An adaptive anti-interference communication method and system based on Star Flash-Bluetooth dual mode

Abstract

The present invention relates to the field of wireless communication technology, and specifically discloses an adaptive anti-interference communication method and system based on Star Flash-Bluetooth dual-mode. The method monitors channel interference parameters in real time through the dual-mode radio frequency module on the communication terminal, including the out-of-band radiation power and pulse interference duty cycle at the switching moment. The instantaneous interference intensity characteristics are extracted using synchronous compression wavelet transform and sparse time-frequency analysis, and the periodic pulse interference characteristics are identified by combining Hawkes process modeling, and a comprehensive anti-interference feature vector is constructed. The feature vector is input into a pre-trained gradient boosting tree model for communication quality assessment, and the optimal communication mode is dynamically selected based on the scoring results to achieve intelligent response to complex interference environments. The present invention improves the stability and reliability of the communication system in high-interference scenarios, takes into account energy efficiency optimization and resource scheduling, and is suitable for application scenarios with strict requirements on communication quality, such as industry, medical care, and the Internet of Things.

Description

Technical Field

[0001] The present invention relates to the field of wireless communication technology, and in particular to a StarFlash-Bluetooth dual-mode adaptive anti-interference communication method and system. Background Art

[0002] With the rapid development of wireless communication technology, dual-mode Starflash and Bluetooth devices have been widely used in smart homes, industrial IoT, and other fields. However, problems such as co-channel interference and pulse noise in complex electromagnetic environments have seriously affected communication reliability. Traditional anti-interference methods mainly use fixed-threshold RSSI detection or static mode switching strategies, which are difficult to deal with the combined effects of transient spectrum leakage and periodic pulse interference. Existing research has mostly focused on optimizing anti-interference for a single communication mode and lacks a joint analysis mechanism for spectrum pollution and pulse interference in the dual-mode Starflash-Bluetooth system. This leads to frequent problems such as false switching and communication interruptions in high-interference scenarios such as industrial sites.

[0003] The existing technology has the following deficiencies:

[0004] The existing technology lacks the transient spectrum regeneration phenomenon caused by dual-mode switching (i.e., out-of-band radiation generated by the transient response of the RF front-end during mode switching). The present invention accurately quantifies the degree of contamination of adjacent frequency channels by this phenomenon through synchronous compression wavelet transform, and combines it with pulse interference analysis based on Hawkes process modeling to solve the misjudgment problem caused by traditional methods that ignore the coupling effect of transient interference and steady-state interference. Summary of the Invention

[0005] The purpose of the present invention is to provide an adaptive anti-interference communication method and system based on StarFlash-Bluetooth dual mode to solve the problems in the above background.

[0006] The purpose of the present invention can be achieved through the following technical solutions:

[0007] An adaptive anti-interference communication method based on Starflash-Bluetooth dual mode, comprising the following steps:

[0008] S1: The dual-mode RF module deployed on the communication terminal monitors the interference parameters of the current channel in real time.

[0009] The interference parameters include: out-of-band radiation power and pulse interference duty cycle at the switching moment;

[0010] S2: Perform time-frequency domain analysis on the out-of-band radiation power, construct a spectrum leakage feature sequence, and calculate the instantaneous interference intensity feature value to assess the degree of spectrum pollution during dual-mode switching.

[0011] S3: Statistically model the pulse interference duty cycle, construct the pulse interference distribution matrix, and calculate the burst noise tolerance eigenvalue to identify whether the channel is subject to periodic strong interference.

[0012] S4: The instantaneous interference intensity characteristic value and the burst noise tolerance characteristic value are combined into a comprehensive anti-interference feature vector, which is input into the pre-trained anti-interference decision model for analysis. The communication mode is then optimized based on the analysis results.

[0013] S5: Dynamically switch between StarFlash and Bluetooth communication modes based on the evaluation results, or enable dual-mode parallel transmission to achieve adaptive anti-interference communication in complex interference environments.

[0014] As a further solution of the present invention, the evaluation of the spectrum pollution degree during dual-mode switching specifically includes:

[0015] The out-of-band radiation power at the switching moment of the current channel is obtained in real time, the out-of-band radiation power is analyzed in the time-frequency domain, a spectrum leakage feature sequence is constructed, the instantaneous interference intensity feature value is calculated, and it is determined whether the instantaneous interference intensity feature value is greater than or equal to the preset threshold. If so, the spectrum pollution during dual-mode switching is abnormal; otherwise, the spectrum pollution during dual-mode switching is normal.

[0016] As a further solution of the present invention: the process of obtaining the instantaneous interference intensity characteristic value is:

[0017] The out-of-band radiation power of the current channel at the switching moment is obtained in real time. The out-of-band radiation power signal at the switching moment is subjected to synchronous compression wavelet transform. The time-frequency energy is decomposed using the Morlet wavelet basis function. The ratio of the maximum out-of-band energy to the average in-band energy is calculated to obtain the maximum energy leakage ratio.

[0018] The out-of-band radiation power signal is subjected to sparse time-frequency distribution analysis. The sparse representation of the time-frequency matrix is ​​solved by L1 regularization. The proportion of non-zero elements in the out-of-band region is counted to obtain the sparsity index.

[0019] The maximum energy leakage ratio and the sparsity index are normalized and calculated to obtain the instantaneous interference intensity characteristic value.

[0020] As a further solution of the present invention, the step of identifying whether a channel is subject to periodic strong interference specifically includes:

[0021] The pulse interference duty cycle of the current channel is obtained in real time, the pulse interference duty cycle is statistically modeled, the pulse interference distribution matrix is ​​constructed, the burst noise tolerance characteristic value is calculated, and it is determined whether the burst noise tolerance characteristic value is greater than or equal to the preset threshold. If so, the channel is subject to periodic strong interference; if not, the channel is not subject to periodic strong interference.

[0022] As a further solution of the present invention: the process of obtaining the burst noise tolerance characteristic value is:

[0023] Real-time acquisition of the time series of pulse interference events in the channel;

[0024] Modeling of pulse interference based on Hawkes process;

[0025] Solve for parameters using maximum likelihood estimation 、 and , specifically achieved by optimizing the log-likelihood function;

[0026] The burst noise tolerance characteristic value is calculated based on the average pulse interval and pulse duration.

[0027] As a further solution of the present invention, the instantaneous interference intensity characteristic value and the burst noise tolerance characteristic value are fused into a comprehensive anti-interference characteristic vector, which is input into a pre-trained anti-interference decision model for analysis, specifically including:

[0028] The instantaneous interference intensity characteristic value and the burst noise tolerance characteristic value are obtained, and the instantaneous interference intensity characteristic value and the burst noise tolerance characteristic value are constructed into a comprehensive anti-interference feature vector as the input of the anti-interference decision model. Minimizing the error between the predicted communication quality score and the actual communication quality score is used as the training goal of the anti-interference decision model. The anti-interference decision model is trained, and the communication quality score is output according to the trained anti-interference decision model. The anti-interference decision model is a gradient boosting tree model.

[0029] As a further solution of the present invention: the training process of the anti-interference decision model is:

[0030] Construct a training data set, the training data set including: historically collected instantaneous interference intensity characteristic value samples; historically collected burst noise tolerance characteristic value samples; corresponding actual communication quality score labels; standardize the characteristic values ​​so that they obey a standard normal distribution with a mean of 0 and a variance of 1; use the SMOTE algorithm to oversample the sample imbalance data; construct a gradient boosting tree model, the model parameters including: the number of basic learners is set to 100 to 200; the learning rate is set to 0.05-0.2; the maximum depth of the tree is set to 3-5 layers; the minimum number of leaf node samples is set to 5-10; use the k-fold cross-validation method to train the model, specifically including: randomly dividing the data set into k mutually exclusive subsets; selecting k-1 subsets as training sets each time, and the remaining 1 subset as a validation set; repeating the training process k times, and taking the average performance as the model evaluation indicator;

[0031] During the model optimization phase, Bayesian optimization is used to tune hyperparameters, with the goal of minimizing the mean absolute error on the validation set.

[0032] The optimal model parameters after training are solidified and deployed to the embedded system of the communication terminal.

[0033] As a further solution of the present invention, the communication mode optimization evaluation based on the analysis results specifically includes:

[0034] Based on the communication quality score output by the anti-interference decision model, a three-level optimization strategy including star flash mode, Bluetooth mode and dual-mode parallel transmission is established: if the score is lower than the first threshold, it is preferentially switched to the high-interference anti-interference star flash mode; if the score is between the first and second thresholds, dual-mode parallel transmission is enabled to take into account both reliability and compatibility; if the score is higher than the second threshold, the low-power Bluetooth mode is adopted; at the same time, the channel state changes are monitored in real time. When it is detected that the score crosses the threshold boundary N times in a row, the mode switching is triggered, and the communication quality is further optimized by dynamically adjusting the transmission power and modulation method, where N is the preset hysteresis count parameter.

[0035] As a further solution of the present invention: dynamically switching between the StarFlash and Bluetooth communication modes according to the evaluation results, or enabling dual-mode parallel transmission, to achieve adaptive anti-interference communication in complex interference environments, specifically includes:

[0036] A three-level optimization strategy is implemented based on the scoring results: when the score is below the first threshold, the star flash mode is prioritized to ensure communication reliability; when the score is between the first and second thresholds, dual-mode parallel transmission is enabled to balance performance and compatibility; when the score is above the second threshold, the Bluetooth mode is switched to reduce power consumption, while a hysteresis counting mechanism is used to avoid frequent mode switching;

[0037] During the mode switching process, the transmission power and modulation method are dynamically adjusted to further optimize the communication quality. For the Star Flash mode, high-order modulation and maximum transmission power are prioritized, while the Bluetooth mode adaptively selects GFSK or LE Coded PHY according to the degree of interference. During dual-mode parallel transmission, time division multiplexing is used to coordinate the resource allocation of the two protocols, and the changes in characteristic values ​​are monitored in real time. When the interference characteristics are detected to be persistently abnormal, the mode is re-evaluated to form a closed-loop anti-interference control mechanism.

[0038] An adaptive anti-interference communication system based on Starflash-Bluetooth dual mode, comprising:

[0039] A real-time interference monitoring module, which monitors the interference parameters of the current channel in real time through a dual-mode radio frequency module deployed on the communication terminal;

[0040] The interference parameters include: out-of-band radiation power and pulse interference duty cycle at the switching moment;

[0041] A spectrum pollution analysis module, which is used to perform time-frequency domain analysis on out-of-band radiation power, construct a spectrum leakage feature sequence, and calculate the instantaneous interference intensity feature value to evaluate the degree of spectrum pollution during dual-mode switching;

[0042] A pulse interference modeling module, which performs statistical modeling on the pulse interference duty cycle, constructs a pulse interference distribution matrix, and calculates a burst noise tolerance eigenvalue to identify whether the channel is subject to periodic strong interference;

[0043] An intelligent decision-making module, which combines the instantaneous interference intensity characteristic value and the burst noise tolerance characteristic value into a comprehensive anti-interference feature vector, inputs it into a pre-trained anti-interference decision model for analysis, and performs communication mode optimization evaluation based on the analysis results;

[0044] A dynamic switching execution module dynamically switches between StarFlash and Bluetooth communication modes according to the evaluation results, or enables dual-mode parallel transmission to achieve adaptive anti-interference communication in complex interference environments.

[0045] Beneficial effects of the present invention:

[0046] (1) The present invention constructs a set of refined interference perception mechanisms for complex electromagnetic environments by deeply exploring and introducing interference parameters such as out-of-band radiation power and pulse interference duty cycle, which are easily overlooked but highly indicative in traditional communication systems. This mechanism not only realizes all-round and multi-dimensional perception of channel status, but also significantly improves the system's detection sensitivity and response speed to transient interference and periodic noise by integrating advanced signal processing technologies, such as synchronous compression wavelet transform for high-precision time-frequency domain analysis, and statistical modeling methods based on Hawkes process for dynamic identification of periodic pulse interference. The system normalizes and fuses the extracted instantaneous interference intensity eigenvalues ​​and burst noise tolerance eigenvalues ​​to construct a comprehensive anti-interference feature vector with clear physical meaning and strong interpretability, which is fed into a fully trained gradient boosting tree model as input to achieve intelligent evaluation and prediction of current communication quality. The decision model is trained with the goal of minimizing the communication quality score error. It has good generalization ability and robustness, and can complete the entire process from interference perception to mode selection within milliseconds, ensuring that the system can quickly adapt to the changing wireless environment. Finally, based on the communication quality score output by the model, the system implements a three-level communication mode optimization strategy: in strong interference scenarios, it prioritizes highly robust Star Flash communication, switches to low-power Bluetooth mode in weak interference environments, and adopts dual-mode parallel transmission under moderate interference conditions to balance stability and compatibility. At the same time, it combines a hysteresis counting mechanism to avoid unnecessary frequent switching, forming a closed-loop adaptive anti-interference control system. This communication architecture, which deeply integrates the "perception-analysis-decision-execution" link, effectively breaks through the performance bottleneck of traditional single communication modes in complex interference environments, significantly improving the stability of the communication link, the reliability of data transmission, and the overall energy efficiency of the system. It is particularly suitable for key application scenarios such as industrial automation, medical monitoring, and smart wearables that have strict requirements on communication quality and energy consumption.

[0047] (2) The present invention not only focuses on improving the communication quality, but also takes into account the energy consumption and resource utilization efficiency of the system. By designing a three-level communication mode optimization strategy, it can flexibly switch between Star Flash, Bluetooth or dual-mode parallel mode under different interference intensities, and at the same time combine the dynamic adjustment of transmission power and modulation mode mechanism to enable the system to reduce power consumption as much as possible under the premise of ensuring communication quality. For example, in a low-interference environment, Bluetooth mode is used first to save energy, and in a high-interference scenario, Star Flash enhanced mode is enabled to ensure performance, while under medium interference conditions, dual-mode parallel is used to achieve a balance between performance and compatibility. In addition, the hysteresis counting mechanism effectively prevents the additional overhead caused by frequent mode switching. This refined resource scheduling mechanism enables the communication system to have stronger environmental adaptability and is suitable for a variety of power-sensitive application scenarios such as the Internet of Things and wearable devices. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] The present invention will be further described below with reference to the accompanying drawings.

[0049] Figure 1 This is a flowchart of an adaptive anti-interference communication method based on StarFlash-Bluetooth dual mode of the present invention;

[0050] Figure 2 This is a flow chart of an adaptive anti-interference communication system based on Starflash-Bluetooth dual mode in the present invention. DETAILED DESCRIPTION

[0051] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0052] See also Figure 1 As shown, the present invention is an adaptive anti-interference communication method based on StarFlash-Bluetooth dual mode, comprising the following steps:

[0053] S1: The dual-mode RF module deployed on the communication terminal monitors the interference parameters of the current channel in real time.

[0054] The interference parameters include: out-of-band radiation power and pulse interference duty cycle at the switching moment;

[0055] S2: Perform time-frequency domain analysis on the out-of-band radiation power, construct a spectrum leakage feature sequence, and calculate the instantaneous interference intensity feature value to assess the degree of spectrum pollution during dual-mode switching.

[0056] S3: Statistically model the pulse interference duty cycle, construct the pulse interference distribution matrix, and calculate the burst noise tolerance eigenvalue to identify whether the channel is subject to periodic strong interference.

[0057] S4: The instantaneous interference intensity characteristic value and the burst noise tolerance characteristic value are combined into a comprehensive anti-interference feature vector, which is input into the pre-trained anti-interference decision model for analysis. The communication mode is then optimized based on the analysis results.

[0058] S5: Dynamically switch between StarFlash and Bluetooth communication modes based on the evaluation results, or enable dual-mode parallel transmission to achieve adaptive anti-interference communication in complex interference environments.

[0059] In S1, the dual-mode radio frequency module deployed on the communication terminal monitors the interference parameters of the current channel in real time. The interference parameters include: the out-of-band radiation power and the pulse interference duty cycle at the switching moment, specifically including:

[0060] At the moment of switching communication modes, the system activates a dedicated high-speed sampling channel, capturing RF signals in real time at a sampling rate of at least 10MS / s. After converting the signal to baseband using digital down-conversion technology, it performs spectrum analysis using a sliding-window fast Fourier transform, focusing on monitoring radiated energy within the ±20MHz range outside the 2.4GHz operating band. The system records three key parameters in real time: the maximum radiated power value in this out-of-band region, the frequency location of maximum radiation, and the average power level in this out-of-band region, forming a complete out-of-band radiation profile.

[0061] The system continuously monitors received signal strength using a highly sensitive envelope detection circuit. Upon detecting a burst exceeding a preset threshold (-60dBm), it immediately initiates high-precision timestamping. Within a set 100ms observation window, the system accurately records the start and end times of each pulse, calculating the ratio of the actual pulse duration to the total observation time as the duty cycle. Simultaneously, the system analyzes the regularity of pulse intervals and distinguishes between random and periodic interference by calculating the coefficient of variation. Ultimately, it outputs a complete description of the pulse interference, including duty cycle, periodicity, and dominant frequency characteristics. Multi-stage anti-aliasing filtering is employed throughout the acquisition process to ensure measurement accuracy.

[0062] In S2, the out-of-band radiation power is analyzed in the time and frequency domains to construct a spectrum leakage feature sequence and calculate the instantaneous interference intensity feature value to assess the degree of spectrum pollution during dual-mode switching. Specifically, the following are performed:

[0063] The out-of-band radiation power at the switching moment of the current channel is obtained in real time, the out-of-band radiation power is analyzed in the time-frequency domain, a spectrum leakage feature sequence is constructed, the instantaneous interference intensity feature value is calculated, and it is determined whether the instantaneous interference intensity feature value is greater than or equal to the preset threshold. If so, the spectrum pollution during dual-mode switching is abnormal; otherwise, the spectrum pollution during dual-mode switching is normal.

[0064] The process of obtaining the instantaneous interference intensity characteristic value is as follows:

[0065] The out-of-band radiation power of the current channel at the switching moment is obtained in real time. The out-of-band radiation power signal at the switching moment is subjected to synchronous compression wavelet transform. The time-frequency energy is decomposed using the Morlet wavelet basis function. The ratio of the maximum out-of-band energy to the average in-band energy is calculated to obtain the maximum energy leakage ratio.

[0066] Among them, the calculation expression of the maximum energy leakage ratio is: ;

[0067] Where, represents the maximum energy leakage ratio, Indicates signal In frequency and time The energy density at Indicates frequency, Indicates time, Indicates the preset out-of-band frequency range, Indicates the preset in-band frequency range, Indicates the number of frequency points in the band, Indicates the out-of-band radiated power signal;

[0068] The sparse time-frequency distribution of the out-of-band radiation power signal is analyzed. The sparse representation of the time-frequency matrix is ​​solved by L1 regularization. The proportion of non-zero elements in the out-of-band region is counted to obtain the sparsity index. The calculation expression is: ;

[0069] Where, represents the sparsity index, represents a sparse coefficient matrix, represents the non-zero element count, Indicates the total number of out-of-band time-frequency grids;

[0070] The maximum energy leakage ratio and the sparsity index are normalized and calculated to obtain the instantaneous interference intensity characteristic value. The calculation expression is: ;

[0071] Where, represents the instantaneous interference intensity characteristic value, and is the preset scale factor, and and Both are greater than 0.

[0072] It should be noted that this step achieves high-precision assessment of spectrum contamination during dual-mode handover through an innovative joint time-frequency analysis method. Its technical effectiveness and innovations lie in the hybrid analysis method of synchronous compressed wavelet transform and sparse time-frequency distribution, which overcomes the time-frequency resolution limitations of traditional fast Fourier transform spectrum analysis. The compressed wavelet transform, using Morlet wavelet basis functions, achieves precise time-frequency localization with μs-level time resolution and 10 MHz-level frequency resolution, effectively capturing transient spectrum leakage characteristics. The sparse representation algorithm based on L1 regularization significantly improves the detection sensitivity of out-of-band interference components. The innovative dual-feature fusion algorithm, combining maximum energy leakage ratio and sparsity index, achieves a comprehensive assessment of both bursty and persistent interference through weighted geometric averaging, reducing the false positive rate of spectrum contamination detection. Furthermore, a dynamic threshold mechanism (updated every 10 ms) adapts to complex electromagnetic environments, providing a key criterion for reliable handover of the StarFlash-Bluetooth dual-mode system.

[0073] In S3, the pulse interference duty cycle is statistically modeled, the pulse interference distribution matrix is ​​constructed, and the burst noise tolerance characteristic value is calculated to identify whether the channel is subject to periodic strong interference. Specifically, the following are performed:

[0074] The pulse interference duty cycle of the current channel is obtained in real time, the pulse interference duty cycle is statistically modeled, the pulse interference distribution matrix is ​​constructed, the burst noise tolerance characteristic value is calculated, and it is determined whether the burst noise tolerance characteristic value is greater than or equal to the preset threshold. If so, the channel is subject to periodic strong interference; if not, the channel is not subject to periodic strong interference.

[0075] The process of obtaining the burst noise tolerance characteristic value is as follows:

[0076] Real-time acquisition of the time series of pulse interference events in the channel;

[0077] The pulse interference is modeled based on the Hawkes process, and its conditional intensity function is defined as: ;

[0078] in, is the background noise rate, which represents the average number of random interference events per unit time, is the pulse intensity, which characterizes the ability of a single interference to stimulate subsequent events. is the attenuation rate, reflecting the duration of the interference effect, Indicates time The conditional intensity function of Indicates the time when the historical interference event occurred, Indicates the time when the current interference event occurs. Represents the base of natural numbers;

[0079] Solve for parameters using maximum likelihood estimation 、 and , specifically achieved by optimizing the log-likelihood function: ;

[0080] in, represents the log-likelihood function, represents the total number of interference events within the observation window, Indicates the first Interference events, represents the length of the observation time window, represents the integral of the conditional intensity function within the observation window, represents the sum of the logarithms of the conditional intensities at the moment of the interference event;

[0081] According to the average pulse interval and pulse duration, the burst noise tolerance characteristic value is calculated. The calculation expression is: ;

[0082] Where, Indicates the burst noise tolerance characteristic value.

[0083] It should be noted that this step achieves accurate identification of periodic pulse interference through an innovative Hawkes process modeling approach. Its technical benefits and innovations are primarily reflected in the following: The use of a self-excited point process model overcomes the limitations of traditional duty cycle statistical methods in interference correlation analysis. Three key parameters in the conditional intensity function (background noise rate μ, pulse intensity α, and decay rate β) simultaneously characterize the spatiotemporal coupling of random and periodic interference. A real-time parameter solution algorithm based on maximum likelihood estimation dynamically models interference event sequences, improving period detection accuracy by more than three times compared to traditional Fourier analysis methods. An innovative burst noise tolerance eigenvalue, through normalization of the average pulse interval / duration, achieves the first joint quantification of interference periodicity and energy intensity. Combined with an adaptive threshold mechanism, this provides a key basis for interference mitigation decision-making in dual-mode systems. This model particularly optimizes the ability to distinguish between arc interference and Bluetooth packet collisions in industrial environments.

[0084] In S4, the instantaneous interference intensity characteristic value and the burst noise tolerance characteristic value are fused into a comprehensive anti-interference feature vector, which is input into the pre-trained anti-interference decision model for analysis. Based on the analysis results, the communication mode optimization evaluation is performed, specifically including:

[0085] The instantaneous interference intensity characteristic value and the burst noise tolerance characteristic value are obtained, and the instantaneous interference intensity characteristic value and the burst noise tolerance characteristic value are constructed into a comprehensive anti-interference feature vector as the input of the anti-interference decision model. Minimizing the error between the predicted communication quality score and the actual communication quality score is used as the training goal of the anti-interference decision model. The anti-interference decision model is trained, and the communication quality score is output according to the trained anti-interference decision model. The anti-interference decision model is a gradient boosting tree model.

[0086] The training process of the anti-interference decision model is:

[0087] Construct a training data set, the training data set including: historically collected instantaneous interference intensity characteristic value samples; historically collected burst noise tolerance characteristic value samples; corresponding actual communication quality score labels; standardize the characteristic values ​​so that they obey a standard normal distribution with a mean of 0 and a variance of 1; use the SMOTE algorithm to oversample the sample imbalance data; construct a gradient boosting tree model, the model parameters include: the number of base learners is set to 100-200; the learning rate is set to 0.05-0.2; the maximum depth of the tree is set to 3-5 layers; the minimum number of leaf node samples is set to 5-10; use the k-fold cross-validation method to train the model, specifically including: randomly dividing the data set into k mutually exclusive subsets; selecting k-1 subsets as training sets each time, and the remaining 1 subset as validation set; repeating the training process k times, and taking the average performance as the model evaluation indicator;

[0088] During the model optimization phase, Bayesian optimization is used to tune hyperparameters, with the goal of minimizing the mean absolute error on the validation set.

[0089] The optimal model parameters after training are solidified and deployed to the embedded system of the communication terminal. The communication mode optimization evaluation based on the analysis results specifically includes:

[0090] Based on the communication quality score output by the anti-interference decision model, a three-level optimization strategy including star flash mode, Bluetooth mode and dual-mode parallel transmission is established: if the score is lower than the first threshold, it is preferentially switched to the high-interference anti-interference star flash mode; if the score is between the first and second thresholds, dual-mode parallel transmission is enabled to take into account both reliability and compatibility; if the score is higher than the second threshold, the low-power Bluetooth mode is adopted; at the same time, the channel state changes are monitored in real time. When it is detected that the score crosses the threshold boundary N times in a row, the mode switching is triggered, and the communication quality is further optimized by dynamically adjusting the transmission power and modulation method, where N is the preset hysteresis count parameter.

[0091] In S5, the StarFlash and Bluetooth communication modes are dynamically switched based on the evaluation results, or dual-mode parallel transmission is enabled to achieve adaptive anti-interference communication in complex interference environments. Specifically,

[0092] Based on the real-time monitoring of out-of-band radiation power and pulse interference duty cycle, the instantaneous interference intensity eigenvalue and burst noise tolerance eigenvalue are calculated respectively through synchronous compression wavelet transform and Hawkes process modeling, and the two are integrated into a comprehensive anti-interference feature vector input into the pre-trained gradient boosting tree decision model to obtain the communication quality score of the current channel state; a three-level optimization strategy is implemented according to the scoring results: when the score is lower than the first threshold, the star flash mode is preferentially adopted to ensure communication reliability; when the score is between the first and second thresholds, dual-mode parallel transmission is enabled to balance performance and compatibility; when the score is higher than the second threshold, it switches to Bluetooth mode to reduce power consumption. At the same time, a hysteresis counting mechanism is used to avoid frequent mode switching.

[0093] During the mode switching process, the transmission power and modulation method are dynamically adjusted to further optimize the communication quality. For the Star Flash mode, high-order modulation and maximum transmission power are prioritized, while the Bluetooth mode adaptively selects GFSK or LE Coded PHY according to the degree of interference. During dual-mode parallel transmission, time division multiplexing is used to coordinate the resource allocation of the two protocols, and the changes in characteristic values ​​are monitored in real time. When the interference characteristics are detected to be persistently abnormal, the mode is re-evaluated to form a closed-loop anti-interference control mechanism.

[0094] See also Figure 2 As shown, an adaptive anti-interference communication system based on Starflash-Bluetooth dual mode includes:

[0095] A real-time interference monitoring module, which monitors the interference parameters of the current channel in real time through a dual-mode radio frequency module deployed on the communication terminal;

[0096] The interference parameters include: out-of-band radiation power and pulse interference duty cycle at the switching moment;

[0097] A spectrum pollution analysis module, which is used to perform time-frequency domain analysis on out-of-band radiation power, construct a spectrum leakage feature sequence, and calculate the instantaneous interference intensity feature value to evaluate the degree of spectrum pollution during dual-mode switching;

[0098] A pulse interference modeling module, which performs statistical modeling on the pulse interference duty cycle, constructs a pulse interference distribution matrix, and calculates a burst noise tolerance eigenvalue to identify whether the channel is subject to periodic strong interference;

[0099] An intelligent decision-making module, which combines the instantaneous interference intensity characteristic value and the burst noise tolerance characteristic value into a comprehensive anti-interference feature vector, inputs it into a pre-trained anti-interference decision model for analysis, and performs communication mode optimization evaluation based on the analysis results;

[0100] A dynamic switching execution module dynamically switches between StarFlash and Bluetooth communication modes according to the evaluation results, or enables dual-mode parallel transmission to achieve adaptive anti-interference communication in complex interference environments.

[0101] Working principle of the present invention: The present invention realizes reliable communication in complex environments through multi-dimensional interference feature extraction and intelligent decision-making. The system first uses a dual-mode RF module to monitor channel interference parameters in real time, including high-speed sampling and sliding window fast Fourier transform analysis to obtain out-of-band radiation power characteristics, and envelope detection and high-precision timestamp recording of pulse interference duty cycle. For spectrum pollution assessment, the synchronous compression wavelet transform and sparse time-frequency distribution analysis are innovatively combined to construct: maximum energy leakage ratio-sparseness index, dual feature fusion algorithm, to achieve accurate interference positioning in μs time domain and 10MHz frequency domain. For periodic interference identification, Hawkes process modeling is adopted, and the conditional intensity function parameters (μ, α, β) are used to quantify the spatiotemporal characteristics of interference, and the burst noise tolerance characteristic value is designed to realize the joint evaluation of periodicity and intensity. The two features are fused and fed into a pretrained gradient boosting tree model. This model, trained through Bayesian optimization and k-fold cross-validation, outputs a communication quality score and implements a three-tiered optimization strategy: switching to the high-interference Star Flash mode for low scores, enabling dual-mode parallel operation for medium scores, and adopting Bluetooth Low Energy mode for high scores. The system innovatively introduces a hysteresis counting mechanism to prevent frequent switching, and further optimizes performance through dynamic power control and adaptive modulation (Star Flash high-order modulation / Bluetooth GFSK / LE Coded PHY). Time division multiplexing coordinates resource allocation when operating in dual-mode parallel operation, forming a closed-loop control system encompassing feature monitoring, analysis, decision-making, and execution. The overall response time is kept within 10ms, significantly improving communication reliability in complex electromagnetic environments.

[0102] The above formulas are all dimensionless and numerical calculations. The formulas are obtained by collecting a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters in the formulas are set by technicians in this field according to actual conditions.

[0103] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer program are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center via wired or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0104] It should be understood that the term "and / or" as used herein simply describes a relationship between associated objects, indicating that three possible relationships exist. For example, "A and / or B" can represent: A alone, A and B together, or B alone. A and B can be singular or plural. Furthermore, the character " / " as used herein generally indicates an "or" relationship between the associated objects, but it may also indicate an "and / or" relationship. For specific understanding, please refer to the context.

[0105] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0106] The above is a detailed description of an embodiment of the present invention. However, the content described is only a preferred embodiment of the present invention and should not be considered to limit the scope of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. An adaptive anti-interference communication method based on Starflash-Bluetooth dual mode, characterized in that: The following steps are involved: S1: The dual-mode RF module deployed on the communication terminal monitors the interference parameters of the current channel in real time. The interference parameters include: out-of-band radiation power and pulse interference duty cycle at the switching moment; S2: Perform time-frequency domain analysis on the out-of-band radiation power, construct a spectrum leakage feature sequence, and calculate the instantaneous interference intensity feature value to assess the degree of spectrum pollution during dual-mode switching. S3: Statistically model the pulse interference duty cycle, construct the pulse interference distribution matrix, and calculate the burst noise tolerance eigenvalue to identify whether the channel is subject to periodic strong interference. S4: The instantaneous interference intensity characteristic value and the burst noise tolerance characteristic value are combined into a comprehensive anti-interference feature vector, which is input into the pre-trained anti-interference decision model for analysis. The communication mode is then optimized based on the analysis results. S5: Dynamically switch between StarFlash and Bluetooth communication modes based on the evaluation results, or enable dual-mode parallel transmission to achieve adaptive anti-interference communication in complex interference environments.

2. The adaptive anti-interference communication method based on StarFlash-Bluetooth dual mode according to claim 1 is characterized in that: The evaluating of the spectrum pollution degree during dual-mode switching specifically includes: The out-of-band radiation power at the switching moment of the current channel is obtained in real time, the out-of-band radiation power is analyzed in the time-frequency domain, a spectrum leakage feature sequence is constructed, the instantaneous interference intensity feature value is calculated, and it is determined whether the instantaneous interference intensity feature value is greater than or equal to the preset threshold. If so, the spectrum pollution during dual-mode switching is abnormal; otherwise, the spectrum pollution during dual-mode switching is normal.

3. The adaptive anti-interference communication method based on StarFlash-Bluetooth dual mode according to claim 2 is characterized in that: The process of obtaining the instantaneous interference intensity characteristic value is as follows: The out-of-band radiation power of the current channel at the switching moment is obtained in real time. The out-of-band radiation power signal at the switching moment is subjected to synchronous compression wavelet transform. The time-frequency energy is decomposed using the Morlet wavelet basis function. The ratio of the maximum out-of-band energy to the average in-band energy is calculated to obtain the maximum energy leakage ratio. The out-of-band radiation power signal is subjected to sparse time-frequency distribution analysis. The sparse representation of the time-frequency matrix is ​​solved by L1 regularization. The proportion of non-zero elements in the out-of-band region is counted to obtain the sparsity index. The maximum energy leakage ratio and the sparsity index are normalized and calculated to obtain the instantaneous interference intensity characteristic value.

4. The adaptive anti-interference communication method based on StarFlash-Bluetooth dual mode according to claim 1, characterized in that: The identifying whether the channel is subject to periodic strong interference specifically includes: The pulse interference duty cycle of the current channel is obtained in real time, the pulse interference duty cycle is statistically modeled, the pulse interference distribution matrix is ​​constructed, the burst noise tolerance characteristic value is calculated, and it is determined whether the burst noise tolerance characteristic value is greater than or equal to the preset threshold. If so, the channel is subject to periodic strong interference; if not, the channel is not subject to periodic strong interference.

5. The adaptive anti-interference communication method based on StarFlash-Bluetooth dual mode according to claim 4 is characterized in that: The process of obtaining the burst noise tolerance characteristic value is as follows: Real-time acquisition of the time series of pulse interference events in the channel; Modeling of pulse interference based on Hawkes process; Solve the parameters by maximum likelihood estimation 、 and , specifically achieved by optimizing the log-likelihood function; The burst noise tolerance characteristic value is calculated based on the average pulse interval and pulse duration.

6. The adaptive anti-interference communication method based on StarFlash-Bluetooth dual mode according to claim 1, characterized in that: The instantaneous interference intensity characteristic value and the burst noise tolerance characteristic value are integrated into a comprehensive anti-interference feature vector, which is input into the pre-trained anti-interference decision model for analysis, specifically including: The instantaneous interference intensity characteristic value and the burst noise tolerance characteristic value are obtained, and the instantaneous interference intensity characteristic value and the burst noise tolerance characteristic value are constructed into a comprehensive anti-interference feature vector as the input of the anti-interference decision model. Minimizing the error between the predicted communication quality score and the actual communication quality score is used as the training goal of the anti-interference decision model. The anti-interference decision model is trained, and the communication quality score is output according to the trained anti-interference decision model. The anti-interference decision model is a gradient boosting tree model.

7. The adaptive anti-interference communication method based on StarFlash-Bluetooth dual mode according to claim 6, characterized in that: The training process of the anti-interference decision model is: Construct a training data set, the training data set including: historically collected instantaneous interference intensity characteristic value samples; historically collected burst noise tolerance characteristic value samples; corresponding actual communication quality score labels; standardize the characteristic values ​​so that they obey a standard normal distribution with a mean of 0 and a variance of 1; use the SMOTE algorithm to oversample the sample imbalance data; construct a gradient boosting tree model, the model parameters including: the number of basic learners is set to 100 to 200; the learning rate is set to 0.05-0.2; the maximum depth of the tree is set to 3-5 layers; the minimum number of leaf node samples is set to 5-10; use the k-fold cross-validation method to train the model, specifically including: randomly dividing the data set into k mutually exclusive subsets; selecting k-1 subsets as training sets each time, and the remaining 1 subset as a validation set; repeating the training process k times, and taking the average performance as the model evaluation indicator; During the model optimization phase, Bayesian optimization is used to tune hyperparameters, with the goal of minimizing the mean absolute error on the validation set. The optimal model parameters after training are solidified and deployed to the embedded system of the communication terminal.

8. The adaptive anti-interference communication method based on Starflash-Bluetooth dual mode according to claim 1 is characterized in that: The communication mode optimization evaluation according to the analysis results specifically includes: Based on the communication quality score output by the anti-interference decision model, a three-level optimization strategy including star flash mode, Bluetooth mode and dual-mode parallel transmission is established: if the score is lower than the first threshold, it is preferentially switched to the high-interference anti-interference star flash mode; if the score is between the first and second thresholds, dual-mode parallel transmission is enabled to take into account both reliability and compatibility; if the score is higher than the second threshold, the low-power Bluetooth mode is adopted; at the same time, the channel state changes are monitored in real time. When it is detected that the score crosses the threshold boundary N times in a row, the mode switching is triggered, and the communication quality is further optimized by dynamically adjusting the transmission power and modulation method, where N is the preset hysteresis count parameter.

9. The adaptive anti-interference communication method based on Starflash-Bluetooth dual mode according to claim 1, characterized in that: The method of dynamically switching between the StarFlash and Bluetooth communication modes or enabling dual-mode parallel transmission based on the evaluation results to achieve adaptive anti-interference communication in complex interference environments specifically includes: A three-level optimization strategy is implemented based on the scoring results: when the score is below the first threshold, the star flash mode is prioritized to ensure communication reliability; when the score is between the first and second thresholds, dual-mode parallel transmission is enabled to balance performance and compatibility; when the score is above the second threshold, the Bluetooth mode is switched to reduce power consumption, while a hysteresis counting mechanism is used to avoid frequent mode switching; During the mode switching process, the transmission power and modulation method are dynamically adjusted to further optimize the communication quality. For the Star Flash mode, high-order modulation and maximum transmission power are prioritized, while the Bluetooth mode adaptively selects GFSK or LECoded PHY according to the degree of interference. During dual-mode parallel transmission, time division multiplexing is used to coordinate the resource allocation of the two protocols, and the changes in characteristic values ​​are monitored in real time. When the interference characteristics are detected to be persistently abnormal, the mode is re-evaluated, forming a closed-loop anti-interference control mechanism.

10. An adaptive anti-interference communication system based on Starflash-Bluetooth dual mode, characterized in that: An adaptive anti-interference communication method based on Starflash-Bluetooth dual mode as claimed in any one of claims 1 to 9, comprising: A real-time interference monitoring module, which monitors the interference parameters of the current channel in real time through a dual-mode radio frequency module deployed on the communication terminal; The interference parameters include: out-of-band radiation power and pulse interference duty cycle at the switching moment; A spectrum pollution analysis module, which is used to perform time-frequency domain analysis on out-of-band radiation power, construct a spectrum leakage feature sequence, and calculate the instantaneous interference intensity feature value to evaluate the degree of spectrum pollution during dual-mode switching; A pulse interference modeling module, which performs statistical modeling on the pulse interference duty cycle, constructs a pulse interference distribution matrix, and calculates a burst noise tolerance eigenvalue to identify whether the channel is subject to periodic strong interference; An intelligent decision-making module, which combines the instantaneous interference intensity characteristic value and the burst noise tolerance characteristic value into a comprehensive anti-interference feature vector, inputs it into a pre-trained anti-interference decision model for analysis, and performs communication mode optimization evaluation based on the analysis results; A dynamic switching execution module dynamically switches between the StarFlash and Bluetooth communication modes according to the evaluation results, or enables dual-mode parallel transmission to achieve adaptive anti-interference communication in complex interference environments.

Images