A Low-Altitude Communication Anti-jamming System and Channel Allocation Method Combining Polarization Diversity and Frequency Diversity

Abstract

This invention provides a low-altitude communication anti-interference system and channel allocation method combining polarization diversity and frequency diversity, belonging to the field of low-altitude communication technology. The system consists of a multi-polarization antenna array, a broadband radio frequency front-end, a digital signal processing unit, an adaptive controller, and an environmental sensing module. The core of this invention is a polarization-frequency joint channel allocation method, which achieves global optimized allocation of channel resources by constructing a polarization-frequency two-dimensional channel quality matrix. The system adopts a multi-layer adaptive architecture, including a fast response layer and a deep optimization layer, balancing real-time performance and optimization level. Furthermore, this invention proposes a dynamic adjustment mechanism for diversity parameters to resist intelligent interference and an interference identification method based on polarization characteristics, significantly improving the system's communication reliability and anti-interference capability in complex electromagnetic environments. It can maintain extremely low bit error rate communication even in strong interference environments.

Description

Technical Field

[0001] This invention relates to the field of low-altitude communication technology, specifically to a low-altitude communication anti-interference system and channel allocation method that combines polarization diversity and frequency diversity. Background Technology

[0002] Low-altitude communication refers to communication activities within an altitude range of approximately 0.1-2 kilometers above the ground, and is widely used in fields such as military reconnaissance, emergency rescue, environmental monitoring, precision agriculture, and border patrol. However, the low-altitude communication environment is extremely complex and volatile. To cope with this complex environment, diversity techniques are widely employed. The core idea of ​​diversity techniques is to improve the reliability of signal reception through diversity.

[0003] Current polarization diversity and frequency diversity correlation techniques have the following main drawbacks in low-altitude communication anti-jamming applications:

[0004] 1. Insufficient Dimensional Synergy: Although existing technologies combine polarization diversity and frequency diversity in some systems, they typically employ simple superposition or hierarchical designs. The two diversity techniques operate largely independently, each optimizing its own parameters, lacking a deep fusion and synergy mechanism. This prevents the system from fully utilizing the complementary advantages of the two diversity techniques, making it difficult to achieve a synergistic gain effect where 1+1>2.

[0005] 2. Inefficient channel allocation: Traditional frequency diversity systems typically only consider frequency domain characteristics (such as signal-to-noise ratio and interference intensity) for channel allocation, while polarization diversity systems mainly focus on polarization domain characteristics, lacking a unified multi-dimensional channel quality assessment mechanism. This fragmented channel assessment method leads to suboptimal resource allocation and low spectrum utilization efficiency, especially in resource-constrained low-altitude communication networks.

[0006] 3. Weak resistance to intelligent interference: Existing systems have some resistance to traditional interference (such as noise interference and narrowband interference), but their effectiveness is poor when facing intelligent interference (such as tracking interference, learning interference, and selective interference). Due to the lack of joint polarization domain and frequency domain interference feature analysis and defense mechanisms, the system is vulnerable to targeted attacks and has low security.

[0007] 4. Limited Adaptability: Existing technologies mostly employ decision-making mechanisms based on preset rules or thresholds, which are insufficiently adaptable to environmental changes. In low-altitude environments, factors such as multipath effects and atmospheric refraction cause rapid changes in channel characteristics, while current systems adapt slowly and struggle to adjust parameter configurations in a timely manner, leading to communication interruptions or performance degradation. Therefore, this invention aims to construct an efficient, reliable, and highly adaptable low-altitude communication anti-interference system to meet the growing security communication needs of modern low-altitude platforms. Summary of the Invention

[0008] To address the shortcomings of existing technologies, the present invention aims to provide a low-altitude communication anti-interference system and channel allocation method that combines polarization diversity and frequency diversity, in order to solve the problems mentioned in the background. The present invention significantly improves the anti-interference capability, spectrum utilization and environmental adaptability of the communication system through joint optimization allocation and intelligent adaptive control of polarization and frequency resources.

[0009] To achieve the above objectives, the present invention provides a low-altitude communication anti-interference system combining polarization diversity and frequency diversity, comprising a multi-polarization antenna array, a broadband RF front-end, a digital signal processing unit, an adaptive controller, and an environment sensing module. The multi-polarization antenna array includes multiple adjustable polarization antenna elements, each of which can independently adjust its polarization state. The broadband RF front-end supports a multi-band parallel processing RF transceiver system, including a low-noise amplifier, mixer, filter, variable gain amplifier, and multi-channel ADC / DAC. The digital signal processing unit, based on a heterogeneous computing platform of FPGA and DSP, is used for signal modulation and demodulation, channel coding, digital filtering, and diversity merging. The adaptive controller is used for resource allocation strategy formulation, diversity parameter optimization, and system configuration management. The environment sensing module is used to collect and analyze electromagnetic environment information to provide a basis for decision-making.

[0010] Furthermore, the parameters of the multi-polarized antenna array include:

[0011] Polarization adjustment range: Full polarization coverage;

[0012] Polarization isolation: >25dB;

[0013] Polarization switching speed: <10 microseconds;

[0014] Antenna gain: 3-6 dBi;

[0015] The broadband radio frequency front-end features include:

[0016] Frequency coverage: 30MHz-3GHz;

[0017] Real-time bandwidth: Maximum 100MHz;

[0018] Number of channels: 4-8 independent channels;

[0019] Dynamic range: >90dB;

[0020] Noise figure: <3dB.

[0021] Furthermore, the digital signal processing unit includes the following modules: a real-time signal processing engine, a diversity merge processor, a channel estimation and tracking module, and an anti-interference algorithm accelerator.

[0022] Furthermore, the adaptive controller includes modules such as a polarization-frequency joint optimizer, a diversity parameter manager, an adaptive control strategy engine, and a system state monitor; the environment perception module includes modules such as a spectrum sensing unit, an interference feature extractor, a channel state estimator, and an environment change predictor.

[0023] A channel allocation method based on the above anti-interference system includes the following steps: after the channel state is collected, a polarization frequency channel quality matrix is ​​constructed, then the channel quality is evaluated, and it is determined whether there is interference. If there is interference, the interference characteristics are analyzed, the interference model is updated, an interference avoidance strategy is generated, then the optimization target and resource allocation are determined, diversity parameters are calculated, the configuration is updated, and the performance is monitored to determine whether the performance meets the requirements. If the requirements are met, the current configuration is maintained.

[0024] Furthermore, if there is no interference, the process of determining the optimization target is carried out directly; if the performance monitoring results do not meet the requirements, the channel state is collected again.

[0025] Furthermore, it also includes a polarization-frequency joint channel quality and diversity potential evaluation algorithm: first, a basic polarization-frequency channel quality index is defined, This metric incorporates key physical layer parameters:

[0026] Q basep,f =W1·[log2(1+SINR)] p,f )·(1–PER p,f )]+W2PDI p,f –W3D f

[0027] Among them: SINR p,f PER represents the signal-to-interference-plus-noise ratio at polarization state p and frequency f. p,f For the corresponding packet error rate, PDI p,f D is a polarization fidelity index. f is the channel time-varying index for frequency f, and w1, w2, w3 are non-negative weighting coefficients.

[0028] Furthermore, the steps to further evaluate diversity potential include:

[0029] a) Calculate the average correlation coefficient modulus between a specific polarization-frequency pair (p,f) and all other available resources (p',f') ≠ (p,f).

[0030]

[0031] Where N and M are the number of available polarization states and frequency channels, respectively, and ρ((p,f),(p',f')) is the Pearson correlation coefficient between the two resource pairs and the channel quality time series.

[0032] b) Define a time-domain correlation suppression factor D based on the average correlation coefficient modulus. p,f This factor rewards combinations that have low correlation with other resources:

[0033]

[0034] Where μ>0 is the sensitivity factor of diversity gain, and k≥1 is the exponential factor.

[0035] Furthermore, the final polarization-frequency joint utility index Q' p,f It combines basic channel quality with diversity potential:

[0036]

[0037] The beneficial effects of this invention are:

[0038] 1. The invention significantly enhances anti-interference capabilities: Through deep joint optimization of polarization diversity and frequency diversity, and quantitative utilization of diversity potential, the invention exhibits strong anti-interference capabilities in complex electromagnetic environments. Experimental results show that, compared with traditional single diversity techniques or simple superposition techniques, this system can reduce the bit error rate by more than 70% under the same interference intensity, and can still maintain extremely low bit error rate communication even in strong interference environments (where the interference signal power is 15dB higher than the useful signal). The improvement in defense against intelligent interference is particularly significant, with an escape success rate increased by more than 90%.

[0039] 2. The present invention significantly improves the efficiency of spectrum and polarization resource utilization: The resource allocation method based on joint utility assessment and risk perception optimizes the utilization of polarization-frequency two-dimensional resources, achieving higher overall resource efficiency while maintaining the same communication quality. Compared to traditional methods, joint resource utilization is increased by (e.g.) 45%, and system capacity or the number of accessible users increases by (e.g.) 60%. This is particularly important for resource-constrained low-altitude communication networks.

[0040] 3. This invention possesses extremely strong environmental adaptability: The hierarchical predictive adaptive control mechanism of this invention enables the system to respond quickly and accurately to environmental changes and adapt to various complex dynamic scenarios. The system's response time to changes in channel and interference is reduced by (for example) 80% compared to traditional feedback control systems, and the performance degradation in high-dynamic scenarios (such as high-speed maneuvering and complex urban electromagnetic environments) does not exceed (for example) 10%, while the performance degradation of traditional systems may exceed 50% under similar conditions.

[0041] 4. This invention possesses superior intelligent interference defense capabilities: Traditional systems struggle to cope with intelligent interference capable of learning and adapting, while this invention achieves efficient and dynamic defense against intelligent interference through refined multi-dimensional polarization feature recognition, machine learning classification, and an adaptive joint hopping mechanism based on environmental awareness. Experiments show that, facing adaptive tracking interference, the communication interruption rate of this system is (for example) 95% lower than that of traditional frequency hopping / polarization systems, significantly improving mission success rate and survivability.

[0042] 5. This invention optimizes energy and computational efficiency: Although a more complex algorithm is introduced, the heuristic optimization algorithm (such as PTS) of this invention ensures high performance while avoiding brute-force computation through intelligent search. The hierarchical control structure also balances computational requirements. Overall, the system can control energy consumption and computational latency within a reasonable range while significantly improving performance. For example, compared to a scheme that continuously performs global optimization, the average energy consumption is reduced by (e.g.) 30%, and the response latency meets real-time requirements, making it suitable for deployment on low-altitude platforms.

[0043] 6. Enhanced Reliability and Stability of the Invention: The joint channel quality assessment of the present invention includes reliability (PER) and stability (Doppler) metrics, and risk-aware resource allocation further enhances the system's stability under harsh conditions. Even in low signal-to-noise ratio or high dynamic environments, the accuracy of parameter estimation and resource decision-making is guaranteed, and the system's operating threshold is 3-5 dB better than traditional methods (e.g.).

[0044] 7. This invention improves deployment flexibility and scalability: The framework proposed in this invention has modular characteristics, and each core algorithm (evaluation, allocation, control, identification, and transition) can be configured and adjusted according to specific application requirements and platform capabilities. It supports software definition, facilitating integration and upgrades. Attached Figure Description

[0045] Figure 1 This is a diagram showing the overall architecture of the low-altitude communication anti-jamming system combining polarization diversity and frequency diversity of the present invention.

[0046] Figure 2 This is a structural diagram of the multi-polarized antenna array of the present invention;

[0047] Figure 3 This is a structural diagram of the adaptive controller of the present invention;

[0048] Figure 4 This is a flowchart of the channel allocation process of the present invention;

[0049] Figure 5 Example of a polarization-frequency two-dimensional channel quality matrix;

[0050] Figure 6A flowchart for interference defense based on polarization characteristics;

[0051] Figure 7 This is a diagram showing the state transitions of the system's operating modes.

[0052] Figure 8 This is a comparison chart of the bit error rates of various schemes under different interference types in the embodiments of the present invention;

[0053] Figure 9 This is a comparison chart of the throughput of various schemes under different interference intensities in the embodiments of the present invention;

[0054] Figure 10 A comparison chart of link interruption probabilities for various schemes in dynamic scenarios in this embodiment of the invention;

[0055] Figure 11 Comparison chart of relative resource consumption of various schemes in the embodiments of the present invention;

[0056] Figure 12 A simplified collaborative flowchart in this embodiment of the invention. Detailed Implementation

[0057] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0058] Please see Figures 1 to 11 The present invention provides the following technical solutions:

[0059] Example 1: A low-altitude communication anti-jamming system and channel allocation method combining polarization diversity and frequency diversity. The system of this invention consists of the following core components:

[0060] 1.1 Multipolar Antenna Array

[0061] It comprises multiple tunable polarization antenna elements, each of which can independently adjust its polarization (linear, circular, or elliptical). The antenna array employs a compact design, making it suitable for installation on space-constrained low-altitude platforms. Key parameters include:

[0062] Polarization adjustment range: Full polarization coverage;

[0063] Polarization isolation: >25dB

[0064] Polarization switching speed: <10 microseconds

[0065] Antenna gain: 3-6 dBi (depending on polarization)

[0066] 1.2 Broadband RF Front-End

[0067] A multi-band parallel processing RF transceiver system, including low-noise amplifiers, mixers, filters, variable gain amplifiers, and multi-channel ADCs / DACs. Key features include:

[0068] Frequency coverage: 30MHz-3GHz

[0069] Real-time bandwidth: Maximum 100MHz

[0070] Number of channels: 4-8 independent channels

[0071] Dynamic range: >90dB

[0072] Noise figure: <3dB

[0073] 1.3 Digital Signal Processing Unit

[0074] A heterogeneous computing platform based on FPGA and DSP is responsible for signal modulation and demodulation, channel coding, digital filtering, and diversity combining. The main modules include:

[0075] Real-time signal processing engine

[0076] Set merge processor

[0077] Channel estimation and tracking module

[0078] Anti-interference algorithm accelerator

[0079] 1.4 Adaptive Controller

[0080] The core decision-making unit of the system is responsible for formulating resource allocation strategies, optimizing diversity parameters, and managing system configuration. It includes the following functional modules:

[0081] Polarization-Frequency Joint Optimizer

[0082] Variety Parameter Manager

[0083] Adaptive control strategy engine

[0084] System status monitor

[0085] 1.5 Environmental Perception Module

[0086] Responsible for collecting and analyzing electromagnetic environment information to provide a basis for decision-making. This mainly includes:

[0087] Spectrum sensing unit

[0088] Interference Feature Extractor

[0089] Channel State Estimator

[0090] Environmental change predictor.

[0091] Structural Description: The system structure of this invention is as follows: Figure 1 As shown. The system adopts a modular design, with each functional unit connected through a standard interface, facilitating upgrades and expansions. Figure 2 The internal structure of the multi-polarization antenna array is shown, including antenna elements, polarization control circuitry, and antenna feed network. Figure 3 The internal structure of the adaptive controller is shown. This module is the core of the invention and is responsible for coordinating the parameters and resource allocation of each diversity technique.

[0092] This embodiment also provides the following key processes and algorithms:

[0093] 2.1 Polarization-Frequency Joint Channel Allocation Method: The core of this invention is to construct a polarization-frequency two-dimensional channel quality matrix and perform global optimized resource allocation based on this matrix. The channel allocation process is as follows: Figure 4 As shown.

[0094] 2.2 Construction of Two-Dimensional Channel Quality Matrix: The polarization-frequency two-dimensional channel quality matrix constructed by the system is as follows: Figure 5 As shown, the horizontal axis represents different frequency channels, the vertical axis represents different polarization states, and the matrix element values ​​represent the corresponding channel quality indicators.

[0095] 2.3 Anti-Intelligent Interference Mechanism: To address intelligent interference, this invention designs an interference identification and dynamic avoidance method based on polarization characteristics, such as... Figure 6 As shown.

[0096] 2.4 System Operating Modes: Based on environmental complexity and task requirements, the system supports three operating modes, such as... Figure 7 As shown.

[0097] The working principle of the above core algorithm flow is as follows:

[0098] Polarization-Frequency Joint Channel Quality and Diversity Potential Evaluation Algorithm: First, a basic polarization-frequency channel quality index, Q, is defined, which integrates key physical layer parameters:

[0099]

[0100] Among them: SINR p,f The signal-to-interference-plus-noise ratio (PER) at polarization state p and frequency f directly reflects the signal strength and interference noise level; p,f This represents the corresponding packet error rate, reflecting the reliability of actual transmission. (1—PER) is used. p,f The PD item is used as a reliability factor. Ip,fThe Polarization Discrimination Index (PDI) is a polarization fidelity metric that comprehensively reflects the isolation capability of a target polarization p relative to its orthogonal polarizations and the cross-polarization distortion introduced by the channel at a specific frequency f. For example, it can be defined as PDI. p,f =XPD p,f —K·CPR p,f XPD p,f For the cross-polarization discrimination at this frequency, CPR p,f K is the cross-polarization power ratio, and K is the penalty factor. This is finer than using only global XPD. -D f The Doppler spread or channel coherence time correlation index for frequency f reflects the time-varying characteristics of the channel; high time-varying characteristics will reduce the channel quality score. w1, w2, and w3 are non-negative weighting coefficients, configured according to service priorities (such as rate, reliability, and polarization purity), satisfying w1 + w2 + w3 = 1. In typical communication scenarios, to prioritize reliability, w1 = 0.5, w2 = 0.4, and w3 = 0.1 can be set. The sensitivity factor μ is usually taken between [1, 5], and k is usually taken as 2 to achieve a balance between diversity potential and computational stability.

[0101] To quantify and utilize the joint diversity potential of the polarization and frequency domains, a time-domain correlation suppression factor D is introduced. p,f This factor is based on the correlation between the channel quality time series of the polarization-frequency pair (p,f) and the time series of other available pairs (p',f'). The average correlation coefficient magnitude of a specific (p,f) with all other available resources (p',f') ≠ (p,f) is calculated.

[0102]

[0103] Where P((p,f),(p',f')) is the time-varying order of the channel quality (e.g., Q or SINR) of the polarization-frequency pairs (p,f) and (p',f').

[0104] Column X p,f (t) and X p’,f’ Pearson correlation coefficient of (t):

[0105]

[0106] Where Cov represents the covariance and σ represents the standard deviation. Low P -p,f The value indicates that the (p,f) combination is statistically independent of other combinations in terms of channel fluctuations and has higher diversity potential.

[0107] Based on this, a time-domain correlation suppression factor D is defined. p,fThis factor rewards (p,f) combinations that have low correlation with other resources:

[0108]

[0109] Where μ>0 is the sensitivity factor for diversity gain, controlling the magnitude of score improvement brought about by low correlation; k≥1 (e.g., k=2) is an exponential factor used to amplify low correlation (P -p,f The reward effect when ≈0).

[0110] The final polarization-frequency joint utility index Q' p,f It combines basic channel quality with diversity potential:

[0111]

[0112] This Q' p,f The metric not only reflects the current communication quality, but also implicitly assesses the potential contribution of selecting the (p,f) resource to the overall diversity performance of the system, and encourages the selection of channels with statistical characteristics that differ greatly from other resources, thereby maximizing overall robustness during joint selection.

[0113] This embodiment also provides an adaptive resource allocation algorithm based on a trade-off between utility and risk.

[0114] Based on the aforementioned joint utility index Q' p,f This invention proposes a risk-aware multi-objective adaptive resource allocation algorithm. The objective is to maximize the total system utility while considering potential channel fluctuation risks, under certain constraints. The resource allocation problem is defined as follows:

[0115]

[0116]

[0117] Where: w i,j For binary decision variables, 1 indicates choosing the combination of polarization state i and frequency channel j, and 0 indicates not choosing; w = [w i,j [ ] represents the resource allocation matrix; N and M are the number of available polarization states and frequency channels, respectively; K is the maximum number of polarization-frequency resource combinations that can be used simultaneously; c p and c f These are the upper limits of the number of times that can be used for a single polarization state and a single frequency channel, respectively (e.g., if c...). p =1,c f =1 means each polarization and frequency is used at most once; Q' i,j It is the joint utility index defined above; R(w) is a risk measure associated with the selected resource combination w, for example, the selected combination Q' i,jR(w) is the sum of the variances of the values ​​or the predicted probabilities of quality degradation in the near future. i,j =1Var[Q' i,j [(t+Δt)] or similar form; λ R A value of ≥0 represents the risk aversion coefficient, used to balance maximizing total utility and minimizing risk.

[0118] Since this problem is a combinatorial optimization problem, typically NP-hard, this invention employs Predictive Tabu Search (PTS) or a similar heuristic algorithm to solve it. Unlike simple greedy algorithms, PTS has the following characteristics:

[0119] 1. Based on prediction: not just relying on the current Q' i,j Furthermore, it combines a short-term channel prediction model to assess the expected utility and risk over a future period after selecting a certain (i,j).

[0120] 2. Neighborhood search: Systematically search for a better solution within the neighborhood of the current solution.

[0121] 3. Taboo List: Maintain a taboo list to prevent the search process from looping through recently visited solutions or regions and encourage the exploration of new solution spaces.

[0122] 4. Aspiration Criterion: Allows the removal of taboos under specific conditions (such as finding the best solution to date), increasing the flexibility of the algorithm.

[0123] 5. Adaptive Adjustment: Algorithm parameters (such as tabu length and neighborhood definition) can be dynamically adjusted based on the rate of change in the environment or the progress of the solution. This algorithm can find a solution closer to the global optimum than a simple greedy algorithm and exhibits better robustness in dynamically changing environments.

[0124] This embodiment also provides a hierarchical predictive adaptive control mechanism.

[0125] To achieve a balance between real-time response and long-term optimization performance, this invention designs a hierarchical adaptive control architecture with predictive capabilities, comprising a fast response layer and a deep planning layer.

[0126] FastResponse Layer: Responsible for handling sudden, severe channel degradation or strong interference events. This layer makes near-instantaneous decisions based on a predefined event-action rule base and real-time monitoring data.

[0127] A fast =RuleSelect(Quantize(s) current ,I detected ,Δs / Δt),R)

[0128] Among them, s current The current system state vector (including Q') p,f (Matrix, resource consumption, etc.), I detected For real-time detected interference features, Δs / Δt is the rate of state change, R is the rule base (e.g., if the SINR of a specific (P,f) drops sharply by >10dB and the PER suddenly increases, then immediately switch to the backup (P',f') in the contingency plan base), Quantize discretizes the input, and RuleSelect selects the fast response action A according to the matching rules. fast (such as switching, power adjustment, etc.)

[0129] The Deep Planning Layer operates on a longer timescale and is responsible for more refined and forward-looking resource optimization. This layer invokes the aforementioned risk-aware multi-objective adaptive resource allocation algorithm.

[0130]

[0131] SolvePTS represents the execution of a predictive tabu search algorithm, with input including the future system state s obtained using a short-term prediction model. predicted and interference model I modeled, and objective function P obj (maximize u(w)) and constraint set C onstraints The output is the optimized resource allocation action A. deep .

[0132] Decision Fusion: The outputs of the two layers are integrated through an intelligent fusion logic to generate the final control command A. final The fusion mechanism is no longer a simple linear weighting, but rather based on confidence level and timeliness assessment:

[0133] A final =Fuse(A fast A deep ,Confidence(A deep Urgency(s) current ,I detected ))

[0134] The Fuse function prioritizes actions with high confidence and those that meet urgent requirements. For example, when the environment changes drastically (high urgency), A... fast The priority is increased; when the environment is stable and the solution given by the deep programming layer has high confidence, A deep The suggestion was adopted. Confidence can be assessed based on factors such as the accuracy of the prediction model and the convergence of the optimization algorithm.

[0135] This embodiment also provides refined interference identification and classification based on multidimensional polarization features and machine learning.

[0136] This invention proposes to utilize the extended polarization domain eigenvector P extI By combining machine learning, more refined interference identification and classification can be achieved. This feature vector not only contains static intensity information, but also incorporates dynamic and phase information.

[0137]

[0138] Among them: I x R is the interference power measured under each fundamental polarization state X; ab It is the interference work ratio between polarization states a and b; R + / - R is the interference power ratio between +45° and -45° linear polarization. RL It is the interference work ratio between right-handed and left-handed circular polarization; I x This corresponds to the rate of change (dynamic characteristic) of the polarization interference power; Δφ VH It is the mean or variance (phase characteristic) of the phase difference between the vertically polarized and horizontally polarized interference components; It is a feature extracted from the power spectral density (PSD) of polarization state X (such as the number of spectral peaks, spectral entropy, and frequency domain features).

[0139] Based on this enhanced feature vector, an ensemble learning method (such as Gradient Boosting Decision Tree (GBDT) or Random Forest) is used for interference classification, instead of a single SVM. Ensemble learning typically provides higher classification accuracy and robustness. The classifier model is represented as:

[0140] y = EnsembleClassify(P extI ;Θ)

[0141] Where y is the predicted interference category (e.g., narrowband, broadband noise, specific signal interference, tracking interference, etc.), and Θ is the parameter of the ensemble model. This model is obtained through offline training.

[0142] Based on the refined interference classification result y, the system can execute a more targeted adaptive defense strategy:

[0143] Polarization-selective interference: accurately identify the affected polarization dimensions and prioritize orthogonal or least affected polarization states.

[0144] Frequency / polarization composite interference: Based on the classification results, understand the distribution characteristics of interference in the joint domain and select resource combinations that can simultaneously avoid interference in the frequency domain and polarization domain.

[0145] Dynamic tracking interference: Activate the following adaptive joint hopping mechanism and adjust the hopping parameters according to the predicted interference tracking mode.

[0146] Unknown / novel interference: trigger anomaly detection, record feature data for subsequent analysis, and adopt general robustness enhancement strategies (such as maximizing diversity).

[0147] This embodiment also provides an environment-aware adaptive joint polarization-frequency hopping mechanism.

[0148] To address intelligent interference capable of learning and predicting communication switching patterns, this invention designs an adaptive chaotic joint switching mechanism based on real-time environmental feedback. The switching sequence {(p t ,f t )} Tt The generation of 1 depends not only on the key but also incorporates awareness of the current electromagnetic environment and potential threats. The underlying transition logic can still use chaotic systems (such as improved Hénon mappings or other high-dimensional chaotic systems) to generate pseudo-random sequences, which are then mapped to polarization states and frequency channels.

[0149] z t+1 =F(z) t ,c t )

[0150] z t+1 =F(z) t ,c t )

[0151]

[0152] in:

[0153] z t It is the state vector of a chaotic system.

[0154] F is a chaotic mapping function.

[0155] c t It is the control parameter vector of the chaotic system, and this parameter is the key to adaptive adjustment.

[0156] N and M represent the number of polarization states and frequency channels, respectively.

[0157] g p ,g f These are functions that map chaotic states to polarization and frequency indices. They can non-uniformly map the state space to available resources and are influenced by real-time environmental information (EnvInfo, such as the type of interference detected, the interference frequency band, and channel quality feedback). For example, if strong interference is detected in a frequency band F_sub, the mapping function g... f This will reduce the probability of selecting a frequency within that band.

[0158] The key lies in the control parameter c t and mapping function g p ,g f Adaptability:

[0159] c t =AdaptParams(Key,InterferenceClass t ,Q' matrix (t-1),HoppingHistory)

[0160] g p / f (·)=UpdateMappingFunctions(EnvInfo t )

[0161] The parameter adjustment function `AdaptParams` uses the key as a basis and dynamically adjusts the chaotic behavior (e.g., changing the shape of the attractor or the traversal region) based on the currently identified interference category, the channel quality matrix of the previous time step, and even inferences about the jammer's behavior (learned from the jump history). The update of the mapping function `gp / f` ensures that the jumps tend to select high-quality, low-interference resources in the current environment. This dual adaptation (chaotic dynamics parameter adaptation + output mapping adaptation) greatly improves the unpredictability and environmental adaptability of the jump sequence, effectively combating intelligent tracking interference.

[0162] The workflow of the above system is summarized as follows:

[0163] 1. The environmental sensing module continuously monitors the electromagnetic environment, extracts interference information using multi-dimensional polarization features, and estimates Q' using a joint evaluation algorithm. p,f matrix.

[0164] 2. The hierarchical adaptive controller integrates the results of fast response and deep planning.

[0165] 3. The controller outputs commands to configure the multi-polarized antenna and RF front-end, select the optimal (p,f) combination, or activate / adjust the adaptive switching mechanism.

[0166] 4. The interference identification module continuously classifies interference, providing input to the controller and switching mechanism.

[0167] 5. The digital signal processing unit performs the corresponding operation.

[0168] 6. System performance is continuously monitored and feedback is given to the evaluation, control, and transition modules, forming a closed-loop adaptive system.

[0169] The system strategy is adjusted according to different interference scenarios:

[0170] No / weak interference: Optimization objectives can be biased towards energy efficiency or throughput, λ R Lower it.

[0171] Strong noise / known fixed interference: Deep planning layer dominates, utilizing Q' p,f The PTS algorithm is used to find the optimal fixed resource allocation.

[0172] from Figure 8 As can be seen, the system of this invention exhibits the lowest bit error rate under all tested interference types, and its performance is significantly better than other schemes. Especially in the simulated tracking-type intelligent interference scenario, the bit error rate of scheme D is as low as 0.0009, representing a 95.5% performance improvement compared to the simple superposition scheme C (0.0200), and a much greater improvement compared to single diversity schemes A and B. This fully demonstrates the superior ability of the joint optimization and adaptive mechanism of this invention to combat complex and intelligent interference. Scheme C (simple superposition diversity) refers to using independent frequency diversity and polarization diversity modules, selecting resources through simple polling or a single-threshold-based switching strategy, lacking a joint optimization process.

[0173] This embodiment also provides a comparison of system throughput under different interference intensities.

[0174] This experiment tested the effective throughput performance of four schemes under different interference intensities (the ratio of interference signal power to signal power, JSR).

[0175] Figure 9 It is clearly shown that the throughput of all schemes decreases as the interference strength (JSR) increases. However, the system of the present invention (Scheme D), with its joint utility assessment, risk-aware allocation, and adaptive mechanism, exhibits extremely strong robustness, with the most gradual decrease in throughput. Especially in high-interference regions (JSR ≥ 10 dB), the performance advantage of the scheme is particularly significant. For example, at JSR = 20 dB, Scheme D can still maintain a throughput of approximately 6.5 Mbps, an improvement of approximately 242% compared to Scheme C, and far exceeding the single diversity schemes A and B, demonstrating the effectiveness of the present invention in ensuring communication bandwidth under high interference.

[0176] This embodiment also provides a comparison of link stability in dynamic scenarios.

[0177] This experiment simulates the link stability of low-altitude platforms (such as drones) in dynamic scenarios involving speed changes and rapid turns, using the average link interruption probability over a period of time as the evaluation metric.

[0178] Figure 10The probability of link interruption for each scheme under different platform speed and turning rate combinations is illustrated using contour plots. The cooler the color (or the green / blue in the map settings), the lower the interruption probability and the more stable the link. Observation shows that the interruption probability increases for all schemes in the high-dynamic region (high speed and high turning rate, marked with a red box in the plot). However, the interruption probability of the system of this invention in this region is significantly lower than other schemes, and its contour line extension in the high-dynamic region is the most limited. Calculation of the average interruption probability within the red box area shows that scheme D is approximately 0.018, far lower than scheme D (an improvement of approximately 73.0%) and schemes A and B. This strongly demonstrates the superior ability of the hierarchical predictive adaptive control and robust resource allocation of this invention to maintain link stability in high-dynamic environments.

[0179] This embodiment also provides a comparison of system resource utilization efficiency.

[0180] This experiment compared the relative resource consumption of each scheme, mainly considering algorithm computational complexity, memory usage, average energy consumption (related to computational intensity and transmission strategy), and response latency (the time required for the system to adjust from environmental changes). Scheme C (simple superposition) was used as the baseline.

[0181] Figure 11 A radar chart was used to visually compare the relative resource consumption of the four schemes in terms of computational complexity, memory usage, average energy consumption, and response latency (with scheme C as the benchmark; the smaller the value, the less consumption / higher efficiency). It can be seen that although the system of this invention (scheme D) is slightly higher than the benchmark scheme C in terms of computational complexity and memory usage (1.3 times and 1.2 times respectively), reflecting its use of more advanced optimization and control algorithms, it shows a significant advantage in average energy consumption (0.7 times) and response latency (only 0.4 times). This is due to the intelligent algorithm design (such as heuristic search to avoid brute-force computation) and the hierarchical control architecture (a fast response layer reduces average latency). Overall, the calculated geometric mean resource consumption index shows that scheme D (0.75) is superior to scheme C (1.00), with an overall resource efficiency improvement of approximately 19.5%. This indicates that the present invention achieves a good balance between significantly improving communication performance and resource efficiency, and its low energy consumption and fast response characteristics are particularly suitable for low-altitude platform applications that are sensitive to energy consumption and real-time performance.

[0182] Based on the above, we can conclude that:

[0183] This invention successfully proposes and verifies a low-altitude communication anti-interference system and method based on deep polarization-frequency fusion and intelligent adaptive control. Addressing the severe challenges of the complex electromagnetic environment at low altitudes, this invention goes beyond the application of single diversity techniques or simple superposition. Instead, it quantifies the comprehensive value of two-dimensional resources through an innovative polarization-frequency joint utility and diversity potential evaluation model. Based on this, the designed risk-aware multi-objective adaptive resource allocation algorithm (emphasizing predictive heuristic search) can balance expected returns and potential risks while satisfying constraints, achieving robust and efficient resource allocation.

[0184] The core hierarchical predictive adaptive control mechanism integrates the agility of rapid response with the optimization of deep planning, achieving seamless collaboration between the two through intelligent decision-making logic. Combined with refined interference identification based on multi-dimensional polarization features and ensemble learning, the system can accurately perceive and understand the interference environment. Crucially, the adaptive joint polarization-frequency hopping mechanism for environment perception, by adjusting chaotic dynamic parameters and output mapping in real time, demonstrates unprecedented resilience against intelligent tracking interference.

[0185] Extensive simulation experiments clearly demonstrate that the system of this invention outperforms existing technologies in a wide range of complex scenarios, including strong interference, dynamic changes, and intelligent countermeasures. These performances are significantly superior in terms of anti-interference performance (reduced bit error rate), communication capacity (increased throughput), link stability (reduced interruption rate), and resource utilization efficiency (energy consumption and latency optimization).

[0186] The system and method proposed in this invention not only provide reliable and efficient communication support for low-altitude unmanned platforms (UAVs, eVTOL, etc.), but its core technical concepts and algorithm modules can also be widely applied to other wireless communication fields facing challenges in complex electromagnetic environments, such as tactical ad hoc networks, satellite communication anti-jamming, and vehicle network security. It has important theoretical value and broad application prospects, and is a key step in promoting the development of next-generation intelligent anti-jamming communication technology.

[0187] Example 2: Adaptive Resource Allocation Scheme Based on Deep Reinforcement Learning

[0188] This alternative uses deep reinforcement learning (DRL) to replace the proposed heuristic optimization algorithm for resource allocation decisions. The system models the polarization-frequency resource allocation problem as a Markov decision process (MDP) and uses methods such as deep Q-networks (DQN) and actor-critic (e.g., A3C, DDPG) to learn the optimal policy.

[0189] The following section uses Deep Q-Network (DQN) as an example to supplement its pseudocode and key implementation details:

[0190] Algorithm 1: Resource Allocation Decision Based on DQN

[0191] 1. Initialization: Initialize the experience replay pool D with a capacity of N. D Initialize the weights θ (in the main network) of the action-value function Q and copy them to the target network. The weight is θ - =θ.

[0192] 2. Loop (for episode = 1 to M):

[0193] 3. Obtain the initial environment state s1 (including Q') p,f Matrix, P extI wait).

[0194] 4. Loop (for t = 1 to T):

[0195] 5. Randomly select an action a with probability ∈. t (explore).

[0196] 6. Otherwise, choose a. t =argmax a Q(s t ,a;θ)(utilization).

[0197] 7. Perform an action a in the environment t (e.g., select a (p,f) combination).

[0198] 8. Observed reward r t and new state s t+1 >

[0199] 9. Transfer (s) t ,a t ,r t ,s t+1 Store it in the experience replay pool D.

[0200] 10. Randomly sample a mini-batch of transitions from D. j ,a j ,r j ,s j+1 ).

[0201] 11. Calculate the target Q value:

[0202]

[0203] 12. Perform one step of gradient descent and update the main network weights θ:

[0204]

[0205] 13. Every C steps, update the target network: θ - ←θ.

[0206] 14.s t ←s t+1 .

[0207] 15. End the inner loop.

[0208] 16. End the outer loop.

[0209] Training data generation: Training data can be generated using a high-fidelity channel simulator. This simulator simulates different low-altitude channel conditions (multipath, Doppler shift) and various types of interference (such as narrowband, wideband, tracking interference, etc.), providing a rich interactive environment for the DRL agent to learn.

[0210] Key hyperparameter: learning rate α (e.g., 10) -4 ), discount factor γ (e.g., 0.99), decay rate of the ∈-greedy strategy, and empirical replay pool capacity N. D (e.g. 10) 5 The target network update frequency C (e.g., 1000 steps).

[0211] The system state is defined as: - the current and historical polarization-frequency joint utility matrix Q' p,f -Enhanced interference feature vector P extI The action space may include: - selecting or releasing a specific (p, f) resource combination - adjusting the transmit power of a specific combination - selecting a hopping mode or adjusting hopping parameters. The reward function is designed as follows:

[0212] R t =f(Throughput) t ,PER t Latency t Energy t RiskPenalty t )

[0213] The reward function needs to be carefully designed to balance multiple objectives, such as maximizing throughput, minimizing error rate and latency, while taking into account energy consumption and risk aversion (e.g., penalizing the selection of resources that will degrade prediction quality).

[0214] Advantages: Its potential lies in its ability to handle highly complex dynamic environments and unknown disturbance patterns, and it can learn to adapt without explicit models.

[0215] Disadvantages: It requires a large amount of training data (simulation or actual interaction), the training process may be unstable, convergence is difficult to guarantee, the decision lacks interpretability, and the computational complexity of real-time inference may be high.

[0216] Example 3

[0217] The core idea of ​​the invention is extended from a single node to a network composed of multiple nodes equipped with the capabilities of this invention. Through cooperation between nodes, stronger situational awareness and interference suppression capabilities are achieved.

[0218] Below is a simplified collaborative flowchart (e.g.) Figure 12 (As shown) and an example of cooperative interference cancellation to illustrate:

[0219] Example of cooperative interference cancellation:

[0220] 1. Information exchange: Nodes A, B, and C each calculate their local channel quality matrix Q using the methods described in steps 1-7. (A) Q (B ),Q (C) and interference feature vectors They broadcast this information to neighboring nodes through a secure control channel.

[0221] 2. Cooperative localization: A node (or a specified cluster head node) merges the received P values ​​from all nodes. extI Information. For example, by using the time difference of arrival (TDOA) or angle of arrival (AOA) of interference signals from the same source received by different nodes, combined with polarization characteristics, the location of the interference source can be estimated more accurately.

[0222] 3. Cooperative Beamforming: Once the location of the interference source is determined, multiple nodes can perform cooperative beamforming. Assume node A is the primary communication node, and nodes B and C are cooperative nodes. They can adjust the phase and amplitude of their respective antenna arrays' transmitted / received signals, ensuring that signals are in-phase superimposed (constructive interference) in the target communication direction and out-of-phase superimposed (destructive interference) in the direction of the located interference source. This creates a "null" pointing towards the interference source, achieving spatial filtering and cancellation of the interference signal.

[0223] Key technical elements:

[0224] 1. Distributed Joint Sensing and Assessment: Each node independently performs assessment and shares the Q'p,f matrix and interference features P through a secure channel. extI Information such as these are combined to generate a global or regional joint utility view.

[0225] 2. Cooperative interference direction finding and localization: Utilize interference signal information (including polarization features) received by multiple nodes for cooperative processing (e.g., based on TDOA / FDOA / AOA, combined with polarization information) to achieve more accurate localization of interference sources.

[0226] 3. Distributed Cooperative Resource Allocation: This extends the resource allocation problem into an optimization problem for multi-agent systems (MAS). Distributed optimization algorithms (such as ADMM), game theory-based methods (finding Nash equilibrium or designing cooperative game strategies), or federated learning frameworks can be used for decision-making to maximize the overall utility of the network or satisfy some fairness criterion.

[0227] 4. Cooperative interference suppression / elimination: In some scenarios, if the interference source has been located, multiple nodes can perform cooperative beamforming to form nulls in the direction of interference, or perform cooperative interference signal elimination.

[0228] 5. Dynamic network topology and role allocation: Based on the overall situation and task requirements, dynamically adjust the connection relationships and cluster structure between nodes, and even dynamically allocate master nodes for centralized coordination or adopt fully distributed control.

[0229] Advantages: Significantly enhances the ability to counter wide-area, mobile, or multiple interference sources; improves overall performance and robustness through spatial diversity and cooperative processing. Disadvantages: System complexity increases dramatically; high communication overhead between nodes; high synchronization requirements; challenges in the design and convergence of cooperative algorithms; susceptible to network connectivity issues.

[0230] The foregoing has shown and described the basic principles and main features of the present invention and its advantages. It will be apparent to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the spirit or basic features of the present invention.

[0231] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A low-altitude communication anti-jamming system combining polarization diversity and frequency diversity, characterized in that: The system comprises a multi-polarized antenna array, a broadband RF front-end, a digital signal processing unit, an adaptive controller, and an environmental sensing module. The multi-polarized antenna array includes multiple tunable polarization antenna elements, each capable of independently adjusting its polarization state. The broadband RF front-end supports a multi-band parallel processing RF transceiver system, including a low-noise amplifier, mixer, filter, variable gain amplifier, and multi-channel ADC / DAC. The digital signal processing unit, based on a heterogeneous computing platform of FPGA and DSP, is used for signal modulation and demodulation, channel coding, digital filtering, and diversity combining. The adaptive controller is used for resource allocation strategy formulation, diversity parameter optimization, and system configuration management. The environmental sensing module collects and analyzes electromagnetic environment information to provide a basis for decision-making. The channel allocation method based on the above anti-interference system includes the following steps: After channel state acquisition, a polarization-frequency channel quality matrix is ​​constructed, followed by channel quality assessment and determination of interference. If interference exists, interference characteristics are analyzed, the interference model is updated, and an interference avoidance strategy is generated. Then, optimization objectives and resource allocation are determined, diversity parameters are calculated, and the configuration is updated. Performance is monitored to determine if the performance meets requirements. If the requirements are met, the current configuration is maintained. If no interference exists, the process of determining optimization objectives proceeds directly. If the performance monitoring results do not meet requirements, channel state acquisition is repeated. The process also includes a polarization-frequency joint channel quality and diversity potential assessment algorithm: first, a basic polarization-frequency channel quality index is defined. This metric incorporates key physical layer parameters: ； in: The signal-to-interference-plus-noise ratio (SIR) is given by the polarization state p and frequency f. For the corresponding packet error rate, As a polarization fidelity index, The channel time-varying index for frequency f. These are non-negative weighting coefficients; further evaluation of diversity potential includes the following steps: a) Calculate specific polarization-frequency pairs With all other available resources Average correlation coefficient magnitude : ； in, and These represent the number of available polarization states and frequency channels, respectively. The Pearson correlation coefficient between the two resource pairs and the channel quality time series. b) Define a time-domain correlation suppression factor based on the average correlation coefficient modulus. This factor rewards combinations that have low correlation with other resources: ； in, The sensitivity factor for diversity gain. As an exponential factor, the final polarization-frequency joint utility index It combines basic channel quality with diversity potential: 。 2. The low-altitude communication anti-jamming system combining polarization diversity and frequency diversity according to claim 1, characterized in that, The parameters of the multi-polarized antenna array include: Polarization adjustment range: Full polarization coverage; Polarization isolation: >25dB; Polarization switching speed: <10 microseconds; Antenna gain: 3-6 dBi; The broadband radio frequency front end is characterized by: Frequency coverage: 30MHz-3GHz; Real-time bandwidth: Maximum 100MHz; Number of channels: 4-8 independent channels; Dynamic range: >90dB; Noise figure: <3dB.

3. The low-altitude communication anti-jamming system combining polarization diversity and frequency diversity according to claim 1, characterized in that, The digital signal processing unit includes the following modules: a real-time signal processing engine, a diversity merge processor, a channel estimation and tracking module, and an anti-interference algorithm accelerator.

4. The low-altitude communication anti-jamming system combining polarization diversity and frequency diversity according to claim 1, characterized in that, The adaptive controller includes modules such as a polarization-frequency joint optimizer, a diversity parameter manager, an adaptive control strategy engine, and a system state monitor; the environment perception module includes modules such as a spectrum sensing unit, an interference feature extractor, a channel state estimator, and an environment change predictor.

Images