accurately capture the non-uniform distribution characteristics of optical power on the entrance pupil plane, especially when subject to spatial disturbances or beam pointing jitter, resulting in a decrease in power monitoring accuracy and affecting subsequent signal processing effects.

[0005] Existing optical power monitoring systems lack adaptive capabilities. When spatial environmental parameters change (such as temperature fluctuations and radiation interference), the system response characteristics drift and gain compensation parameters cannot be adjusted in real time, resulting in unstable measurement results and reducing the reliability of the communication system.

[0006] The signal modulation and processing links in traditional technologies have problems such as narrow linear range and slow dynamic response. When the entrance pupil light power fluctuates violently, it is difficult for the system to achieve accurate measurement of a wide dynamic range. Especially in scenarios where weak light signals and strong light signals appear alternately, it is impossible to simultaneously guarantee the measurement requirements of high sensitivity and wide dynamic range, which restricts the application performance of intersatellite laser communication systems in complex space environments. Summary of the Invention

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

[0008] A first aspect of an embodiment of the present invention provides a test method for real-time monitoring of the entrance pupil optical power of an inter-satellite laser communication terminal, comprising:

[0009] 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;

[0010] 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;

[0011] 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;

[0012] The response curve of the photodetector is adjusted according to the gain compensation instruction to obtain a compensated electrical signal; 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 nanoresonance unit array, and piecewise linear mapping of the electrical signal amplitude is achieved by adjusting the structural parameters of the nanoresonance unit.

[0013] 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:

[0014] 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;

[0015] 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;

[0016] 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;

[0017] 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;

[0018] 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;

[0019] 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.

[0020] 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:

[0021] 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;

[0022] 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;

[0023] 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;

[0024] 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;

[0025] 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.

[0026] 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:

[0027] 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.

[0028] 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.

[0029] 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:

[0030] 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;

[0031] 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;

[0032] 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;

[0033] 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;

[0034] 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.

[0035] 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:

[0036] 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;

[0037] 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;

[0038] 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.

[0039] A second aspect of an embodiment of the present invention provides a test system for real-time monitoring of the entrance pupil optical power of an inter-satellite laser communication terminal, comprising:

[0040] 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;

[0041] 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;

[0042] 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;

[0043] The fourth unit is used to adjust the response curve of the photodetector according to the gain compensation instruction to obtain a compensated electrical signal; and nonlinearly modulate the compensated electrical signal using a point-enhanced plasmon metasurface, wherein the plasmon metasurface is composed of a periodically arranged array of metal-dielectric-metal nanoresonance units, and a piecewise linear mapping of the electrical signal amplitude is achieved by adjusting the structural parameters of the nanoresonance units.

[0044] According to a third aspect of an embodiment of the present invention, an electronic device is provided, including:

[0045] processor;

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

[0047] The processor is configured to call the instructions stored in the memory to execute the aforementioned method.

[0048] According to a fourth aspect of an embodiment of the present invention, a computer-readable storage medium is provided, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the method described above is implemented.

[0049] The beneficial effects of this application are as follows:

[0050] The present invention combines photonic integrated microcavity array technology with adaptive photonic neural networks to achieve high-precision real-time monitoring of the entrance pupil light power of intersatellite laser communication terminals, effectively improving the detection sensitivity and signal-to-noise ratio of weak optical signals, and solving technical problems such as slow response speed and insufficient accuracy of traditional detection methods in space applications.

[0051] By constructing an optical power distribution mapping matrix and analyzing the spatial distribution of optical power at the entrance pupil plane in real time, the present invention can dynamically adjust the response curve of the photodetector according to the actual light intensity, achieve adaptive gain compensation, and ensure the stable operating performance of the system under complex and changing spatial conditions, while reducing the impact of power fluctuations on communication quality.

[0052] The present invention introduces a point-enhanced plasmon metasurface for nonlinear modulation. By precisely controlling the structural parameters of the nanoresonance unit, it achieves piecewise linear mapping of the electrical signal amplitude. This not only optimizes signal processing efficiency, but also significantly reduces system power consumption and complexity. It is suitable for space application scenarios such as intersatellite communications that have strict restrictions on weight, volume and power consumption. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 Schematic diagram of a flow chart of a test method for real-time monitoring of entrance pupil optical power of an intersatellite laser communication terminal according to an embodiment of the present invention;

[0054] Figure 2 This is a flow chart of real-time analysis of optical power spatial distribution by an adaptive photonic neural network according to an embodiment of the present invention;

[0055] Figure 3 This is a bar chart comparing and analyzing the nonlinear modulation performance of plasmon metasurfaces according to embodiments of the present invention. DETAILED DESCRIPTION

[0056] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0057] The technical solution of the present invention is described in detail below with reference to specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments.

[0058] Figure 1 FIG. 1 is a flow chart of a test method for real-time monitoring of the entrance pupil optical power of an intersatellite laser communication terminal according to an embodiment of the present invention. Figure 1 As shown, the method includes:

[0059] 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;

[0060] 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;

[0061] 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;

[0062] The response curve of the photodetector is adjusted according to the gain compensation instruction to obtain a compensated electrical signal; 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 nanoresonance unit array, and piecewise linear mapping of the electrical signal amplitude is achieved by adjusting the structural parameters of the nanoresonance unit.

[0063] In an optional embodiment, 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:

[0064] 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;

[0065] 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;

[0066] 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;

[0067] 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;

[0068] 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;

[0069] 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.

[0070] A microring resonator array was fabricated using a silicon-based photonic integration process. The array consists of 64 microring resonator units arranged in 8 rows and 8 columns. Each unit has a radius of 10 microns, a ring width of 0.5 microns, and a coupling gap of 200 nanometers with the straight waveguide. The fabrication process used an SOI wafer as the substrate, with a top silicon layer thickness of 220 nanometers and a buried oxide layer thickness of 2 microns. Electron beam lithography was used to define the pattern of the microring resonator array, followed by inductively coupled plasma etching to form the microring resonator structure.

[0071] After the preparation is completed, the refractive index distribution of the micro-ring resonator unit is adjusted by ion implantation. In the specific implementation, boron ions are used for implantation with an implantation dose of 1×10 13 to 5×10 14 ions per square centimeter, with injection energy controlled between 50 and 150 keV. By precisely controlling the injection area and dose, the refractive index of each microring resonator unit can be precisely tuned. The resonant wavelength spacing between adjacent microring resonators is set at 0.8 nanometers, creating a wavelength gradient across the entire array. Annealing treatment achieves a quality factor of over 10,000 for each microring resonator unit, meeting the system's preset threshold requirement of 8,000.

[0072] When coupling incident light into the microring resonant cavity array, a focusing grating coupler is used to couple the optical signal from an external light source into the chip. The incident light source is a broadband light source with a central wavelength of 1550 nanometers and a linewidth of 100 nanometers. The microring resonant cavity array modulates the incident light to produce a specific transmission spectrum. Each microring resonant cavity unit strongly absorbs the light signal at its resonant wavelength, forming a valley in the transmission spectrum. The transmission spectrum is photoelectrically converted by the germanium silicon photodetector array integrated on the chip. Each detector has a responsivity of 0.8 amps / watt, a dark current of less than 100 nanoamperes, and a bandwidth greater than 10 gigahertz, outputting light intensity signals from multiple microcavity units.

[0073] The output light intensity signal is subjected to the first wavelet multi-scale decomposition process. The db4 wavelet basis function is selected to decompose the signal into 5 scale levels. By calculating the signal-to-noise ratio evaluation index at each scale, in this embodiment, the signal-to-noise ratio of the third scale reaches the highest value of 15.8 decibels, so the third scale is selected as the optimal decomposition scale. The wavelet coefficients of the optimal decomposition scale are processed using an adaptive soft threshold function. The initial threshold 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. After processing, the signal is reconstructed to obtain the light intensity signal after the first denoising, and the signal-to-noise ratio is improved to 18.2 decibels.

[0074] Crosstalk correction is performed on the light intensity signal after the first noise reduction. Due to the optical coupling effect between adjacent microcavity units in the microcavity array, signal crosstalk will occur. Correction is performed using a pre-measured crosstalk matrix between microcavity units. The matrix consists of 64×64 elements, and each element represents the crosstalk coefficient between the corresponding microcavity units. In actual tests, the crosstalk coefficients of adjacent microcavity units are between 0.05 and 0.15, and the crosstalk coefficients of diagonally adjacent units are between 0.02 and 0.08. The inverse matrix of the crosstalk matrix is ​​applied to correct the light intensity signal to generate a corrected light intensity signal.

[0075] The corrected light intensity signal was subjected to a second wavelet multiscale decomposition. Using the same db4 wavelet basis function, the signal-to-noise ratio (SNR) at the second scale reached a maximum of 20.1 dB, and this scale was selected for processing. The wavelet coefficients were processed using a semi-soft threshold function, with the threshold set at 2.8 times the standard deviation of the wavelet coefficients. After reconstruction, the second denoised light intensity signal was obtained, with the SNR improved to 22.6 dB.

[0076] A third wavelet multiscale decomposition is performed on the light intensity signal after the second denoising step. Using the sym8 wavelet basis function, decomposition is performed into three scale levels. Directional adaptive thresholding is applied to each scale coefficient to enhance the signal's directional characteristics. After reconstruction, the spatial distribution of optical power is obtained, meeting the preset spatial resolution requirement. In this embodiment, the final spatial resolution achieved is 2 microns, meeting the preset resolution requirement of less than 5 microns.

[0077] The resulting optical power distribution mapping matrix is ​​an 8×8 two-dimensional matrix, with each element corresponding to the light intensity value measured by a microcavity unit. This matrix is ​​expanded to a 64×64 high-resolution matrix using a bicubic interpolation algorithm, enabling precise characterization of the spatial distribution of optical power at the entrance pupil plane. In actual tests, this method accurately reflects the spatial non-uniformity of the incident light field, achieving a ratio of maximum to minimum detected light intensity of 28.5:1 and a dynamic range exceeding 25 decibels.

[0078] In an optional embodiment, an adaptive photonic neural network is used to perform real-time analysis on the spatial distribution of optical power. The adaptive photonic neural network dynamically adjusts weight coefficients through a reconfigurable optical waveguide structure, automatically optimizes network parameters based on the spatial distribution characteristics of the entrance pupil optical power, and generates gain compensation instructions including:

[0079] 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;

[0080] 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;

[0081] 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;

[0082] 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;

[0083] 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.

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

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

[0086] To construct a photonic neural network with a reconfigurable optical waveguide structure, a waveguide array was fabricated on a single-crystal silicon substrate using a silicon-based photonic integration process. The array consists of 64×64 microcavity units, each measuring 10μm×10μm. The microcavity units utilize a ring resonator structure, and their resonant frequencies are tunable through the electrocaloric effect. The coupling distance between each microcavity unit and its adjacent units is set at 200nm, with an initial coupling coefficient of 0.3. The quality factor (Q) of the microcavity units ranges from 104 to 105, with a resonant wavelength around 1550nm and a wavelength tuning range of ±5nm. The initial optical transfer function of the photonic neural network is determined by setting the resonant frequency and coupling coefficient of each microcavity unit, with a transmission wavelength range of 1545nm to 1555nm.

[0087] When integrating the DNA molecular switch into the reconfigurable optical waveguide structure, a G-quadruplex DNA molecular switch with a length of 22 base pairs was selected. The DNA molecular switch was fixed on the surface of the microcavity by silanization treatment, with an initial concentration of 5×1012 / cm 2 . The DNA molecular switch presents two conformations, folded and unfolded, under different ionic strengths, and the switching time between the two conformations is 5ms. When the DNA is in the folded state, its spatial size is about 2nm; when it is in the unfolded state, its spatial size is about 8nm. The DNA molecular switch affects the effective refractive index of the microcavity through near-field interaction. The refractive index change caused in the folded state is +0.005, and in the unfolded state it is +0.001. The temperature response coefficient is set to 2×10-4 / ℃, and the operating temperature range is 20℃ to 40℃. At the standard operating temperature of 25℃, the controllable switching of DNA conformation is achieved by regulating the K+ ion concentration in the range of 0.1mM to 100mM, thereby correcting the initial optical transmission function, and the coupling coefficient adjustment range is ±20%.

[0088] When constructing a periodic time crystal structure within a photonic neural network, an electro-optic modulator array is used to periodically modulate the resonant frequency of the microcavity. The modulation period is set to 100 ps, ​​and the modulation depth is ±0.5% of the resonant frequency. The time crystal structure consists of eight periodic units, each modulated by an independent electrical control signal. The modulation signal is a square wave with a 50% duty cycle and a peak voltage of 5V. The modulation frequency of the time crystal structure can be adjusted from 1 GHz to 10 GHz, and the generated frequency sidebands are spaced at integer multiples of the modulation frequency.

[0089] Through dynamic modulation of the time crystal structure, the initial optical transfer function is transformed into a time-varying transmission characteristic, causing the eigenvalues ​​of the transmission matrix to exhibit periodic variations in the time domain, with periods being integer multiples of the modulation period. The bandwidth of the time-varying transmission characteristic is extended to three times the original bandwidth, providing richer dynamic characteristics.

[0090] When setting up a nonlinear optical resonant cavity array as a chaotic computing layer, doped lithium niobate (LiNbO 3 ) material, each cavity has a diameter of 50 μm and a thickness of 500 nm. The third-order nonlinear coefficient of the cavity is 2×10-19 m 2 / W, with an input power threshold of 10mW. When the spatial distribution of optical power is input into a nonlinear optical resonator array, the intensity distribution of the light field in the resonator exhibits chaotic characteristics as the input power varies. During feature extraction, the input optical power spatial distribution data (with a resolution of 1024×1024 pixels) is reduced to a 16×16 region using an optical lens system, with each region corresponding to a nonlinear resonator.

[0091] Incorporating time-varying transmission characteristics, a nonlinear resonant cavity array transforms the input light field nonlinearly, generating a 256×256-dimensional feature mapping matrix. The self-assembly properties of the DNA molecular switch are used to modulate the feature mapping matrix. By varying the K+ ion concentration gradient (linearly increasing from 0.1 mM at one end of the array to 100 mM at the other end), the spatial distribution of the DNA conformation is controlled, with a modulation coefficient ranging from 0.8 to 1.2. The optimization objective function constructed using chaotic dynamics uses light field uniformity as the primary metric, aiming to keep the standard deviation of the output light field less than 10% of the input light field.

[0092] When generating gain compensation instructions based on the calculated results of the optimization objective function, the feature mapping matrix is ​​compared with a preset ideal distribution template to calculate the deviation. If the deviation exceeds a threshold (set at 15%), the system generates gain compensation instructions containing individual adjustment parameters for each of the 64×64 microcavity units. These compensation instructions are transmitted to each microcavity unit via an electrothermal modulator array, with an adjustment range of ±2nm from the resonant frequency and a response time of less than 10ms. The system samples and analyzes the spatial distribution of optical power 100 times per second, generating compensation instructions in real time. This reduces the nonuniformity of the output light field by over 80% and improves power efficiency by 25%.

[0093] In an optional embodiment, 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:

[0094] 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.

[0095] 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.

[0096] This method utilizes the self-assembly properties of DNA molecular switches to modulate the characteristic mapping matrix and combines this with the chaotic dynamics of a nonlinear optical resonator array to construct an optimization objective function. Based on the self-assembly properties of the DNA molecular switch, the change in its free energy must first be calculated. In practice, DNA molecular switches exhibit varying free energy levels in different conformational states. Changes in environmental conditions, such as temperature and ion concentration, can cause the DNA molecule to transition from one conformation to another.

[0097] For a typical DNA hairpin switch, the free energy change can be calculated using the nearest neighbor thermodynamic model. For example, for a DNA hairpin with the sequence 5'-GCGAGCTTTTGCTCGC-3', the free energy change from the closed state to the open state at 25°C and 0.1M NaCl is approximately 6.2 kcal / mol.

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

[0099] The calculated refractive index modulation is input into the DNA modulation kernel function and convolved with the original feature map matrix to generate the DNA-modulated feature map matrix. The DNA modulation kernel function can be designed as a Gaussian kernel function, whose parameters are related to the spatial distribution characteristics of the DNA molecular switch. Assuming the original feature map matrix is ​​a 10×10 matrix with each element ranging from [0, 1], and the DNA modulation kernel function is a 3×3 Gaussian kernel with a standard deviation of 0.8, after the convolution operation, the value of the central region of the generated DNA-modulated feature map matrix will change according to the refractive index modulation, for example, the central element will be modulated from the original value of 0.5 to 0.482.

[0100] Constructing the chaotic dynamics equations for a nonlinear optical resonator requires considering the coupling relationships between variables such as light field intensity, phase, and carrier density. By adjusting control parameters in the chaotic dynamics equations, such as the injection current density and optical feedback intensity, the degree of chaos in the system can be determined. In an optical resonator array, the system exhibits deterministic chaotic characteristics when the injection current density is 1.5 times the threshold current, the optical feedback intensity is 0.15, and the feedback delay time is 2 ns.

[0101] The maximum Lyapunov exponent is used to quantify the degree of chaos in the system. For the nonlinear optical resonator 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 optimization objective function is constructed by combining the squared difference between the DNA-modulated feature mapping matrix and the preset target mapping matrix with a regularization term based on the maximum Lyapunov exponent. Assume that the preset target mapping matrix is ​​also a 10×10 matrix, with each element representing the desired feature mapping value, and the regularization coefficient is set to 0.3. The difference between the DNA-modulated feature mapping matrix and the target mapping matrix is ​​calculated, and the squared sum of the differences for each element is calculated, resulting in an error term of 0.087. Adding this error term to the regularization term based on the maximum Lyapunov exponent (0.3×0.85=0.255) yields a final optimization objective value of 0.342.

[0103] In practical applications, the optimization objective function can be repeatedly calculated by varying the design parameters of the DNA molecular switch (such as sequence length and GC content) and the control parameters of the nonlinear optical resonator (such as injection current and feedback intensity). Optimization algorithms, such as gradient descent, gradually adjust the parameters until the parameter combination that minimizes the optimization objective function is found. For example, after 10 iterations of optimization, it was found that when the DNA sequence was modified to 5'-GCGATCTTTTGATCGC-3' and the injection current density was adjusted to 1.65 times the threshold current, the optimization objective function value could be reduced to 0.128, indicating a significant improvement in system performance.

[0104] The advantage of this method is that it fully utilizes the programmability of DNA molecular switches and the rich dynamic characteristics of nonlinear optical resonators to achieve precise control of the characteristic mapping matrix, providing new technical solutions for optical computing, biosensing and other fields.

[0105] In an optional embodiment, a point-enhanced plasmon metasurface is used to perform nonlinear modulation on the compensated electrical signal, wherein the plasmon metasurface is composed of a periodically arranged array of metal-dielectric-metal nanoresonant units, and the piecewise linear mapping of the electrical signal amplitude is achieved by adjusting the structural parameters of the nanoresonant units, including:

[0106] 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;

[0107] 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;

[0108] 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;

[0109] 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;

[0110] 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.

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

[0112] The effective dielectric constant was calculated using a nonlinear optimization algorithm based on a prefabricated metal-dielectric-metal nanoresonator array. Using a nonlinear optimization method based on a genetic algorithm, the metal-dielectric-metal structure was transformed into a uniform dielectric with a specific dielectric constant. The calculations considered the influence of the lateral dimensions and thickness of the nanoresonator on the effective dielectric constant. Through iterative calculations, it was determined that when the lateral dimensions of the resonator are 200 nm × 200 nm, the thickness of the upper and lower metal layers is 40 nm, and the thickness of the intermediate dielectric layer is 30 nm, the real part of the effective dielectric constant at an operating frequency of 600 terahertz is -2.35, and the imaginary part is 0.27.

[0113] Based on the calculated effective dielectric constant, a nonlinear optimization algorithm was used to further calculate the eigenfrequency and wave vector parameters of the metal-dielectric-metal nanoresonator array. The eigenfrequency characterizes the resonant properties of the structure, while the wave vector parameters describe the propagation characteristics of electromagnetic waves in the structure. In this example, the calculated eigenfrequency is 598.5 terahertz, with a normalized real part of the wave vector of 0.83 and an imaginary part of 0.15.

[0114] To enhance the electric field strength, a point enhancement layer is pre-deposited on the surface of the metal-dielectric-metal nanoresonator array. The point enhancement layer uses gold nanoparticles with a diameter of 10 nanometers and a spacing of 40 nanometers. Based on the polarizability of the gold nanoparticles and the Deide Green function, the local field enhancement factor (LFEF) is calculated. The LFEF represents the enhancement factor of the electric field strength. In this embodiment, the LFEF reaches 15.6, which means that the electric field strength at the point enhancement layer is enhanced by 15.6 times.

[0115] To process the input electrical signal, its amplitude range is divided into multiple subranges, and an electrical signal modulation function is constructed based on the effective dielectric constant, local field enhancement factor, and eigenfrequency. In this embodiment, the 0-1 volt input signal range is divided into four subranges: 0-0.25 volts, 0.25-0.5 volts, 0.5-0.75 volts, and 0.75-1 volt. Within each subrange, the electrical signal modulation function exhibits a linear relationship, but the slope varies across subranges, achieving piecewise linear mapping.

[0116] To optimize the performance of the electrical signal modulation function, a nonlinear optimization algorithm is used to optimize the slope of each subinterval. The optimization goal is to ensure that the slopes of adjacent subintervals exhibit a predetermined difference, thereby achieving nonlinear modulation of the input signal. In this embodiment, the slope ratio of adjacent subintervals is required to be 2:1, meaning that the slope of the subsequent subinterval is twice that of the previous subinterval. Through iterative optimization calculations, the slopes of the four subintervals are determined to be 0.5, 1.0, 2.0, and 4.0, respectively.

[0117] Based on the optimization results of the nonlinear optimization algorithm, the structural parameters of the metal-dielectric-metal nanoresonator array are adjusted. Specifically, the adjustment method is to change the lateral dimensions of different regions within the metal-dielectric-metal nanoresonator array. The lateral dimensions of the resonant units in the first region are adjusted to 190 nm x 190 nm, the second region to 200 nm x 200 nm, the third region to 210 nm x 210 nm, and the fourth region to 220 nm x 220 nm. This adjustment causes the electric field strength and phase response of different regions to change, thereby achieving piecewise 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 nanoresonant unit array based on the effective dielectric constant using the nonlinear optimization algorithm includes:

[0123] 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;

[0124] 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;

[0125] 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.

[0126] Based on the effective dielectric constant of the metal-dielectric-metal nanoresonant unit array, the eigenfrequency and wave vector parameters are calculated through a nonlinear optimization algorithm to achieve accurate characterization of the performance of the metal-dielectric-metal nanoresonant unit array.

[0127] Obtain the structural and material parameters of the metal-dielectric-metal nanoresonant unit array. Structural parameters include the thickness of the metal layer, the thickness of the dielectric layer, and the geometric dimensions and arrangement period of the nanoresonant unit. For example, the metal layer can be made of 30nm thick silver, the dielectric layer can be made of 60nm thick silicon dioxide, and the nanoresonant unit can be designed as a cylindrical structure with a diameter of 200nm and an arrangement period of 400nm. Material parameters include the complex dielectric constants of the metal and dielectric. For example, the complex dielectric constant of silver in the visible light band can be obtained using the Drude model, and the complex dielectric constant of silicon dioxide is approximately 2.1.

[0128] Based on the obtained structural parameters and material parameters, the effective dielectric constant of the metal-dielectric-metal nanoresonance unit array is calculated. This process can be achieved through the effective medium theory, that is, 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 the example, when the volume fraction of silver is 40% and the volume fraction of silicon dioxide is 60%, at a wavelength of 600nm, the real part of the effective dielectric constant of the composite structure is approximately -5.2, and the imaginary part is approximately 0.8.

[0129] A nonlinear optimization algorithm is used to calculate the electromagnetic field distribution within the metal-dielectric-metal nanoresonator array and construct a field intensity mapping matrix. In practice, a distribution model of the electromagnetic field within the resonant unit array can be established, which accounts for the distribution patterns of the electric and magnetic fields within the resonant unit. Numerical calculation methods such as the finite element method or the finite-difference time-domain method can be used to solve Maxwell's equations and obtain the electromagnetic field distribution within the resonant unit.

[0130] During the calculation process, the resonant unit is meshed into multiple computational nodes, and the electric and magnetic field values ​​at each node form the electromagnetic field vector. For N computational nodes, an N×N field intensity mapping matrix can be constructed, which describes the electromagnetic field coupling relationship between different nodes. In actual calculations, the resonant unit can be meshed into 10,000 nodes. The electric and magnetic field distributions at each node are calculated through iterative calculations, thus constructing a 10,000×10,000 field intensity mapping matrix.

[0131] A nonlinear optimization algorithm is used to perform eigenvalue decomposition on the field intensity mapping matrix to obtain the eigenfrequencies of the metal-dielectric-metal nanoresonator array. During eigenvalue decomposition, the eigenvalues ​​and eigenvectors of the field intensity mapping matrix are calculated. The eigenvalues ​​correspond to the eigenfrequencies of the resonant units, and the eigenvectors correspond to the corresponding resonant modes. For large matrices, efficient numerical methods such as the Arnoldi iteration method or the Lanczos algorithm can be used for eigenvalue decomposition.

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

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

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

[0135] The above method can accurately calculate the eigenfrequency and wave vector parameters of metal-dielectric-metal nanoresonant unit arrays, providing theoretical guidance for the design and optimization of nano-optical devices. This method has the advantages of high calculation accuracy and wide applicability, and can effectively characterize the electromagnetic properties of metal-dielectric-metal nanoresonant unit arrays.

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

[0137] 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;

[0138] 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;

[0139] 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;

[0140] The fourth unit is used to adjust the response curve of the photodetector according to the gain compensation instruction to obtain a compensated electrical signal; and nonlinearly modulate the compensated electrical signal using a point-enhanced plasmon metasurface, wherein the plasmon metasurface is composed of a periodically arranged array of metal-dielectric-metal nanoresonance units, and a piecewise linear mapping of the electrical signal amplitude is achieved by adjusting the structural parameters of the nanoresonance units.

[0141] According to a third aspect of an embodiment of the present invention, an electronic device is provided, including:

[0142] processor;

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

[0144] The processor is configured to call the instructions stored in the memory to execute the aforementioned method.

[0145] According to a fourth aspect of an embodiment of the present invention, a computer-readable storage medium is provided, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the method described above is implemented.

[0146] The present invention may be a method, an apparatus, a system and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for executing various aspects of the present invention.

[0147] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. However, these 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 invention.

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