A multilayer transparent antenna system based on frequency selective surfaces

By using a multi-layer transparent antenna system based on a frequency-selective surface, the mechanical limitations, performance bottlenecks, and environmental sensitivity of traditional vehicle antennas are solved, realizing a vehicle communication system that is multi-band compatible, dynamically adaptable to the environment, and highly integrated.

CN120127413BActive Publication Date: 2025-11-11DONGGUAN LONGYI ELECTRONICS TECH
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202510377660.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2025-03-28
Publication Date
2025-11-11
Estimated Expiration
2045-03-28

AI Technical Summary

Technical Problem

Traditional vehicle-mounted antennas have mechanical limitations, performance bottlenecks, and environmental sensitivity issues, making them unable to meet the requirements of multi-band, high-gain, and dynamic environment adaptability.

Method used

A multi-layer transparent antenna system based on frequency selective surface is adopted, including a transparent FSS interlayer, a dynamic impedance matching layer, a conformal waveguide coupling structure and a control module. Dynamic impedance matching and electromagnetic wave coupling are achieved through electronically controlled liquid crystal materials and environmental sensors, and intelligent spectrum management is carried out by combining SDR equipment and convolutional neural networks.

Benefits of technology

It achieves multi-band compatibility, dynamic environmental adaptability and high integration, significantly improving antenna performance stability and communication quality, and reducing manufacturing costs and process difficulty.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure SMS_1
    Figure SMS_1
  • Figure SMS_2
    Figure SMS_2
  • Figure SMS_3
    Figure SMS_3
Patent Text Reader

Abstract

This invention relates to the field of smart antenna design technology, specifically to a multi-layer transparent antenna system based on a frequency selective surface (FSS). The system includes a transparent FSS interlayer, a dynamic impedance matching layer, a conformal waveguide coupling structure, a control module, and a receiving module signal-connected to the control module. The purpose of this invention is to provide a multi-layer transparent antenna system based on a frequency selective surface, characterized by high multi-band compatibility, dynamic environmental adaptability, high integration, and concealment.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of smart antenna design technology, and more specifically to a multilayer transparent antenna system based on a frequency selective surface. Background Technology

[0002] With the rapid development of intelligent connected vehicles, the demand for multi-band, high-gain antennas in vehicle communication systems is becoming increasingly urgent. Traditional vehicle antennas mostly adopt external designs (such as shark fin antennas and rearview mirror integrated antennas), which have the following drawbacks:

[0003] Mechanical limitations: External antennas increase the vehicle's drag coefficient (approximately 0.03-0.05 Cd), and long-term outdoor exposure can easily lead to aging and corrosion;

[0004] Performance bottleneck: Existing glass integrated antennas mainly rely on indium tin oxide (ITO) conductive film, whose conductivity is negatively correlated with light transmittance (sheet resistance > 50Ω / sq when light transmittance > 80%), resulting in transmission loss > 3dB in high-frequency bands (such as 5G millimeter waves), and only supporting a single frequency band;

[0005] Environmental sensitivity: In rainy or snowy weather, the voltage standing wave ratio (VSWR) of a fixed impedance design deteriorates to >2.5 due to changes in the dielectric constant of the glass surface, and the signal bit error rate increases by 2-3 orders of magnitude.

[0006] Existing technical solutions (such as CN113451700A) attempt to embed a metal mesh antenna within a glass interlayer, but the following problems still exist:

[0007] Visible light diffraction of the metal mesh causes optical distortion (moiré fringes) in the windshield; the fixed frequency band design cannot adapt to the multi-standard coordination requirements of 5G / V2X / satellite communication; and the lack of dynamic environmental adaptation mechanism makes it unable to suppress interference such as rain and fog scattering and temperature drift. Summary of the Invention

[0008] In order to overcome the shortcomings and deficiencies of the existing technology, the purpose of this invention is to provide a multi-layer transparent antenna system based on a frequency selective surface, which has the characteristics of high efficiency compatibility of multiple frequency bands, dynamic environmental adaptability, high integration and concealment.

[0009] This invention is achieved through the following technical solution:

[0010] In a first aspect, the present invention discloses a multilayer transparent antenna system based on a frequency selective surface, which includes a transparent FSS interlayer, a dynamic impedance matching layer, a conformal waveguide coupling structure, a control module, and a receiving module that is signal-connected to the control module.

[0011] The transparent FSS interlayer consists of at least two glass substrates, inner and outer, and an intermediate PVB layer. The PVB layer contains a transparent FSS array. The frequency selective surface array includes periodically arranged conductive units for selectively reflecting Ku-band and 5G millimeter-wave communication signals.

[0012] The dynamic impedance matching layer is disposed on the inner surface of the inner glass substrate and includes an electro-controlled liquid crystal material and a driving electrode, which is used to dynamically adjust the dielectric constant according to environmental parameters to achieve impedance matching.

[0013] The conformal waveguide coupling structure is used to couple the electromagnetic waves reflected by the transparent FSS interlayer to the receiving module.

[0014] In conjunction with the first aspect, the conductive unit material of the FSS array is silver nanowires or graphene, and the diameter of the silver nanowires is 20-50 nm with an aspect ratio ≥150; the sheet resistance of the graphene is ≤15Ω / sq and the transmittance is ≥85%.

[0015] In conjunction with the first aspect, further, the unit pattern of the FSS array is a nested cross structure, which includes an outer cross unit and an inner cross unit.

[0016] The outer cross-shaped unit has a side length of 4-6mm and a line width of 0.08-0.12mm, corresponding to Ku-band resonance;

[0017] The inner cross unit has a side length of 1.5-2.5mm and a line width of 0.03-0.07mm, corresponding to 5G millimeter wave resonance.

[0018] In conjunction with the first aspect, the dynamic impedance matching layer further includes:

[0019] Nematic liquid crystal materials have an adjustable dielectric constant ε ranging from 2.3 to 5.5.

[0020] Transparent indium tin oxide driving electrode, with an applied voltage range of 0-5V;

[0021] And an environmental sensor array, including a humidity sensor and a temperature sensor for communicating with the control module, for triggering dielectric constant adjustment.

[0022] In conjunction with the first aspect, the adjustment logic of the dynamic impedance matching layer is further as follows:

[0023] When the humidity sensor detects humidity >80% or the temperature sensor detects temperature <5℃, the control module controls the indium tin oxide driving electrode to apply a 2-4V voltage, so that the dielectric constant ε of the nematic liquid crystal material is adjusted to 3.8-4.5.

[0024] In conjunction with the first aspect, the conformal waveguide coupling structure is further formed at the glass edge by a laser metallization process, comprising:

[0025] Copper plating with a thickness of 8-12μm and a width of 2-4mm;

[0026] Waveguide curvature radius matching unit used to ensure that the deviation between the waveguide path curvature and the windshield edge curvature is ≤5%.

[0027] And a balun converter with an operating frequency band of 18-40GHz for balanced-to-unbalanced signal conversion.

[0028] In conjunction with the first aspect, the control module further includes an SDR device, the working process of which includes the following steps:

[0029] S100. At sampling rate f s ≥94GHz radio frequency signals are captured, time-frequency maps are generated by windowed overlapping FFT, and power spectral density is calculated;

[0030] The formula for calculating power spectral density is:

[0031]

[0032] Where x[n] is the time-domain sampled signal, w[n] is the Hanning window function, H is the frame shift, and S is the time-domain sampled signal. win Here, k is the energy normalization factor for the window function, and k is the frequency index.

[0033] S200. Based on spectrum sensing results, dynamically select the frequency band with the highest signal-to-interference-plus-noise ratio, calculate the optimal impedance matching parameters, and construct a multi-objective optimization function to solve for the optimal frequency band and dielectric constant:

[0034]

[0035] stf∈F available ,∈ min ≤∈≤∈ max

[0036] Where α is the signal-to-interference-plus-noise ratio (SINR) weighting factor, β is the dielectric constant adjustment cost weighting factor, SINR(f) is the SINR of frequency band f, ∈ current This represents the dielectric constant of the current liquid crystal layer;

[0037] S300. Adjust the equivalent phase distribution of the FSS array according to the current attitude. The equivalent calculation formula for the equivalent phase distribution of the FSS array is as follows:

[0038]

[0039] in, Let m be the phase adjustment amount in the m-th row and n-th column of the FSS array. Let a be a Jacobian matrix. desired For the desired beam pattern, a measured The actual measured beam pattern is given, and η is the learning rate.

[0040] In conjunction with the first aspect, the control module further includes a convolutional neural network model, which includes the following steps:

[0041] S400. Input the time-frequency map captured by the SDR device into the convolutional neural network model to extract interference mode features. The equivalent equation for extracting interference mode features is:

[0042]

[0043] Among them, C in The number of input feature channels, For depthwise convolution kernels, For pointwise convolution kernels, y i,j,k To output the value of the k-th channel of the feature map at position (i,j);

[0044] S500. Based on the classification results of the convolutional neural network model, select the optimal combination of communication parameters from the predefined policy library, and define the policy reward function as follows:

[0045] R(s,a)=λ1·Throughput(a)-λ2·BER(a)+λ3·Latency -1 (a)

[0046] Where λ1 is the throughput weight, λ2 is the bit error rate penalty coefficient, λ3 is the latency reward coefficient, Throughput(a) is the data transmission rate corresponding to action a, BER(a) is the bit error rate of action a, and Latency -1 (a) is the reciprocal of the communication delay for action a;

[0047] S600. Distributed training based on the FedProx optimizer is used to update the parameters of the convolutional neural network model. The equivalent equation for distributed training with the FedProx optimizer is:

[0048]

[0049] Where θ represents the neural network model parameters, F k (θ) is the local loss function for the k-th sample, μ is the model aggregation constraint coefficient, and n k Let be the data volume of the k-th sample.

[0050] Secondly, the present invention also discloses a vehicle comprising a multi-layer transparent antenna system as described above, wherein the FSS interlayer area accounts for 60-80% of the total area of ​​the vehicle glass.

[0051] The multilayer transparent antenna system based on a frequency-selective surface of the present invention has the following beneficial effects:

[0052] Multi-band compatibility: By designing periodically arranged conductive units, the transparent FSS interlayer can selectively reflect Ku-band and 5G millimeter-wave communication signals, achieving multi-band compatibility and meeting the diverse needs of modern communication systems;

[0053] Dynamic impedance matching: The dynamic impedance matching layer uses electro-controlled liquid crystal material and driving electrodes, which can dynamically adjust the dielectric constant according to environmental parameters (such as temperature, humidity, etc.) to achieve real-time impedance matching and significantly improve the performance stability of the antenna in different environments.

[0054] High-efficiency electromagnetic wave coupling: The conformal waveguide coupling structure can efficiently couple electromagnetic waves reflected by the transparent FSS interlayer to the receiving module, reducing signal loss and improving communication quality;

[0055] Transparency and Aesthetics: The transparent FSS interlayer adopts a glass substrate and PVB layer structure, which has good transparency and mechanical strength. It can be widely used in smart windows, automotive glass and other scenarios, taking into account both functionality and aesthetics.

[0056] Low cost and easy manufacturing: This invention uses mature glass substrate and PVB layer technology, combined with electro-controlled liquid crystal material and conformal waveguide coupling structure, which reduces manufacturing cost and process difficulty, and is conducive to large-scale promotion and application. Detailed Implementation

[0057] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, a detailed description is provided below in conjunction with specific embodiments. Many specific details are set forth in the following description to provide a thorough understanding of the present invention. However, the present invention can be practiced in many other ways different from those described herein, and those skilled in the art can make similar modifications without departing from the spirit of the invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0058] With the rapid development of intelligent connected vehicles, the demand for multi-band, high-gain antennas in vehicle communication systems is becoming increasingly urgent. Traditional vehicle antennas mostly adopt external designs (such as shark fin antennas and rearview mirror integrated antennas), which have the following drawbacks:

[0059] Mechanical limitations: External antennas increase the vehicle's drag coefficient (approximately 0.03-0.05 Cd), and long-term outdoor exposure can easily lead to aging and corrosion;

[0060] Performance bottleneck: Existing glass integrated antennas mainly rely on indium tin oxide (ITO) conductive film, whose conductivity is negatively correlated with light transmittance (sheet resistance > 50Ω / sq when light transmittance > 80%), resulting in transmission loss > 3dB in high-frequency bands (such as 5G millimeter waves), and only supporting a single frequency band;

[0061] Environmental sensitivity: In rainy or snowy weather, the voltage standing wave ratio (VSWR) of a fixed impedance design deteriorates to >2.5 due to changes in the dielectric constant of the glass surface, and the signal bit error rate increases by 2-3 orders of magnitude.

[0062] Existing technical solutions (such as CN113451700A) attempt to embed a metal mesh antenna within a glass interlayer, but the following problems still exist:

[0063] Visible light diffraction of the metal mesh causes optical distortion (moiré fringes) in the windshield; the fixed frequency band design cannot adapt to the multi-standard coordination requirements of 5G / V2X / satellite communication; and the lack of dynamic environmental adaptation mechanism makes it unable to suppress interference such as rain and fog scattering and temperature drift.

[0064] To address the aforementioned issues, this embodiment discloses a multilayer transparent antenna system based on a frequency selective surface, comprising a transparent FSS interlayer, a dynamic impedance matching layer, a conformal waveguide coupling structure, a control module, and a receiving module signal-connected to the control module.

[0065] The transparent FSS interlayer consists of at least two glass substrates, inner and outer, and an intermediate PVB layer. The PVB layer contains a transparent FSS array. The frequency selective surface array includes periodically arranged conductive units for selectively reflecting Ku-band and 5G millimeter-wave communication signals.

[0066] The dynamic impedance matching layer is disposed on the inner surface of the inner glass substrate and includes an electro-controlled liquid crystal material and a driving electrode, which is used to dynamically adjust the dielectric constant according to environmental parameters to achieve impedance matching.

[0067] The conformal waveguide coupling structure is used to couple the electromagnetic waves reflected by the transparent FSS interlayer to the receiving module.

[0068] Compared to existing technologies, the antenna system in this embodiment achieves reflection efficiencies of 92% and 88% in the Ku band (12-18GHz) and 5G millimeter wave (24-28GHz) respectively, which is more than 40% higher than that of traditional metal patch antennas.

[0069] With a transmittance of ≥85%, the sheet resistance of silver nanowires is ≤10Ω / sq, and the signal transmission loss at 28GHz is ≤0.3dB (compared to ≥3dB for traditional ITO solutions).

[0070] In rainy or snowy weather (humidity > 85%), by dynamically adjusting the dielectric constant of the liquid crystal layer (ε = 3.8–4.5), the VSWR was reduced from 2.7 to 1.4, and the bit error rate was reduced from 1 × 10⁻⁶. -3 Improved to 5×10 -6 ;

[0071] After temperature drift compensation (-30℃~85℃), the antenna gain fluctuation is ≤0.5dB;

[0072] The conformal waveguide has a curvature matching error of ≤3% and a 28GHz signal transmission loss of <0.5dB / cm, which is 76% better than the traditional microstrip line coupling scheme.

[0073] Through the aforementioned collaborative design, a multi-band, highly environmentally adaptable vehicle antenna system is achieved within a limited space. Example data shows that its key performance indicators (gain, loss, bit error rate) are more than three times higher than traditional solutions, and it meets automotive-grade reliability requirements (vibration test ≥20g RMS, high temperature and humidity test 85℃ / 85%RH / 1000h), providing a mass-producible stealth antenna solution for intelligent connected vehicles.

[0074] Furthermore, the conductive unit material of the FSS array is silver nanowire, and the diameter of the silver nanowire is 20-50nm, with an aspect ratio of ≥150. With low cost and mature technology, it meets the current 5G millimeter wave communication requirements and is suitable for economical car models.

[0075] In this embodiment, the silver nanowires are synthesized by the following method:

[0076] Silver nanowires with a diameter of 30±5 nm and a length of 15-20 μm (aspect ratio ≥500) were synthesized by polyol reduction method.

[0077] The dispersion concentration was 0.1 wt%, and the solvent was an ethanol / water mixture (volume ratio 7:3).

[0078] A transparent conductive film with a sheet resistance of 9 Ω / sq was formed by spin-coating a silver nanowire dispersion onto a PET substrate (50 μm thick) at 3000 rpm.

[0079] The dual-frequency cross array is etched using a 355nm ultraviolet laser (4W power, 2m / s scanning speed) with a linewidth accuracy of ±3μm.

[0080] According to performance testing, the FSS array of this embodiment has a transmittance of 86.3% at 550nm wavelength as measured by a spectrophotometer (PerkinElmer Lambda950), a transmission loss of 0.28dB / cm at 28GHz as measured by a vector network analyzer (Keysight N5245B), and a sheet resistance change rate of <2% after 1000 cycles at a bending radius of 5mm.

[0081] Furthermore, the unit pattern of the FSS array is a nested cross structure, which includes an outer cross unit and an inner cross unit.

[0082] The outer cross-shaped unit has a side length of 4-6mm and a line width of 0.08-0.12mm, corresponding to Ku-band (12-18GHz) resonance, covering the needs of satellite navigation and long-distance communication;

[0083] The inner cross unit has a side length of 1.5-2.5mm and a line width of 0.03-0.07mm, corresponding to 5G millimeter wave (24-47GHz) resonance, which meets the requirements of high-capacity vehicle network communication.

[0084] The dual-band performance metrics of the FSS array show:

[0085]

[0086] The spacing and linewidth of the inner and outer cross units are optimized through electromagnetic simulation to suppress inter-band coupling (coupling coefficient <-30dB); the rotational symmetry layout of the units (90° periodic arrangement) reduces polarization sensitivity and adapts to multi-incident angle scenarios; the micron-level linewidth design (0.03-0.12mm) makes the unit pattern invisible to the naked eye (moiré fringe intensity <0.1%); the total area ratio of the nested structure is ≤8% to avoid obstructing the driver's view.

[0087] Furthermore, the dynamic impedance matching layer includes:

[0088] Nematic liquid crystal materials have an adjustable dielectric constant ε ranging from 2.3 to 5.5, which can compensate for impedance shifts caused by rain, snow, and temperature changes in real time, maintain VSWR < 1.5, and eliminate the defect of traditional fixed impedance design where VSWR deteriorates to > 2.5 when humidity > 80%;

[0089] Transparent indium tin oxide driving electrode, with an applied voltage range of 0-5V, avoids the energy loss of traditional mechanical tuning mechanisms;

[0090] And an environmental sensor group, including a humidity sensor and a temperature sensor for communicating with the control module, used to trigger dielectric constant adjustment to solve the problem of false triggering caused by single parameter feedback (such as the difference in humidity response between condensate and rainwater).

[0091] Furthermore, the core of dynamic impedance matching layers lies in solving the problem of precise control of the dynamic impedance matching layer under harsh environments. The adjustment logic of the dynamic impedance matching layer is as follows:

[0092] When the humidity sensor detects humidity >80% or the temperature sensor detects temperature <5℃, the control module controls the indium tin oxide driving electrode to apply a 2-4V voltage, so that the dielectric constant ε of the nematic liquid crystal material is adjusted to 3.8-4.5.

[0093] Specifically, when the humidity threshold is >80%, corresponding to scenarios such as rain and fog, the water film (thickness >10μm) on the windshield surface will significantly change the equivalent dielectric constant (Δε≈0.8), requiring active tuning to maintain impedance matching;

[0094] The temperature threshold is <5℃, which corresponds to the increase in viscosity and the extension of response time of the liquid crystal material due to low temperature. It is necessary to compensate for the dielectric constant drift in advance (ε decreases with temperature at a rate of ≈0.01 / ℃).

[0095] Voltage-dielectric constant nonlinear mapping:

[0096] Limit the driving voltage range (2-4V) to avoid overvoltage causing liquid crystal molecule orientation disorder (breakdown voltage > 5V) or undervoltage preventing effective regulation (threshold voltage < 1.5V);

[0097] High-precision control (Δε / ΔV≈0.7 / V) is achieved by experimentally calibrating the exponential relationship between voltage and dielectric constant.

[0098] Balancing energy consumption and reliability:

[0099] 2-4V low voltage drive reduces power consumption (<0.05W / cm²). 2 This also avoids the electromigration effect caused by high-frequency electric fields aging the ITO electrodes.

[0100] The above features enable the electrical constant adjustment error of this embodiment to be <±0.1, the total delay from sensor detection of environmental change to completion of tuning to be <15ms, and the lifespan to be greater than 100,000 hours under 2-4V drive.

[0101] Furthermore, traditional planar waveguides cannot conform to the complex curved surface of windshields (curvature radius range 500-2000mm), resulting in millimeter-wave signal scattering loss >3dB / cm. This embodiment uses laser metallization to form a waveguide path conforming to the curved surface at the glass edge, reducing transmission loss to <0.5dB / cm.

[0102] The conformal waveguide coupling structure includes:

[0103] A copper plating layer with a thickness of 8-12μm and a width of 2-4mm, exhibiting an adhesion strength >20MPa and a resistivity change rate <3% after salt spray testing;

[0104] A waveguide curvature radius matching unit is used to ensure that the deviation between the waveguide path curvature and the windshield edge curvature is ≤5%. The curvature matching deviation of the waveguide in this unit is ≤3%, and the coating shows no cracks after bending fatigue testing.

[0105] And a balun converter with an operating frequency band of 18-40GHz for balanced-to-unbalanced signal conversion.

[0106] In this embodiment, the waveguide can be obtained through the following design and fabrication:

[0107] Glass curvature modeling:

[0108] The 3D laser scanner was used to obtain the edge curvature data of the windshield of a certain car model, and the average radius of curvature was obtained by fitting.

[0109] Waveguide path planning:

[0110] Algorithm steps:

[0111] Input edge point cloud data of the glass CAD model; generate initial waveguide centerline based on B-spline curve interpolation; iteratively optimize path curvature to satisfy ≤5%.

[0112] Output results: Waveguide path length 1.2m, maximum curvature deviation 2.7%.

[0113] Laser metallization process:

[0114] Glass edge pretreatment; laser-induced electroless plating: local heating to activate the catalyst, chemically depositing a copper layer; electroplating to thicken to 10μm, surface passivation treatment.

[0115] Structural parameters of the balun converter:

[0116] Type: Marchand balun (quarter-wavelength converter);

[0117] Substrate: Rogers RO4350B (thickness 0.254mm, ε=3.66);

[0118] Dimensions: 3.2mm × 1.6mm × 0.8mm.

[0119] Performance testing:

[0120]

[0121] System testing conditions:

[0122] Signal source: Keysight M8195A AWG (generates 28GHz QPSK modulated signal);

[0123] Test environment: Microwave anechoic chamber (ETS-Lindgren AMS-8500).

[0124] Result comparison:

[0125]

[0126]

[0127] Furthermore, the control module includes an SDR device, and the workflow of the SDR device includes the following steps:

[0128] S100. At sampling rate f s ≥94GHz radio frequency signals are captured, time-frequency maps are generated by windowed overlapping FFT, and power spectral density is calculated;

[0129] The formula for calculating power spectral density is:

[0130]

[0131] Where x[n] is the time-domain sampled signal, w[n] is the Hanning window function, H is the frame shift, and S is the time-domain sampled signal. win Here, k is the energy normalization factor for the window function, and k is the frequency index.

[0132] S200. Based on spectrum sensing results, dynamically select the frequency band with the highest signal-to-interference-plus-noise ratio, calculate the optimal impedance matching parameters, and construct a multi-objective optimization function to solve for the optimal frequency band and dielectric constant:

[0133]

[0134] stf∈F available ,∈ min ≤∈≤∈ max

[0135] Where α is the signal-to-interference-plus-noise ratio (SINR) weighting factor, β is the dielectric constant adjustment cost weighting factor, SINR(f) is the SINR of frequency band f, ∈ current This represents the dielectric constant of the current liquid crystal layer;

[0136] S300. Adjust the equivalent phase distribution of the FSS array according to the current attitude. The equivalent calculation formula for the equivalent phase distribution of the FSS array is as follows:

[0137]

[0138] in, Let m be the phase adjustment amount in the m-th row and n-th column of the FSS array. Let a be a Jacobian matrix. desired For the desired beam pattern, a measuredThe actual measured beam pattern is given, and η is the learning rate.

[0139] The core of steps S100-S300 lies in realizing intelligent spectrum management and dynamic tuning of the antenna system through SDR equipment. This aims to solve problems such as the rigidity of frequency bands that traditional antennas only support preset fixed frequency bands and cannot adapt to the dynamic access requirements of multiple standards such as 5G / 6G / V2X, the difficulty of static tuning mechanisms in real time to cope with dynamic interference such as rain, snow and multipath fading, and the inability of fixed beamforming to optimize spectrum efficiency, resulting in limited channel capacity.

[0140] Through the above steps, real-time scanning of the 18-47GHz spectrum and identification of the optimal available channel across the entire frequency band were achieved; dynamic parameter tuning to improve signal quality was achieved by jointly optimizing FSS impedance matching and beam pointing; and intelligent interference suppression was achieved by dynamically switching anti-interference strategies based on machine learning to identify interference patterns.

[0141] Furthermore, the control module also includes a convolutional neural network model, which includes the following steps:

[0142] S400. Input the time-frequency map captured by the SDR device into the convolutional neural network model to extract interference mode features. The equivalent equation for extracting interference mode features is:

[0143]

[0144] Among them, C in The number of input feature channels, For depthwise convolution kernels, For pointwise convolution kernels, y i,j,k To output the value of the k-th channel of the feature map at position (i,j);

[0145] S500. Based on the classification results of the convolutional neural network model, select the optimal combination of communication parameters from the predefined policy library, and define the policy reward function as follows:

[0146] R(s,a)=λ1·Throughput(a)-λ2·BER(a)+λ3·Latency -1 (a)

[0147] Where λ1 is the throughput weight, λ2 is the bit error rate penalty coefficient, λ3 is the latency reward coefficient, Throughput(a) is the data transmission rate corresponding to action a, BER(a) is the bit error rate of action a, and Latency -1 (a) is the reciprocal of the communication delay for action a;

[0148] S600. Distributed training based on the FedProx optimizer is used to update the parameters of the convolutional neural network model. The equivalent equation for distributed training with the FedProx optimizer is:

[0149]

[0150] Where θ represents the neural network model parameters, F k (θ) is the local loss function for the k-th sample, μ is the model aggregation constraint coefficient, and n k Let be the data volume of the k-th sample.

[0151] The core of steps S400-S600 lies in achieving intelligent interference suppression and dynamic parameter optimization through deep learning models. This aims to solve problems such as the inability of traditional threshold detection methods to effectively distinguish diverse interference modes such as rain and snow scattering, multipath fading, and co-channel interference; the difficulty of predefined anti-interference strategies to adapt to dynamically changing electromagnetic environments, resulting in low spectrum utilization; and the excessive response delay of rule-based processing algorithms in high-frequency (such as millimeter wave) scenarios, which affects the real-time performance of communication.

[0152] Through the above steps, we have achieved accurate interference classification with an accuracy of ≥94% in identifying 12 typical interference patterns; dynamic strategy generation that selects the optimal frequency band, modulation method and FSS parameter combination in real time according to the interference type; and adaptive learning capability that improves the ability to suppress interference in unfamiliar environments by enabling online fine-tuning of the model through federated learning.

[0153] Example 2

[0154] In this embodiment, the conductive unit material of the FSS array is graphene, which has a sheet resistance ≤15Ω / sq and a light transmittance ≥85%, providing technical reserves for high-end vehicle models in the future 6G terahertz communication.

[0155] In this embodiment, graphene is synthesized by the following method:

[0156] Monolayer graphene (1-2 layers) was grown on copper foil using CVD and then transferred to a PET substrate (50 μm thick).

[0157] Doping treatment: Immersion in silver nitrate solution (concentration 0.1mol / L) to reduce sheet resistance to 8Ω / sq;

[0158] A dual-frequency cross pattern was defined using SU-8 photoresist, followed by oxygen plasma etching (100W power, 30s time).

[0159] Hydrofluoric acid vapor passivates the etched edges, reducing surface roughness (Ra<1nm).

[0160] Performance testing revealed that the white light transmittance of the FSS array in this embodiment was 87.2% (including the PET substrate), the surface resistance at 40 GHz was 7.5 Ω / sq, the dielectric loss tangent tanδ < 0.02, and the sheet resistance change rate after 100 thermal shocks at -40℃ was < 3%.

[0161] The graphene material in this embodiment differs from the silver nanowire material in Example 1 in the following ways:

[0162] Test Items Silver nanowires FSS Graphene FSS <![CDATA[Material cost ($ / m 2 )]]> 35 120 process complexity Low (solution coating) High (CVD transfer) Maximum operating frequency 60GHz 1THz (theoretical) Flexural properties Excellent (ΔR / R < 2%) Good (ΔR / R<5%)

[0163] Example 3

[0164] This embodiment discloses a vehicle that includes a multi-layer transparent antenna system as described in Embodiments 1-2, wherein the FSS interlayer area accounts for 60-80% of the total glass area of ​​the vehicle.

[0165] In this embodiment, the lower limit of area (60%) ensures that the element density of the FSS array is sufficient to cover the antenna operating wavelength (e.g., the wavelength corresponding to 28GHz is approximately 10.7mm), avoiding a decrease in gain due to sparse elements (for every 10% reduction in area, the gain loss is approximately 1.2dBi).

[0166] The upper limit of area (80%) prevents the FSS pattern from being too dense and affecting the driver's vision (the light transmittance must meet the ECE R43 regulation ≥70%), while avoiding the risk of cracking due to stress concentration in the metal layer during the hot bending of the glass.

[0167] The performance and light transmittance test table is as follows:

[0168]

[0169] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit the scope of protection of the present invention. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the essence and scope of the technical solutions of the present invention.

Claims

1. A multilayer transparent antenna system based on a frequency-selective surface, characterized in that, It includes a transparent FSS interlayer, a dynamic impedance matching layer, a conformal waveguide coupling structure, a control module, and a receiving module that is connected to the control module for signal transmission. The transparent FSS interlayer consists of at least two glass substrates, inner and outer, and an intermediate PVB layer. The PVB layer contains a transparent FSS array. The frequency selective surface array includes periodically arranged conductive units for selectively reflecting Ku-band and 5G millimeter-wave communication signals. The dynamic impedance matching layer is disposed on the inner surface of the inner glass substrate and includes an electro-controlled liquid crystal material and a driving electrode, which is used to dynamically adjust the dielectric constant according to environmental parameters to achieve impedance matching. The conformal waveguide coupling structure is used to couple the electromagnetic waves reflected by the transparent FSS interlayer to the receiving module. The control module includes an SDR device, and the SDR device's workflow includes the following steps: S100. Based on sampling rate Radio frequency signals are captured, time-frequency plots are generated by windowed overlapping FFT, and power spectral density is calculated; The formula for calculating power spectral density is: in, For time-domain sampled signals, For the Hanning window function, For frame shift, The window function energy normalization factor. For frequency index, This refers to the total number of sampling points within an analysis window; This refers to the total number of frames used for averaging; S200. Based on the spectrum sensing results, dynamically select the frequency band with the highest signal-to-interference-plus-noise ratio, calculate the optimal impedance matching parameters, and construct a multi-objective optimization function to solve for the optimal frequency band and dielectric constant: in, The signal-to-interference-plus-noise ratio (SINR) weighting factor. Adjusting the cost weighting factor for dielectric constant frequency band Signal-to-interference-to-noise ratio, This represents the dielectric constant of the current liquid crystal layer; S300. Adjust the equivalent phase distribution of the FSS array according to the current attitude. The equivalent calculation formula for the equivalent phase distribution of the FSS array is as follows: in, For the FSS array OK The phase adjustment amount of the column, For Jacobian matrices, For the desired beam pattern, This is the actual measured beam pattern. The learning rate; Indicates the first During the iteration, the FSS array... line, number The phase adjustment amount of the column; It is the sign of a partial derivative; It is a partial derivative operator, representing the phase adjustment amount of the m-th row and n-th column cell in the FSS array. Find the partial derivative.

2. The multilayer transparent antenna system based on a frequency-selective surface according to claim 1, characterized in that, The conductive unit material of the FSS array is silver nanowire or graphene, and the diameter of the silver nanowire is 20-50nm with an aspect ratio ≥150; the sheet resistance of the graphene is ≤15Ω / sq and the transmittance is ≥85%.

3. The multilayer transparent antenna system based on a frequency-selective surface according to claim 1, characterized in that, The unit pattern of the FSS array is a nested cross structure, which includes an outer cross unit and an inner cross unit. The outer cross-shaped unit has a side length of 4-6mm and a line width of 0.08-0.12mm, corresponding to Ku-band resonance; The inner cross unit has a side length of 1.5-2.5mm and a line width of 0.03-0.07mm, corresponding to 5G millimeter wave resonance.

4. A multilayer transparent antenna system based on a frequency-selective surface according to claim 1, characterized in that, The dynamic impedance matching layer includes: Nematic liquid crystal materials, their dielectric constant The adjustable range is 2.3-5.5; Transparent indium tin oxide driving electrode, with an applied voltage range of 0-5V; And an environmental sensor array, including a humidity sensor and a temperature sensor for communicating with the control module, for triggering dielectric constant adjustment.

5. A multilayer transparent antenna system based on a frequency-selective surface according to claim 4, characterized in that, The adjustment logic of the dynamic impedance matching layer is as follows: When the humidity sensor detects humidity >80% or the temperature sensor detects temperature <5℃, the control module controls the indium tin oxide driving electrode to apply a 2-4V voltage, thereby adjusting the dielectric constant of the nematic liquid crystal material. Adjusted to 3.8-4.

5.

6. A multilayer transparent antenna system based on a frequency-selective surface according to claim 1, characterized in that, The conformal waveguide coupling structure is formed at the glass edge using a laser metallization process, and includes: Copper plating with a thickness of 8-12μm and a width of 2-4mm; Waveguide curvature radius matching unit used to ensure that the deviation between the waveguide path curvature and the windshield edge curvature is ≤5%. And a balun converter with an operating frequency band of 18-40GHz for balanced-to-unbalanced signal conversion.

7. A multilayer transparent antenna system based on a frequency-selective surface according to claim 1, characterized in that, The control module further includes a convolutional neural network model, which includes the following steps: S400. Input the time-frequency map captured by the SDR device into the convolutional neural network model to extract interference mode features. The equivalent equation for extracting interference mode features is: in, The number of input feature channels, For depthwise convolution kernels, For pointwise convolution kernels, To output the feature map of the th The passage is in the location The value on; The input feature map is in the th... Each channel, location The value on; Indicates a depthwise convolution operation; This represents pointwise convolution; S500. Based on the classification results of the convolutional neural network model, select the optimal combination of communication parameters from the predefined policy library, and define the policy reward function as follows: in, As throughput weight, This is the bit error rate penalty coefficient. This is the delayed reward coefficient. For action The corresponding data transmission rate For action bit error rate, For action The communication delay is taken as the reciprocal; S600. Distributed training based on the FedProx optimizer is used to update the parameters of the convolutional neural network model. The equivalent equation for distributed training with the FedProx optimizer is: in, For neural network model parameters, For the first The local loss function for each sample. These are the aggregation constraint coefficients for the model. For the first The amount of data in each sample; Refers to the first Global model parameters after round aggregation; This represents the total amount of data from all samples participating in distributed training.

8. A vehicle, characterized in that, Including the multi-layer transparent antenna system as described in any one of claims 1-7, wherein the FSS interlayer area accounts for 60-80% of the total area of ​​the vehicle glass.

Citation Information

Patent Citations

  • Spiral deacidification safety valve for storage battery shell

    CN113451700A

  • FSS-loaded antenna glass

    CN213026483U