Dynamic compensation method for polarization signals in low-altitude complex terrain, UAV communication device and system

Abstract

This invention discloses a method for dynamic compensation of polarization signals in low-altitude complex terrain, an unmanned aerial vehicle (UAV) communication device, and a system, belonging to the field of wireless communication. The method includes: utilizing an edge computing platform to collect and analyze channel polarization state data in real time; employing an adaptive dynamic compensation algorithm; setting a set of candidate modulation schemes based on the channel's signal-to-noise ratio; and constructing a constraint function based on the set of candidate modulation schemes to achieve high-fidelity transmission of polarization signals. The system also integrates interference prediction, federated learning cluster collaboration, and secure communication modules, thereby achieving high security and robust communication while ensuring low latency and low power consumption. This invention effectively solves the problems of polarization signal mismatch, weak anti-interference capability, and poor environmental adaptability in existing technologies under complex terrain, improving the reliability and security of UAV communication systems, and has significant application value and market prospects.

Description

Technical Field

[0001] This invention belongs to the field of wireless communication, and particularly relates to a method for dynamic compensation of polarization signals in low-altitude complex terrain, a UAV communication device and system. Background Technology

[0002] Low-altitude unmanned aerial vehicles (UAVs) are playing an increasingly important role in military reconnaissance, emergency rescue, logistics delivery, and urban security. However, in complex terrains (such as mountainous areas, urban canyons, and forested areas), severe polarization signal mismatch and distortion occur due to terrain obstruction, multipath fading, and uneven local electromagnetic environments, leading to a significant deterioration in the quality of UAV communication links. Furthermore, traditional polarization compensation methods mostly employ fixed compensation strategies, which cannot track the dynamic changes in channel polarization under complex terrain conditions in real time, thus failing to meet the stringent requirements of high reliability, low bit error rate, and high security in practical applications.

[0003] Furthermore, while some current systems utilize edge computing technology for real-time data processing, they still suffer from issues such as data processing latency and insufficient algorithm robustness in dynamic polarization signal compensation. Therefore, there is an urgent need to propose an innovative method for real-time monitoring and compensation of dynamic polarization signals in low-altitude complex terrain environments, combined with UAV communication devices to achieve efficient, stable, and intelligent anti-interference communication. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention provides a method for dynamic compensation of polarization signals in low-altitude complex terrain, comprising:

[0005] S1. Real-time acquisition of channel polarization status data based on adaptive polarization antenna array and full-duplex SDR platform;

[0006] S2. Calculate the signal-to-noise ratio of the channel based on the polarization state data;

[0007] S3. Based on the signal-to-noise ratio of the channel, set a set of candidate modulation schemes, and construct a constraint function based on the set of candidate modulation schemes;

[0008] S4. Based on the constraint function, a global optimization algorithm is used to solve for the optimal compensation parameters and modulation scheme. Preferably, the process of collecting channel polarization state data further includes:

[0009] Collect the channel polarization state data, and construct the channel polarization matrix based on the channel polarization state data;

[0010] The polarization mismatch, polarization correlation coefficient, and polarization ellipse parameters are calculated based on the channel polarization matrix.

[0011] Preferably, the expression for the channel polarization matrix is:

[0012]

[0013] Among them, h ij H represents the channel gain between different polarization directions, and H is the channel polarization matrix.

[0014] The expression for the polarization mismatch is:

[0015]

[0016] XPD represents polarization mismatch.

[0017] Preferably, the expression for the polarization correlation coefficient is:

[0018]

[0019] Where ρ is the polarization correlation coefficient.

[0020] Preferably, the polarization ellipse parameters include polarization angle and polarization ellipticity;

[0021] The expression for calculating the polarization angle is:

[0022]

[0023] Where θ is the polarization angle;

[0024] The expression for calculating the polarization ellipticity is as follows:

[0025]

[0026] Where ∈ represents the polarization ellipticity.

[0027] Preferably, the expression for the constraint function is:

[0028]

[0029] Where BER(P,M) is the bit error rate, P is the polarization compensation parameter, M is the modulation scheme, and β is the preset bit error rate threshold.

[0030] The constraint conditions of the constraint function are:

[0031] BER(P,M)≤β.BER(P,M)≤β.

[0032] To address the aforementioned technical problems, the present invention also provides a UAV communication device for dynamic compensation of polarization signals in low-altitude complex terrain, comprising:

[0033] An adaptive polarization antenna array is used to switch between multiple polarization modes and control voltage or current to adjust the polarization direction and ellipticity.

[0034] A full-duplex SDR platform for acquiring bandwidth signals and performing real-time digital signal processing;

[0035] High-performance edge computing units are used to run polarization state monitoring, dynamic compensation algorithms, interference prediction, and cluster collaborative decision-making algorithms.

[0036] A high-precision positioning module is used to provide positioning information;

[0037] The secure storage and communication module is used to protect sensitive data based on physically isolated storage areas and encryption algorithms.

[0038] Preferably, the adaptive polarization antenna array is based on a four-element polarization antenna array with an adjustable frequency range of 2-6 GHz, the frequency switching time is no more than 5 μs, and each antenna element supports independent polarization control.

[0039] The full-duplex SDR platform is based on an RF front-end, a high-speed ADC / DAC, an FPGA, and a high-performance DSP chip.

[0040] The high-performance edge computing unit is based on a quad-core ARM Crtex-A76 processor, an AI accelerator, and an FPGA logic unit.

[0041] The high-precision positioning module is based on RTK-GPS and a nine-axis IMU;

[0042] The secure storage and communication module is based on a hardware encryption chip.

[0043] To address the aforementioned technical problems, this invention also provides a UAV communication system with dynamic compensation for polarization signals in low-altitude complex terrain, comprising:

[0044] A real-time polarization state monitoring module is used to dynamically adjust the acquisition frequency according to the channel change rate, and

[0045] The dynamic compensation algorithm module is used to automatically adjust the compensation parameters based on the real-time collected polarization state data;

[0046] A predictive interference identification module is used to identify and predict interference types and trends.

[0047] The cluster collaborative decision-making module is used to realize collaborative anti-interference decision-making for UAV clusters;

[0048] The secure communication module provides high-strength encrypted communication to ensure data transmission security.

[0049] Compared with the prior art, the present invention has the following advantages and technical effects:

[0050] 1. Dynamic Polarization Compensation: This invention overcomes the limitations of fixed compensation methods by real-time acquisition of polarization states and dynamic optimization of compensation parameters, achieving optimal matching of channel polarization states. Experimental verification shows that compared with traditional methods, this invention can reduce the bit error rate by more than 60% in complex terrain, significantly improving communication link stability. Especially in complex environments such as mountains and forests, the bit error rate of this method is only 38% of that of traditional methods. This is because this invention can track and compensate for rapidly changing polarization mismatches in complex terrain in real time, which traditional fixed compensation methods cannot do.

[0051] 2. Enhanced Environmental Adaptability: Employing predictive interference identification and federated learning collaborative decision-making, the system can proactively and rapidly respond to environmental changes, dynamically adjusting modulation schemes and coding strategies. Experiments show that the convergence time of the proposed method after abrupt polarization state changes is no more than 2 seconds, while traditional methods require more than 5 seconds to stabilize, significantly improving the system's adaptability in dynamic environments. Furthermore, by predicting interference changes, the system can adjust the modulation and coding scheme in advance, further reducing the impact of interference and improving communication reliability.

[0052] 3. Cluster Collaboration Advantages: Through distributed collaborative decision-making, multiple UAV nodes jointly optimize anti-interference strategies, effectively reducing the risk of single-point failures. When the cluster size expands from 2 nodes to 16 nodes, the overall anti-interference capability of the system improves by approximately 140%, and the impact of a single node failure on the overall system performance is reduced to less than 25% of the original. This is because federated learning can aggregate environmental awareness information from multiple nodes to form a more comprehensive interference situation map, thereby achieving a better collaborative anti-interference strategy.

[0053] 4. Low Power Consumption and High Security: The lightweight algorithm designed in this invention reduces computational complexity by approximately 65% ​​compared to traditional solutions, while reducing power consumption by approximately 42% to maintain equivalent performance. Simultaneously, the integrated secure communication module effectively prevents man-in-the-middle attacks and replay attacks, increasing security strength to 128-bit equivalent security, meeting military-grade security requirements. The low-power design makes this invention more suitable for UAV platforms, extending UAV endurance. The high-security design ensures data and control security of the UAV communication link. Attached Figure Description

[0054] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings:

[0055] Figure 1 This is a system architecture diagram according to an embodiment of the present invention;

[0056] Figure 2 This is a diagram of the internal system architecture of a drone node according to an embodiment of the present invention;

[0057] Figure 3 This is a flowchart illustrating the core information of the system in an embodiment of the present invention.

[0058] Figure 4 This is a schematic diagram comparing the polarization compensation effects under different terrains according to an embodiment of the present invention;

[0059] Figure 5 This is a schematic diagram illustrating the real-time performance test of polarization adaptive capability according to an embodiment of the present invention.

[0060] Figure 6 This is a schematic diagram comparing the system throughput under different interference intensities according to an embodiment of the present invention;

[0061] Figure 7 This is a comparison chart of bit error rate and throughput in an embodiment of the present invention;

[0062] Figure 8 This is a flowchart of a method according to an embodiment of the present invention. Detailed Implementation

[0063] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0064] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0065] Example 1

[0066] like Figure 1-3 and Figure 8 As shown, this embodiment provides a method for dynamic compensation of polarization signals in low-altitude complex terrain, including:

[0067] Real-time acquisition of channel polarization status data based on adaptive polarization antenna array and full-duplex SDR platform;

[0068] The signal-to-noise ratio of the channel is calculated based on the polarization state data;

[0069] A set of candidate modulation schemes is set based on the signal-to-noise ratio of the channel, and a constraint function is constructed based on the set of candidate modulation schemes;

[0070] Based on the constraint function, a global optimization algorithm is used to solve for the optimal compensation parameters and modulation scheme.

[0071] Based on the channel polarization matrix HH and interference parameter II collected in real time under low-altitude complex terrain, the optimal polarization compensation parameter P*P^* and the optimal modulation scheme M*M^* are calculated to maximize the system channel capacity and keep the bit error rate below the preset threshold.

[0072] Step 1: Polarization index calculation:

[0073] Given the channel polarization matrix:

[0074]

[0075] Among them, h HH h HV h VH h VV , represents the channel gain between different polarization directions, and H is the channel polarization matrix;

[0076] Calculate the polarization mismatch (XPD):

[0077]

[0078] XPD represents polarization mismatch.

[0079] Calculate the polarization correlation coefficient:

[0080]

[0081] Where ρ is the polarization correlation coefficient.

[0082] Calculate the polarization ellipse parameters:

[0083] The polarization ellipse parameters include polarization angle and polarization ellipticity;

[0084] The expression for calculating the polarization angle is:

[0085]

[0086] Where θ is the polarization angle;

[0087] The expression for calculating the polarization ellipticity is as follows:

[0088]

[0089] Where ∈ represents the polarization ellipticity.

[0090] Step 2: Channel Quality Indicator Assessment

[0091] Calculate the channel signal-to-noise ratio (SNR):

[0092]

[0093] Define the channel quality metric Q as a function:

[0094] Q = f(XPD, ρ) p (SNR);

[0095] Step 3: Modulation method selection:

[0096] Define a set of candidate modulation schemes {M} i (e.g., BPSK, QPSK, 16-QAM, 64-QAM, etc.), and calculate the bit error rate BER(P,M) for each modulation scheme. i Requirements must be met:

[0097] BER(P, Mi)≤β;

[0098] Where β is the preset bit error rate threshold.

[0099] Step 4: Optimize the problem-solving process:

[0100] Construct the optimization objective function:

[0101]

[0102] Where BER(P,M) is the bit error rate, P is the polarization compensation parameter, M is the modulation scheme, and β is the preset bit error rate threshold.

[0103] The constraint conditions of the constraint function are:

[0104] BER(P,M)≤β.BER(P,M)≤β.

[0105] The optimal compensation parameters P* and modulation scheme M* are solved by gradient descent or other global optimization algorithms.

[0106] This embodiment uses the gradient descent algorithm.

[0107] initialization:

[0108] Set the initial polarization parameter P (0) With modulation method M (0) ;

[0109] Read the location of the interference source p i And the polarization matrix H.

[0110] Calculate the objective function and its gradient:

[0111] Calculate BER(P) (k) M (k) ) and log2(M (k) );

[0112] If BER > β, then perform constraint correction;

[0113] beg

[0114] Parameter update:

[0115]

[0116] Where α is the learning rate, and round indicates that the modulation order should take discrete values ​​(2, 4, 16, 64, etc.).

[0117] Convergence criterion: If P(k+1)-P(k)<δ and M(k+1)=M(k), or the number of iterations exceeds the upper limit, then output (P*,M*).

[0118] Data flow:

[0119] The interference source localization module periodically outputs P i ;

[0120] The status monitoring module updates parameters such as H, XPD, and ρ;

[0121] The optimization algorithm performs gradient descent and other global searches on the edge computing platform, and sends updates (P). * M * ).

[0122] Real-time performance:

[0123] Edge computing platforms feature CPU+GPU / FPGA collaboration and can complete an iteration in the 10ms range.

[0124] The optimization cycle (e.g., 100ms) can be set for parameter updates to ensure a rapid response to disturbed environments.

[0125] Elastic polarization switching:

[0126] Once obtained (P) * M * The polarization mode (θ) of the dual-polarized antenna is adjusted by optimizing the switching control model. * ,∈ * And configure the modulation method M * .

[0127] The switching time is controlled in the range of 5μs to 10μs (depending on the performance of the antenna array and RF switch), which has little impact on communication continuity.

[0128] Step 5: Output the decision results

[0129] Returns the optimal parameters P* and the optimal modulation scheme M*, which are used to dynamically adjust the UAV communication transmission parameters.

[0130] Performance verification experiment:

[0131] To verify the effectiveness of the algorithm in this embodiment, three sets of comparative experiments were designed, and detailed performance data were collected:

[0132] Experiment 1: Comparison of polarization compensation effects under different terrains

[0133] This embodiment tests the performance difference between the polarization compensation algorithm of this embodiment and the traditional fixed compensation method under three typical terrains (open terrain, urban environment, and mountain forest). Experimental parameters were set as follows: carrier frequency 2.4 GHz, signal bandwidth 20 MHz, modulation scheme QPSK, channel models using the CST2100 urban microcell model and the ITU-RP.1411 mountain terrain model, and a fixed signal-to-noise ratio of 20 dB. 1000 independent channel simulations were performed for each terrain type, and the bit error rate was statistically analyzed. Figure 4 As shown. Figure 4 The method of this embodiment is shown to significantly outperform traditional methods in terms of bit error rate under three different terrains, with the effect being more pronounced in complex terrains.

[0134] Experiment 2: Real-time performance test of polarization adaptive capability:

[0135] This embodiment tests the tracking ability of different algorithms for polarization changes under rapidly changing channel conditions. In the experiment, a simulated UAV flying through complex terrain caused the channel polarization state to change rapidly over time. The polarization state change model adopted a random walk model, with the polarization angle and ellipticity randomly jumping within a certain range at a jump rate of 10Hz. The polarization tracking errors of the method in this embodiment, the traditional fixed compensation method, and the rule-based adaptive compensation method (switching polarization modes according to an XPD threshold) are compared. The tracking error is defined as the Euclidean distance between the actual polarization state and the compensated polarization state.

[0136] Figure 5 This paper demonstrates the advantages of the proposed method in terms of polarization tracking error convergence speed and steady-state accuracy. The embodiment compares the throughput performance of the proposed system with that of the conventional system under different interference intensities (signal-to-interference ratio from -10dB to 20dB). In the experiment, narrowband interference was introduced, with an interference bandwidth of 10% of the signal bandwidth and a randomly varying center frequency. The throughput of the proposed system (using dynamic polarization compensation and adaptive modulation), the conventional system (fixed polarization compensation and fixed modulation QPSK), and the theoretical upper limit throughput (ideal polarization matching and optimal modulation) are compared. The throughput calculation formula is: Throughput = Bandwidth multiplied by log2(M) × (1-BER), where M is the modulation order and BER is the bit error rate. Figure 6 The method of this embodiment can achieve significant throughput improvement under various interference environments, especially under moderate interference environments, the improvement exceeds 100%.

[0137] The following is another experimental procedure in this embodiment:

[0138] Experimental objective:

[0139] This study verifies the anti-interference performance of the proposed UAV dual-polarized antenna elastic switching method based on spatial localization of interference sources (EPS method) in low-altitude complex terrain environments, and compares it with the traditional fixed polarization method (VP method) and the polarization switching method based on signal strength detection (RSSI-based PS method). Key performance indicators include bit error rate (BER), system throughput, and spectral efficiency.

[0140] Experimental setup and simulation parameters:

[0141] Experimental scenario: Simulating complex low-altitude terrain environment, using urban and mountain environment models (based on CST2100 urban microcell model and ITU-RP.1411 mountain model respectively).

[0142] Carrier frequency: 2.4GHz

[0143] Signal bandwidth: 20MHz

[0144] Modulation method: QPSK (extendable to other modulation schemes, such as 16-QAM, 64-QAM)

[0145] Signal-to-noise ratio (SNR): 20dB (desired signal under interference-free conditions)

[0146] Interference type: Single narrowband interference, interference bandwidth is 10% of the signal bandwidth, center frequency varies randomly.

[0147] Interference signal power (INR): ranging from 0dB to 20dB

[0148] Number of samplings: 1000 Monte Carlo simulations were performed under each condition, and statistical data were collected to calculate the mean and standard deviation.

[0149] Performance index calculation:

[0150] Bit Error Rate (BER): The average bit error rate for each method, calculated through simulation.

[0151] Throughput: The calculation formula is as follows

[0152] Throughput = Bandwidth × Bitrate × (1 - BER)

[0153] Spectral efficiency:

[0154] Data statistical methods:

[0155] For each interference intensity condition, the average bit error rate, throughput, and spectral efficiency were calculated for all simulation results, and the standard deviation was statistically analyzed to reflect data fluctuations. Standard mean and standard deviation formulas were used for data statistics.

[0156] Comparative experimental design:

[0157] Fixed Vertical Polarization (VP Method): The UAV antenna is fixed in a vertical polarization mode.

[0158] Polarization switching based on signal strength detection (RSSI-based PS method): dynamically switches between horizontal and vertical polarization based on the received signal strength.

[0159] The elastic polarization switching method (EPS method) in this embodiment is based on the spatial positioning information of the interference source, intelligently decides the optimal polarization mode and performs elastic switching.

[0160] The methods were compared through simulation under the same experimental conditions. Curves were plotted showing the changes in bit error rate, throughput, and spectral efficiency with respect to INR. The average value and standard deviation of each indicator were also marked on the graphs.

[0161] Experiment code description:

[0162] The code defines three methods: fixed vertical polarization (VP), polarization switching based on RSSI detection (RSSI), and elastic polarization switching (EPS) of the present invention.

[0163] Each method was subjected to 1000 Monte Carlo simulations under different interference signal power ratios (INRs) to calculate the average bit error rate and standard deviation, and the throughput was calculated accordingly.

[0164] Finally, two charts were drawn: the left side is a comparison chart of bit error rate, and the right side is a comparison chart of throughput. Both charts include error bars to show the data fluctuations.

[0165] The experimental results are described as follows: Figure 7 As shown:

[0166] Figure 7 The results show that under strong interference (INR>10dB), the average bit error rate of the EPS method is significantly lower than that of the VP and RSSI methods, and the throughput is higher.

[0167] For example, at INR=15dB, the average BER of the EPS method is less than 0.06, while the VP method may reach 0.08, resulting in a throughput improvement of over 100%.

[0168] The small error bar (standard deviation) indicates that the data is stable, verifying the robustness and real-time advantages of this method.

[0169] This embodiment has the following technical effects: 1. Dynamic polarization compensation: This embodiment overcomes the limitations of fixed compensation methods by real-time acquisition of polarization state and dynamic optimization of compensation parameters, achieving optimal matching of channel polarization state. Experimental verification shows that compared with traditional methods, this embodiment can reduce the bit error rate by more than 60% in complex terrain, and significantly improve the stability of communication links. Especially in complex environments such as mountains and forests, the bit error rate of this method is only 38% of that of traditional methods. This is because this embodiment can track and compensate for rapidly changing polarization mismatch in complex terrain in real time, which traditional fixed compensation methods cannot do.

[0170] 2. Enhanced Environmental Adaptability: Employing predictive interference identification and federated learning collaborative decision-making, the system can proactively and rapidly respond to environmental changes, dynamically adjusting modulation schemes and coding strategies. Experiments show that the convergence time of the method in this embodiment after abrupt polarization state changes is no more than 2 seconds, while traditional methods require more than 5 seconds to stabilize, significantly improving the system's adaptability in dynamic environments. Furthermore, by predicting interference changes, the system can adjust the modulation and coding scheme in advance, further reducing the impact of interference and improving communication reliability.

[0171] 3. Cluster Collaboration Advantages: Through distributed collaborative decision-making, multiple UAV nodes jointly optimize anti-interference strategies, effectively reducing the risk of single-point failures. When the cluster size expands from 2 nodes to 16 nodes, the overall anti-interference capability of the system improves by approximately 140%, and the impact of a single node failure on the overall system performance is reduced to less than 25% of the original. This is because federated learning can aggregate environmental awareness information from multiple nodes to form a more comprehensive interference situation map, thereby achieving a better collaborative anti-interference strategy.

[0172] 4. Low Power Consumption and High Security: The lightweight algorithm designed in this embodiment reduces computational complexity by approximately 65% ​​compared to traditional solutions, while reducing power consumption by approximately 42% to maintain equivalent performance. Simultaneously, the integrated secure communication module effectively prevents man-in-the-middle attacks and replay attacks, increasing security strength to 128-bit equivalent security, meeting military-grade security requirements. The low-power design makes this embodiment more suitable for UAV platforms, extending the UAV's flight time. The high-security design ensures data and control security of the UAV communication link.

[0173] Example 2

[0174] like Figure 1-3 As shown, this embodiment provides a UAV communication device for dynamic compensation of polarization signals in low-altitude complex terrain, including:

[0175] Adaptive Polarization Antenna Array: Employing a four-element polarization antenna array with an adjustable frequency range of 2-6 GHz, it enables rapid switching between multiple polarization modes (linear, circular, and elliptic polarization) with a frequency switching time not exceeding 5 μs. Each antenna element supports independent polarization control, adjusting the polarization direction and ellipticity by controlling voltage or current. The antenna array features a compact design, with dimensions no larger than 10cm x 10cm x 5cm and a weight not exceeding 200g, meeting the payload requirements of UAVs.

[0176] Full-duplex SDR platform: Supports 100MHz bandwidth signal acquisition and real-time digital signal processing, with a sampling rate up to 200MSPS and a dynamic range of no less than 80dB. The SDR platform integrates an RF front-end, high-speed ADC / DAC, FPGA, and high-performance DSP chip, supports software-defined radio functions, and allows for flexible configuration of modulation / demodulation, channel coding, and other communication protocols. The platform adopts a full-duplex design, supporting simultaneous data transmission and reception to meet real-time communication requirements.

[0177] High-performance edge computing unit: Integrates a quad-core ARM Crtex-A76 processor (2.2GHz), a dedicated AI accelerator (supporting 8TPS computing power), and an FPGA logic unit (equivalent to 1 million logic gates), with power consumption controlled below 10W. The edge computing unit is responsible for running core software modules such as polarization state monitoring, dynamic compensation algorithms, interference prediction, and cluster collaborative decision-making. The AI ​​accelerator is used to accelerate the inference computation of deep learning models, and the FPGA logic unit is used to implement high-speed data processing and hardware acceleration.

[0178] High-precision positioning module: Integrating RTK-GPS and a nine-axis IMU, it achieves centimeter-level precise positioning for the UAV, provides position information for the compensation algorithm, and supports inertial navigation in the event of GNSS signal interruption. The positioning module outputs position, velocity, attitude, and other information, with an update frequency of up to 100Hz. In the event of GNSS signal loss, the IMU can provide short-term high-precision inertial navigation to ensure the continuity of positioning information.

[0179] Secure Storage and Communication Module: Employs physically isolated storage areas and high-strength encryption algorithms to protect sensitive data and achieve high-strength encrypted communication. The secure storage module uses hardware encryption chips and tamper-proof design to prevent data leakage and malicious attacks. The secure communication module supports SM2 / SM3 / SM4 Chinese national cryptographic algorithms and AES-256 international standard encryption algorithms to ensure the confidentiality and integrity of data transmission.

[0180] The system consists of a low-altitude unmanned aerial vehicle (UAV) swarm, a ground control station (GCS), and a wireless communication channel. Each UAV node is equipped with the communication device described in this embodiment to achieve environmental monitoring, polarization state estimation, dynamic compensation, and collaborative decision-making. The ground control station is responsible for mission planning, command and control, data processing, and safety management. The UAV swarm and the ground control station exchange data via the wireless communication channel.

[0181] The UAV node consists of an adaptive polarization antenna, an SDR platform, a high-performance edge computing unit, and various functional modules. These modules are connected through an internal high-speed bus (such as PCIe or AXI) to collaboratively complete signal acquisition, processing, compensation, and communication tasks.

[0182] Example 3

[0183] This embodiment provides a UAV communication system with dynamic compensation for polarization signals in low-altitude complex terrain, including:

[0184] Real-time polarization state monitoring module: Acquisition frequency: adjustable from 100Hz to 1kHz, dynamically adjusted according to the channel change rate. The acquisition frequency is increased when channel changes are drastic and decreased when they are mild, balancing real-time performance and power consumption.

[0185] Polarization parameter estimation accuracy: XPD error < 0.5 dB, phase error < 2°. An estimation algorithm based on minimum mean square error (MMSE) is employed, combined with Kalman filtering for time-domain smoothing, improving estimation accuracy and robustness.

[0186] Processing latency: <5ms, meeting real-time compensation requirements. The module employs optimized algorithms and hardware acceleration to reduce processing latency.

[0187] Dynamic compensation algorithm module:

[0188] Compensation parameter update rate: up to 200Hz, matching the acquisition frequency of the polarization state monitoring module.

[0189] Compensation accuracy: polarization angle error <3°, polarization ellipticity error <5%. An iterative optimization algorithm is used to continuously adjust the compensation parameters to approach the optimal compensation effect.

[0190] Algorithm complexity: O(n·lgn), ensuring real-time performance. Efficient algorithms such as the Fast Fourier Transform (FFT) are employed to reduce computational complexity.

[0191] Predictive interference identification module:

[0192] Interference type identification: Supports 10 typical interference types, including narrowband interference, frequency sweep interference, and impulse interference. A deep learning-based interference classification model is employed, utilizing a convolutional neural network (CNN) to extract interference features for high-precision identification.

[0193] Including but not limited to:

[0194] The spectral characteristics of interference signals, such as center frequency, bandwidth, and power spectral density, are analyzed. Through spectral analysis, the distribution of interference signals in the frequency domain is accurately identified, and the spectral characteristics of different types of interference are distinguished.

[0195] The time-domain waveform characteristics of the interference signal, such as pulse width, pulse repetition frequency, modulation method, etc.; analyze the variation law of the interference signal in the time domain, and distinguish the interference types with different time-domain characteristics, such as pulse interference and continuous wave interference.

[0196] The polarization characteristics of interference signals, such as polarizability and polarization angle, are used to distinguish between co-polarized interference and cross-polarized interference, thereby improving the ability to identify interference with specific polarizations.

[0197] Spatial characteristics of interference signals, such as angle of arrival (AOA); using array signal processing techniques, the direction of arrival of interference signals is estimated to help locate the interference source and distinguish interference signals from different directions.

[0198] High-precision identification: A deep learning-based interference classification model utilizes a convolutional neural network (CNN) to extract interference features, achieving high-precision identification. The specific CNN model structure is as follows:

[0199] Input layer: Receives preprocessed interference signal data, such as time-domain sampled data or frequency-domain spectrum data. The dimensions of the input data can be adjusted according to the actual application scenario and the dimensions of the interference characteristics; for example, a fixed-length time-domain sampled sequence or a two-dimensional spectrum can be input.

[0200] Convolutional Layers: Multiple layers of convolutional layers are stacked to automatically extract deep features from interference signals.

[0201] The first convolutional layer uses a small kernel size (e.g., 3x3 or 5x5) to extract local temporal or frequency domain features, such as edges and textures.

[0202] Subsequent convolutional layers: As network depth increases, the kernel size can be gradually increased to extract more abstract and global features. Different numbers and sizes of convolutional kernels can be used to increase the diversity of feature extraction.

[0203] Activation function: Each convolutional layer is usually followed by an activation function, such as ReLU (Rectified Linear Unit) or LeakyReLU, to increase the non-linear expressive power of the network.

[0204] Pooling layers: Pooling layers, such as max pooling or average pooling, are inserted between or after convolutional layers to reduce the dimensionality of feature maps, reduce computation, and enhance the translation invariance of features.

[0205] Fully Connected Layers: After multiple convolutional and pooling layers, several fully connected layers are connected to map the extracted features to the class space.

[0206] Hidden layers: Fully connected layers can contain multiple hidden layers, increasing the network's learning ability.

[0207] Activation functions: Fully connected layers can also use activation functions such as ReLU or LeakyReLU.

[0208] Output layer: The last fully connected layer serves as the output layer, employing the Softmax activation function to output the probability values ​​corresponding to each interference type. The output dimension corresponds to the number of interference types (e.g., 10 typical interferences).

[0209] Training method: The network is trained using labeled perturbation sample data. Backpropagation algorithm and optimizer (such as Adam or SGD) are used to optimize the network parameters. Cross-entropy loss function can be selected as the loss function.

[0210] Model parameters, such as the number of convolutional kernels, kernel size, pooling layer type and parameters, number of fully connected layers and nodes, can be adjusted and optimized according to the actual application scenario and data characteristics.

[0211] Recognition accuracy: >92% (under SINR>-10dB conditions). It maintains a high recognition rate even under low signal-to-noise ratio conditions.

[0212] Prediction time window: adjustable from 50ms to 500ms, depending on the characteristics of the disturbance and the system response speed.

[0213] Recognition accuracy: >92% (under SINR>-10dB conditions). It maintains a high recognition rate even under low signal-to-noise ratio conditions.

[0214] Prediction time window: adjustable from 50ms to 500ms, depending on the characteristics of the disturbance and the system response speed.

[0215] Cluster collaborative decision-making module:

[0216] Node size: Supports up to 32 drone nodes working together. Employs a distributed architecture for easy expansion.

[0217] Federated learning iteration cycle: Adjustable from 100ms to 1s, dynamically adjusted according to channel change rate and computing resources. Communication overhead: <10kbps / node, low communication overhead, suitable for bandwidth-constrained UAV communication networks. Employs compression coding and sparse update techniques to reduce communication overhead.

[0218] Secure communication module: Encryption algorithms: Supports national cryptographic standards such as SM2 / SM3 / SM4 and AES-256, as well as international standards. Encryption algorithms can be flexibly configured to meet different security level requirements.

[0219] Key update frequency: Configurable, 10s-600s. Supports both periodic key updates and event-triggered key updates, improving key security.

[0220] Encryption computation latency: <1ms, with minimal impact on real-time communication. Low-latency encryption is achieved through hardware acceleration and optimized algorithms.

[0221] Secure communication module:

[0222] Encryption algorithms: Supports Chinese national cryptographic standards such as SM2 / SM3 / SM4 and AES-256, as well as international standards. Encryption algorithms can be flexibly configured to meet different security level requirements.

[0223] Key update frequency: Configurable, 10s-600s. Supports both periodic key updates and event-triggered key updates, improving key security.

[0224] Encryption computation latency: <1ms, with minimal impact on real-time communication. Low-latency encryption is achieved through hardware acceleration and optimized algorithms.

[0225] The communication protocol is:

[0226] Polarization State Negotiation Protocol: This protocol defines the reference signal structure (preamble + training sequence), the polarization state feedback format (compressed representation of the polarization matrix H), and the compensation parameter negotiation process (three-way handshake mechanism). The reference signal uses a pseudo-random sequence with good autocorrelation and cross-correlation. The polarization state feedback format employs differential coding and quantization compression to reduce feedback overhead. This protocol primarily operates between the polarization processing module and the control module. The polarization processing module is responsible for adjusting the polarization state based on the negotiation results, while the control module is responsible for initiating and managing the negotiation process, and storing and transmitting the negotiation parameters.

[0227] Interference Information Sharing Protocol: This protocol supports efficient transmission of interference feature data between nodes, employing differential coding. It primarily operates between the interference identification module and the information fusion module (or control module). The interference identification module is responsible for detecting and extracting interference features, and then transmitting the data to the information fusion module (or control module) via this protocol for centralized processing and analysis, enabling global interference situational awareness and coordinated interference suppression. If the system lacks a dedicated information fusion module, the control module is responsible for receiving and processing the interference information.

[0228] Federated Learning Collaboration Protocol: This protocol defines the model training, parameter aggregation, and decision distribution processes, including node weighting mechanisms, differential privacy protection, and anomaly node detection. The node weighting mechanism dynamically adjusts weights based on node contributions. The differential privacy protection mechanism prevents the leakage of node data privacy during model training. The anomaly node detection mechanism identifies malicious and faulty nodes, improving system robustness. This protocol primarily operates between the model training module and the control module. The model training module is responsible for executing the federated learning training tasks, while the control module coordinates and manages the entire federated learning process, including parameter aggregation, model distribution, node management, and policy decision-making.

[0229] Secure Communication Protocol: Ensures the security of confidential key distribution, storage, and management, supporting authentication, session protection, and integrity verification. Key distribution employs a key exchange protocol based on Public Key Infrastructure (PKI). Authentication utilizes digital certificates and two-way authentication mechanisms. Session protection employs encrypted tunnels and secure transport protocols (such as TLS / SSL). Integrity verification employs Message Authentication Code (MAC) and digital signature technology. This protocol primarily operates between the security module and the control module, as well as between all modules within the system that require secure communication. The security module is responsible for providing various security functions, such as key management, authentication, data encryption, and integrity verification. The control module is responsible for configuring and managing security policies and coordinating the secure communication needs of various modules. Other modules within the system that require secure communication, such as the data acquisition module, signal processing module, and information fusion module, must collaborate with the security module and exchange data according to the secure communication protocol.

[0230] The above are merely preferred embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for dynamic compensation of polarization signals in low-altitude complex terrain, the method being applied to a UAV node, the UAV node comprising an adaptive polarization antenna array, a full-duplex SDR platform and a high-performance edge computing unit, the adaptive polarization antenna array being used to switch between multiple polarization modes and control voltage or current to adjust the polarization direction and ellipticity; The full-duplex SDR platform is used to acquire bandwidth signals and perform real-time digital signal processing. The high-performance edge computing unit is used to run polarization state monitoring, dynamic compensation algorithms, interference prediction, and cluster collaborative decision-making algorithms; characterized in that the method includes: S1. Real-time acquisition of channel polarization state data based on an adaptive polarization antenna array and a full-duplex SDR platform; the process of acquiring channel polarization state data further includes: Collect the channel polarization state data, and construct the channel polarization matrix based on the channel polarization state data; Calculate the polarization mismatch, polarization correlation coefficient, and polarization ellipse parameters based on the channel polarization matrix; S2. Calculate the signal-to-noise ratio of the channel based on the polarization state data; S3. Based on the signal-to-noise ratio of the channel, set a set of candidate modulation schemes, and construct a constraint function based on the set of candidate modulation schemes; S4. Based on the constraint function, the optimal compensation parameters and modulation scheme are solved using the gradient descent algorithm; the expression of the constraint function is: in, β is the constraint function; BER(P,M) is the bit error rate, P is the polarization compensation parameter, M is the modulation scheme, and β is the preset bit error rate threshold. The constraint conditions of the constraint function are: BER(P, M)≤β; initialization: Set the initial polarization parameter P (0) With modulation method M (0) ; Read the location p of the interference source i and polarization matrix H; Calculate the objective function and its gradient: Calculate BER(P) (k) M (k) ) and log2(M (k) ); If BER(P) (k) M (k) If β > 0, then constraint correction is performed; beg Parameter update: Where α is the learning rate, and round indicates that the modulation order needs to be a discrete value; Convergence criterion: If P (k+1) -P (k) <δ and M (k+1) =M (k) If the number of iterations exceeds the upper limit, then output (P) * M * ); The optimization algorithm performs gradient descent on the edge computing platform and sends updates (P). * M * ); Once obtained (P) * M * The polarization mode (θ) of the dual-polarized antenna is adjusted by optimizing the switching control model. * ,∈ * And configure the modulation method M * P * These are the optimal compensation parameters.

2. The method according to claim 1, characterized in that, The expression for the channel polarization matrix is: Among them, h HH h HV h VH h VV , represents the channel gain between different polarization directions, and H is the channel polarization matrix; The expression for the polarization mismatch is: XPD represents polarization mismatch.

3. The method according to claim 2, characterized in that, The expression for the polarization correlation coefficient is: Where, ρ p This represents the polarization correlation coefficient.

4. The method according to claim 3, characterized in that, The polarization ellipse parameters include polarization angle and polarization ellipticity; The expression for calculating the polarization angle is: Where θ is the polarization angle; The expression for calculating the polarization ellipticity is as follows: Where ∈ represents the polarization ellipticity.

Images