Signal extraction method and device based on ica-r algorithm and medium

Abstract

The application discloses a signal extraction method and device based on an ICA-R algorithm and a medium, and aims to solve the problem of co-frequency signal interference. The scheme generates a cosine pulse signal with the same carrier frequency as the target signal as a reference signal, which can still effectively separate the carrier frequency in the absence of target signal symbol information; the threshold of the characteristic parameter is set by using the prior information of the target signal, and after the estimated signal is obtained, whether the estimated signal meets the preset requirements is judged through the threshold of the characteristic parameter. The introduction of the threshold and the dynamic verification of the extracted signal ensure that the accuracy of the extraction result can be continuously optimized until the preset optimization number is reached or the estimated signal meets the requirements of the target signal. The application overcomes the defects of insufficient accuracy in the current technology to a certain extent, improves the robustness of the ICA-R algorithm in accurately extracting the target signal, and meets the high requirements of the communication field for signal extraction quality.

Description

Technical Field

[0001] This invention relates to the field of blind source separation algorithm optimization, and in particular to a signal extraction method, apparatus and medium based on the ICA-R algorithm. Background Technology

[0002] With the rapid development of electronic and communication technologies, a large number of frequency-using devices such as radios, mobile phones, and satellite communication terminals have emerged, making the electromagnetic environment of radio communication increasingly complex. Mutual interference and even co-channel interference are highly likely to occur between various frequency-using devices, severely impacting communication quality. Traditional anti-interference methods based on filtering, frequency hopping, direct sequence spread spectrum, and antenna nulling cannot effectively handle co-channel interference. As a digital signal processing technique based on the assumption of signal independence, the ICA (Independent Component Analysis) algorithm can achieve blind separation of co-channel signals under conditions of unknown signal sources and signal mixing methods. By utilizing some prior information about the target signal, such as carrier frequency, phase, and symbol rate, a reference signal can be constructed to constrain the optimization objective of the ICA algorithm, directly extracting one or more target signals of interest.

[0003] The ICA algorithm with a reference signal constraint is called ICA-R (Independent Component Analysis with Reference). The original ICA-R algorithm uses the symbol function of the target signal as the reference signal, achieving a target signal extraction accuracy of over 90%. However, in applications such as communication and reconnaissance, the receiver cannot know the symbol information of the target signal in advance, thus making it impossible to construct the required reference signal. If a cosine pulse signal with the same frequency as the carrier of the target signal is used as the reference signal, the target signal extraction accuracy will drop to around 15%. This accuracy cannot meet the engineering application requirements in fields such as communication and reconnaissance.

[0004] To address the aforementioned technical problems, existing technologies employ non-random initialization of the separation vector or use metaheuristic optimization algorithms to optimize the objective function of ICA-R, thereby improving the accuracy of target signal extraction. However, the accuracy of the target signal obtained using these two methods in existing technologies remains low. Summary of the Invention

[0005] The purpose of this invention is to provide a signal extraction method, device, and medium based on the ICA-R algorithm, which to a certain extent overcomes the shortcomings of insufficient accuracy in the current technology, improves the robustness of the ICA-R algorithm in accurately extracting target signals, and meets the high requirements of the communication field for signal extraction quality.

[0006] To address the aforementioned technical problems, this invention provides a signal extraction method based on the ICA-R algorithm, comprising:

[0007] Receive input signals, the input signals including target signals and interference signals;

[0008] A reference signal is generated based on the carrier frequency of the target signal, wherein the reference signal is a cosine pulse signal with the same carrier frequency as the target signal;

[0009] Calculate feature parameters based on prior information of the target signal, and set threshold values ​​for the feature parameters;

[0010] The estimated signal of the target signal is extracted from the input signal according to the ICA-R algorithm and the reference signal;

[0011] Extract the feature parameters from the estimated signal, and determine whether the estimated signal meets the preset requirements based on the feature parameters and the feature parameter threshold.

[0012] If the condition is not met, the process returns to the step of extracting the estimated signal of the target signal from the input signal according to the ICA-R algorithm and the reference signal, until the estimated signal meets the preset requirements.

[0013] Preferably, calculating feature parameters based on prior information of the target signal includes:

[0014] Perform a Fourier transform on the target signal to obtain a frequency domain signal;

[0015] Calculate the power spectral density of the frequency domain signal;

[0016] The power spectral density symmetry coefficient is calculated based on the power spectral density.

[0017] Preferably, the expression for the power spectral density symmetry coefficient is:

[0018] Wherein, PSI is the power spectral density symmetry coefficient, M is the number of sampling points in the power spectral density that deviate from the center frequency, k represents the k-th deviation sampling point, and Y... L (k) represents the power spectral density value of the k-th sampling point in the portion of the target signal where the power spectral density is less than the center frequency, Y R (k) represents the power spectral density value of the k-th sampling point in the portion of the target signal where the power spectral density is greater than the center frequency. (k) is Y R (k) horizontal flip, Represents the modulus of a vector.

[0019] Preferably, the expression for the power spectral density symmetry coefficient is:

[0020] Wherein, PSI is the power spectral density symmetry coefficient, M is the number of sampling points in the power spectral density that deviate from the center frequency, k represents the k-th deviation sampling point, and Y... L (k) represents the power spectral density value of the k-th sampling point in the portion of the target signal where the power spectral density is less than the center frequency, Y R (k) represents the power spectral density value of the k-th sampling point in the portion of the target signal where the power spectral density is greater than the center frequency. (k) is Y R (k) horizontal flip, The modulus of a vector. This represents the second-order normal form.

[0021] Preferably, calculating the feature parameter threshold based on prior information of the target signal includes:

[0022] The target signal is subjected to Fourier transform to obtain the frequency domain signal.

[0023] Calculate the power spectral density of the frequency domain signal;

[0024] Calculate the first power value within the signal bandwidth based on the distribution of the power spectral density;

[0025] Calculate the second power value within the same bandwidth outside the signal bandwidth based on the distribution of the power spectral density;

[0026] Calculate the ratio between the first power value and the second power value to obtain the in-band and out-of-band power ratio.

[0027] Preferably, the feature parameter is a power spectral density symmetry coefficient. Determining whether the estimated signal meets the preset requirements based on the feature parameter and the feature parameter threshold includes:

[0028] Determine whether the power spectral density coefficient in the estimated signal is greater than the power spectral density coefficient threshold;

[0029] If the power spectral density coefficient is greater than the threshold value, then the estimated signal is determined to meet the preset requirement.

[0030] If the power spectral density coefficient is not greater than the threshold value, then the estimated signal is determined not to meet the preset requirements.

[0031] Preferably, after determining that the preset requirements are not met, the method further includes:

[0032] Determine whether the number of times the step of extracting the estimated signal of the target signal from the input signal according to the ICA-R algorithm and the reference signal is performed is greater than the set maximum number of optimizations;

[0033] If the number of optimization attempts exceeds the set maximum number, then the estimated signal is determined to be the target signal.

[0034] If the number of optimization attempts is not greater than the set maximum number of attempts, then the process re-enters the step of extracting the estimated signal of the target signal from the input signal according to the ICA-R algorithm and the reference signal.

[0035] Preferably, the ICA-R algorithm includes a built-in number of iterations, and after determining that the number of times the step of extracting the estimated signal of the target signal from the input signal according to the ICA-R algorithm and the reference signal is performed is not greater than the set maximum number of optimizations, it further includes:

[0036] The built-in iteration count of the ICA-R algorithm is reset to 0, and the process re-enters the step of extracting the estimated signal of the target signal from the input signal according to the ICA-R algorithm and the reference signal.

[0037] To address the aforementioned technical problems, this invention provides a signal extraction device based on the ICA-R algorithm, comprising:

[0038] Memory, used to store computer programs;

[0039] A processor is used to implement the steps of the signal extraction method based on the ICA-R algorithm described above when executing a computer program.

[0040] To address the aforementioned technical problems, this invention provides a non-volatile storage medium storing a computer program, which, when executed by a processor, implements the steps of the signal extraction method based on the ICA-R algorithm described above.

[0041] This invention provides a signal extraction method, apparatus, and medium based on the ICA-R algorithm, aiming to solve the problem of co-channel signal interference. This scheme generates a cosine pulse signal with the same carrier frequency as the target signal as a reference signal, enabling effective separation using the carrier frequency even in the absence of target signal symbol information. It sets a feature parameter threshold using prior information of the target signal. After obtaining the estimated signal, the threshold is used to determine whether the estimated signal meets preset requirements. The introduction of this threshold and the dynamic verification of the extracted signal ensure that the accuracy of the extraction result can be continuously optimized until a preset number of optimizations is reached or the estimated signal meets the requirements of the target signal. This application not only improves the reliability of signal extraction but also overcomes the shortcomings of insufficient accuracy in current technologies to a certain extent, improving the robustness of the ICA-R algorithm in accurately extracting target signals and meeting the high requirements for signal extraction quality in the communication field. Attached Figure Description

[0042] To more clearly illustrate the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0043] Figure 1 This invention provides a schematic flowchart of a signal extraction method based on the ICA-R algorithm.

[0044] Figure 2 This invention provides a schematic diagram for calculating the power spectral density symmetry coefficient.

[0045] Figure 3 A block diagram of a target signal blind extraction system based on the ICA-R algorithm provided by the present invention;

[0046] Figure 4 A schematic diagram of an ICA-R algorithm processing module provided by the present invention;

[0047] Figure 5 A flowchart of an algorithm for extracting a target signal using prior information of the target signal is provided for this invention.

[0048] Figure 6 This invention provides a graph showing the change in target signal extraction accuracy with the number of algorithm reruns when setting a threshold using PSI as a feature parameter.

[0049] Figure 7 This is a schematic diagram of a signal extraction device based on the ICA-R algorithm provided by the present invention. Detailed Implementation

[0050] The core of this invention is to provide a signal extraction method, device, and medium based on the ICA-R algorithm, which to a certain extent overcomes the shortcomings of insufficient accuracy in the current technology, improves the robustness of the ICA-R algorithm in accurately extracting target signals, and meets the high requirements of the communication field for signal extraction quality.

[0051] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0052] First, we introduce the existing ICA algorithms, specifically ICA-R (Independent Component Analysis with Reference) which uses a reference signal constraint. ICA-R algorithms include equality and inequality constraints. They typically use the Lagrange multiplier method to transform the constrained optimization problem into an unconstrained one, and then use a Newton-like optimization method with a learning rate to solve for the optimal value of the separation vector. Because the Newton-like optimization method is sensitive to the learning rate and the initial value of the separation vector, inappropriate initial values ​​may cause the algorithm to fail to converge to the global optimum or even diverge. Global convergence is only guaranteed when the initial point is sufficiently close to the global optimum. Therefore, when the initial value of the separation vector is generated randomly, the accuracy of the ICA-R algorithm in extracting a target signal is probabilistic. Consequently, different ICA-R algorithms exhibit varying robustness in accurately extracting a target signal.

[0053] The original ICA-R algorithm uses the symbol function of the target signal as a reference signal, achieving a target signal extraction accuracy of over 90%. However, in applications such as communication and reconnaissance, the receiver cannot know the symbol information of the target signal in advance, thus making it impossible to construct the required reference signal. If a cosine pulse signal with the same frequency as the carrier of the target signal is used as the reference signal, the target signal extraction accuracy will drop to around 15%. This accuracy cannot meet the engineering application requirements in fields such as communication and reconnaissance.

[0054] To address the aforementioned technical problems, existing technologies employ a non-random initialization of the separation vector to improve extraction accuracy. One approach is to assume the reference signal r(t) is the convergence target of the ICA-R algorithm. Based on the signal extraction formula... The initialization formula for the separating vector can be obtained through the Moore-Penrose generalized inverse operation:

[0055] ;

[0056] in The Moore-Penrose generalized inverse matrix is ​​called X, where X is the whitened observed signal. Compared to the original ICA-R algorithm with random initialization, this method uses reference prior information not only in the contrast function but also in the initialization of the separation vector. This method can improve the accuracy of extracting the target signal to about 95%. Nevertheless, for some communication applications with high reliability requirements, this accuracy still needs to be further improved. In addition, this method requires the calculation of the pseudo-inverse of the signal matrix, which has high computational complexity, making it complex for engineering implementation and requiring a large amount of memory resources.

[0057] Another approach is to use metaheuristic optimization algorithms to optimize the objective function of ICA-R. Metaheuristic optimization algorithms are a class of algorithms that solve complex optimization problems by simulating phenomena or processes in nature. This method has advantages such as strong global search capability and good robustness, but it also has disadvantages such as slow convergence speed, difficulty in parameter tuning, high computational resource consumption, and susceptibility to getting trapped in local optima.

[0058] To address the aforementioned technical problems, this invention provides a signal extraction method based on the ICA-R algorithm, such as... Figure 1 As shown, Figure 1 This invention provides a flowchart illustrating a signal extraction method based on the ICA-R algorithm, which includes:

[0059] S11: Receive input signals, including target signals and interference signals;

[0060] During signal reception, the input signal typically contains both the target signal (the useful information to be extracted) and interference signals (which may be interference or noise). This is because the real-world communication environment is complex and variable, filled with various unpredictable electromagnetic waves and physical phenomena. These phenomena may originate from nature, such as lightning and solar radiation, or from human activities, such as the operation of other electronic devices, radio broadcasts, and industrial equipment. The electromagnetic waves generated by these external factors mix with the target signal, forming interference signals. Furthermore, signals may be affected by physical effects such as attenuation, reflection, refraction, and scattering during transmission, leading to a decrease in signal quality and further increasing the interference signal component. Therefore, the received input signal inevitably contains both the target signal and interference signals. Input signal processing requires effective separation to identify and extract the target signal in subsequent steps. The quality of this step directly affects the accuracy of subsequent signal processing; ensuring the integrity and accuracy of the input signal is crucial.

[0061] S12: Generate a reference signal based on the carrier frequency of the target signal. The reference signal is a cosine pulse signal with the same carrier frequency as the target signal.

[0062] A reference signal is generated based on the carrier frequency of the target signal. This reference signal is a cosine pulse signal with the same carrier frequency as the target signal. This reference signal provides a benchmark, which helps in signal separation in the subsequent ICA-R algorithm. The generated reference signal should have good time and frequency domain characteristics to preserve the characteristics of the target signal to the greatest extent possible, providing effective support for subsequent blind signal separation. As a preferred embodiment, the relationship of the cosine reference signal is: r(t) = cos(2πf c t); where r(t) is the reference signal, f c Let t be the carrier frequency of the target signal and t be time.

[0063] S13: Calculate the feature parameters based on the prior information of the target signal and set the feature parameter thresholds;

[0064] Feature parameters are calculated using prior information about the target signal. These parameters can include characteristics such as signal amplitude, phase, and frequency. Based on these feature parameters, thresholds are set, and the selection of these thresholds is crucial for the effective extraction of the target signal. Appropriate feature parameter thresholds can effectively suppress interference signals and improve the accuracy of target signal extraction.

[0065] As a preferred embodiment, calculating feature parameters based on prior information of the target signal includes: performing a Fourier transform on the target signal to obtain a frequency domain signal; calculating the power spectral density of the frequency domain signal; and calculating the power spectral density symmetry coefficients based on the power spectral density. In this preferred embodiment, to calculate feature parameters based on prior information of the target signal, a Fourier transform is first performed on the target signal, converting it from the time domain to the frequency domain. The Fourier transform reveals the frequency components of the signal and their amplitude and phase characteristics at different frequencies. Next, the power spectral density of the frequency domain signal is calculated. This process involves statistically analyzing the squared amplitude of the frequency domain signal to obtain the energy distribution of the signal at each frequency. The power spectral density not only reflects the main frequency components of the signal but also reveals potential noise and interference. Finally, the power spectral density symmetry coefficients are calculated based on the power spectral density to analyze the spectral characteristics and symmetry of the signal, helping to identify the nature and quality of the signal. The value of the symmetry coefficients can indicate whether the signal is evenly distributed within the positive and negative frequency ranges, which is crucial for further signal processing and target signal extraction. Through this series of steps, the frequency domain characteristics can be fully utilized to provide an effective basis for target signal extraction.

[0066] In a preferred embodiment, the characteristic parameter is the power spectral density symmetry coefficient. Determining whether the estimated signal meets preset requirements based on the characteristic parameter and its threshold includes: judging whether the power spectral density coefficient in the estimated signal is greater than a power spectral density coefficient threshold; if it is greater than the threshold, the estimated signal is determined to meet the preset requirements; if it is not greater than the threshold, the estimated signal is determined not to meet the preset requirements. Specifically, it first judges whether the power spectral density coefficient in the estimated signal is greater than a set power spectral density coefficient threshold. This threshold is set based on the results of previous signal analysis and aims to provide a benchmark to distinguish between valid signals and interference or noise signals. If the power spectral density coefficient of the estimated signal is greater than the threshold, the signal is determined to meet the preset requirements, meaning that the signal characteristics match the target signal and have good signal quality and usability. Conversely, if the power spectral density coefficient is not greater than the threshold, the estimated signal is determined not to meet the preset requirements, which may indicate that the signal is affected by interference or noise and requires further processing or re-extraction. This judgment process provides an effective quality control mechanism for signal extraction, ensuring that the finally extracted signal has high reliability and accuracy.

[0067] As a preferred embodiment, the expression for the power spectral density symmetry coefficient is:

[0068] Where PSI is the power spectral density symmetry coefficient, M is the number of sampling points in the power spectral density that deviate from the center frequency, k represents the k-th sampling point that deviates, and Y... L (k) represents the power spectral density value of the k-th sampling point in the portion of the target signal where the power spectral density is less than the center frequency, Y R (k) represents the power spectral density value of the k-th sampling point in the portion of the target signal where the power spectral density is greater than the center frequency. (k) is Y R (k) horizontal flip, Represents the modulus of a vector.

[0069] Figure 2 This invention provides a schematic diagram for calculating the power spectral density symmetry coefficient, wherein... Y represents the carrier frequency of the target signal. L Y represents the portion of the target signal whose power spectral density is less than the center frequency. R This represents the portion of the target signal whose power spectral density is greater than the center frequency. It represents the distance by which the signal frequency deviates from the center frequency.

[0070] Assuming the target signal is a digitally modulated communication signal, then for the receiver, the carrier frequency of the communication signal is... This is known prior information. The power spectral density symmetry coefficient (PSI) of the signal is defined as follows:

[0071] ;

[0072] in yes Horizontal flip.

[0073] In the absence of noise and interference signals, the power spectral density of a pure digital modulated communication signal is... It is centrally symmetric, i.e., PSI=0. However, after the signal is superimposed with Gaussian white noise, due to the random perturbation of the Gaussian white noise, the power spectral density of the mixed signal will no longer be centrally symmetric. For a perfectly symmetrical system, the PSI will become a value greater than 0. If a power spectral density other than PSI is superimposed... For centrally symmetric interference signals, the PSI value of the mixed signal will become larger. Therefore, the PSI value under a certain signal-to-noise ratio condition can be calculated in advance and then used as a characteristic parameter threshold for determining whether the estimated signal output by the ICA-R algorithm is the target signal.

[0074] As a preferred embodiment, the expression for the power spectral density symmetry coefficient is:

[0075] Where PSI is the power spectral density symmetry coefficient, M is the number of sampling points in the power spectral density that deviate from the center frequency, k represents the k-th sampling point that deviates, and Y... L (k) represents the power spectral density value of the k-th sampling point in the portion of the target signal where the power spectral density is less than the center frequency, Y R (k) represents the power spectral density value of the k-th sampling point in the portion of the target signal where the power spectral density is greater than the center frequency. (k) is Y R (k) horizontal flip, The modulus of a vector. This represents the second-order normal form. The PSI exponent ranges from [0,1]. The larger the exponent, the stronger the spectral symmetry of the signal. Ideally, the PSI is 0, but it becomes a non-zero value after noise is added. This value can be calculated in advance and used as a parameter threshold to determine whether the algorithm accurately extracts the target signal.

[0076] In a preferred embodiment, calculating the characteristic parameter threshold based on prior information of the target signal includes: performing a Fourier transform on the target signal to obtain a frequency domain signal; calculating the power spectral density of the frequency domain signal; calculating a first power value within the signal bandwidth based on the distribution of the power spectral density; calculating a second power value within the same bandwidth outside the signal bandwidth based on the distribution of the power spectral density; and calculating the ratio between the first power value and the second power value to obtain the in-band / out-of-band power ratio.

[0077] The in-band / out-of-band power ratio (PRIO) is another signal characteristic parameter that can be considered. The PRIO is the ratio between the power value within the signal's bandwidth and the power value within the same bandwidth outside the signal's bandwidth, calculated based on the signal's power spectral density distribution. The PRIO is used as a threshold when the target signal is superimposed with channel noise. Thr At the end of the ICA-R algorithm, the PRIO value of the extracted signal is calculated (labeled as PRIOy), and the PRIO is compared. Thr The relationship between PRIOy and PRIOy: if PRIOy is greater than PRIO... Thr If this happens, the ICA-R algorithm's optimization process is restarted until PRIOy is not less than PRIO. Thr Or the number of re-optimization attempts m exceeds the set maximum value m. max .

[0078] S14: Extract the estimated target signal from the input signal based on the ICA-R algorithm and the reference signal;

[0079] In this step, the estimated target signal is extracted from the input signal using the ICA-R algorithm and the previously generated reference signal. The ICA-R algorithm, by assuming signal independence, enables blind signal separation under conditions of unknown signal sources and mixing patterns. This process aims to maximize the extraction of the target signal while minimizing the impact of interference and noise; the specific process can be found in [reference needed]. Figure 4 .

[0080] S15: Extract the feature parameters from the estimated signal, and determine whether the estimated signal meets the preset requirements based on the feature parameters and the feature parameter threshold; if not, return to S14 until the estimated signal meets the preset requirements;

[0081] S16: End.

[0082] Feature parameters are extracted from the estimated signal and compared with set feature parameter thresholds to determine whether the estimated signal meets preset requirements. This step is the quality control stage of signal extraction. By analyzing the feature parameters, it is ensured that the quality of the extracted signal meets the actual application requirements, thereby achieving effective signal verification.

[0083] If the estimated signal does not meet the preset requirements, the process returns to step S14 to extract the target signal again. This feedback mechanism ensures the flexibility and reliability of the signal extraction process. Through multiple iterations and optimizations, the extraction effect of the target signal can be continuously improved until the set quality standards are finally met. This process not only improves the accuracy of signal extraction but also guarantees the application effect in complex electromagnetic environments.

[0084] As a preferred embodiment, after determining that the preset requirements are not met, the method further includes: determining whether the number of times the step of extracting the estimated signal of the target signal from the input signal according to the ICA-R algorithm and the reference signal is performed is greater than the set maximum number of optimizations; if it is greater than the set maximum number of optimizations, the estimated signal is determined to be the target signal; if it is not greater than the set maximum number of optimizations, the step of extracting the estimated signal of the target signal from the input signal according to the ICA-R algorithm and the reference signal is re-entered.

[0085] The maximum number of optimization attempts is: Figure 5 m max In this preferred embodiment, after determining that the estimated signal does not meet the preset requirements, it is further determined whether the number of times the step of extracting the estimated signal of the target signal from the input signal according to the ICA-R algorithm and the reference signal has been performed exceeds the set maximum number of optimization attempts (m). max The purpose of this mechanism is to control processing time and resource usage while ensuring signal extraction effectiveness. If the number of extraction steps exceeds the set maximum number of optimization attempts, the current estimated signal will be directly identified as the target signal, even if it may not fully meet the preset requirements. This step aims to prevent excessive iterations from wasting resources and causing time delays, while ensuring that a relatively usable signal can still be obtained within a certain number of iterations. Conversely, if the maximum number of optimization attempts has not been exceeded, the extraction step S14 will be restarted to continue optimizing the estimated signal. In this way, a balance is achieved between effectively extracting the target signal and utilizing resources, ensuring a rapid response and providing usable signal extraction results in practical applications.

[0086] As a preferred embodiment, the ICA-R algorithm includes a built-in number of iterations. After determining that the number of times the step of extracting the estimated signal of the target signal from the input signal according to the ICA-R algorithm and the reference signal is performed is not greater than the set maximum number of optimizations, the algorithm further includes: resetting the built-in number of iterations of the ICA-R algorithm to 0 and re-entering the step of extracting the estimated signal of the target signal from the input signal according to the ICA-R algorithm and the reference signal.

[0087] Specifically, the ICA-R algorithm has a built-in iterative mechanism to enhance the accuracy and efficiency of signal extraction. After determining that the number of extraction steps has not exceeded the set maximum number of optimization attempts, the built-in iteration count (k) of the ICA-R algorithm is reset to 0, thus restarting the extraction process. Specifically, the ICA-R algorithm undergoes multiple iterations during execution, with each iteration incrementing the value of k to track the current iteration progress. When k reaches its maximum iteration count, the ICA-R algorithm terminates its execution. This design allows the system to run the extraction process from scratch, helping to improve the accuracy of signal extraction and avoiding insufficient signal extraction due to insufficient iterations. The mechanism of resetting the iteration count ensures that the algorithm can fully explore possible signal features each time it re-extracts, thereby optimizing the extraction results. This process is similar to the flow of the ICA-R algorithm in the prior art, and will not be described in detail here, ensuring effective extraction of the target signal in complex electromagnetic environments.

[0088] Furthermore, to better explain Figure 1 How does the blind target signal extraction module based on the ICA-R algorithm use the reference signal r(t) to separate and extract the target signal s(t)? Please refer to [link to documentation]. Figure 3 and Figure 4 ,in, Figure 3 This invention provides a block diagram of a blind target signal extraction system based on the ICA-R algorithm, where s(t) is the target signal, j(t) is the interference signal in the input signal, x1(t) and x2(t) are the two antenna signals, n1(t) and n2(t) are the two noise signals superimposed at the receiver, r(t) is the reference signal, y(t) is the extracted signal, b(n) is the source information bits, b'(n) is the information bits recovered at the receiver, and A represents the channel mixing matrix. Of course, the representation of the wireless channel is not limited to using a matrix; other implementation methods can also be used, and this application does not impose any particular limitations.

[0089] The 2x2 full-rank matrix in this embodiment can complete the signal transmission, and the implementation method is simple and reliable.

[0090] This section provides a general description of the principles of the ICA-R (Independent Component Analysis with Reference) algorithm:

[0091] The ICA-R-based blind signal extraction method effectively addresses the shortcomings of ICA-based blind source separation methods. By introducing a reference signal containing prior information about the target signal, the ICA-R-based blind source separation method can accurately separate and extract a unique target signal. The key to the success of this method lies in selecting a suitable reference signal and properly initializing the separation vector. The reference signal must be a non-Gaussian signal containing prior information about the target signal. A review of existing literature reveals that methods for generating reference signals mainly include: based on the signal's sign function, based on the statistical characteristics of the target signal, and based on signal frequency. Initialization methods for the separation vector mainly include random initialization methods and non-random initialization methods assisted by prior information.

[0092] Please refer to Figure 4 , Figure 4 This is a schematic diagram of an ICA-R algorithm processing module provided by the present invention. Its main idea is to introduce a proximity metric between the output signal y(t) and the reference signal r(t) into the difference function of the traditional ICA algorithm. Then, the separation vector is learned through a certain iterative loop. After the separation vector converges, it is multiplied by the observed signal x(t) to achieve the separation and extraction of the target signal s(t). The specific steps are as follows:

[0093] (1) Initialize Laplace parameters ( ), learning rate ( ), randomly generate separation vector ( ), and set convergence criteria;

[0094] (2) Choose a proximity measurement function ( The proximity metric is calculated by comparing the output signal y(t) with the reference signal r(t). The proximity metric reaches its minimum when the output signal y(t) is the desired communication signal s(t). It can be implemented in three different forms:

[0095] , (Formula 1);

[0096] in, It represents the mathematical expectation.

[0097] (3) Constructing inequality constraints ( ). The proximity between the output signal and the reference signal is controlled within a certain threshold. (), used to distinguish target signals from other communication signals, its expression form is: , (Formula 2);

[0098] (4) Constructing equality constraints ( This constraint ensures that the extracted signal has unit variance, and its expression is: , (Formula 3);

[0099] (5) Construct the optimal objective function ( It can be based on the difference function of negative entropy, and its expression is: , (Formula 4);

[0100] in, It is a positive constant. It is any non-quadratic function, and v is a Gaussian variable with zero mean and unit variance.

[0101] (6) Using the augmented method, the optimization problem under the above constraints is transformed into one with Lagrange parameters ( The problem of finding the extrema of the augmented Lagrangian function is as follows: The extrema are found to be:

[0102] , (Formula 5);

[0103] In the formula and These are the Lagrange multipliers corresponding to the two constraints. For penalty parameters, It is an Euclidean norm.

[0104] (7) Solve equation (5) for the following... The extreme values ​​are used to obtain the separation vector. The iterative learning formula is: , (Formula 6);

[0105] in, and L pairs respectively First and second partial derivatives, R zz Let z be the covariance matrix of the whitened observation matrix. The learning rate. Simultaneously, the Lagrangian parameters. and The iterative update formula is:

[0106] , (Formula 7);

[0107] (8) Iterate through formulas 6 and 7, and determine the result. If the convergence criterion is met, stop the loop calculation and save the final separation vector. Otherwise, continue iterative calculation.

[0108] (9) Through the formula The calculation directly yields the expected independent component y(t) that is closest to the reference signal r(t), which is the recovered sample of the target signal s(t), thus completing the separation and extraction of the target signal.

[0109] In order to ensure the stability and reliability of the signal during the calculation process, this embodiment first centers and whitens the N antenna signals before (1).

[0110] As can be seen, the above calculation method in this embodiment can extract the communication signal from the N antenna signals, and the implementation method is simple and reliable, thereby achieving anti-interference in the communication signal transmission process.

[0111] Specifically, this embodiment aims to provide a specific implementation method for improving the robustness of the ICA-R algorithm in extracting target signals by utilizing prior information of the signal. The prior information in this embodiment is not limited to the power spectral density symmetry coefficient (PSI) mentioned in the above implementation method. As long as it can reflect the uniqueness of the target signal, it is acceptable. No special limitation is made here.

[0112] Please refer to Figure 5 , Figure 5 This invention provides a flowchart of an algorithm for extracting a target signal using prior information of the target signal. The portion of the flowchart without the gray background represents the conventional ICA-R algorithm; the gray background represents the innovative improvements of this invention. Specifically, based on the conventional ICA-R algorithm, an estimated value y(t) of the extracted signal is obtained. Then, the characteristic parameter values ​​of this signal are calculated and compared with a characteristic parameter threshold of the target signal s(t) under a set signal-to-noise ratio condition. If the value is greater than the threshold, it indicates that the signal estimated by the algorithm in this instance is not the target signal. The learning step k of the ICA-R algorithm's single optimization needs to be reset to 0, allowing the algorithm to restart with the optimal separation vector. The search task continues until the value of the characteristic parameters of the estimated signal y(t) is not less than or equal to a set threshold, or the number of algorithm restarts m exceeds the set maximum value m. Max .

[0113] Please refer to Figure 6 , Figure 6 This invention provides a graph showing the change in target signal extraction accuracy with the number of algorithm reruns when setting a threshold using PSI as a feature parameter. Figure 6 The separation vector ω corresponding to the diagram on the left p Initialization is performed using a random generation method. Figure 6 The separation vector ω corresponding to the diagram on the right p Initialization is performed using a non-random generation method.

[0114] get Figure 6The simulation parameters for the experimental results shown are set as follows.

[0115] The target signal is set as a BPSK modulated digital bandpass communication signal with a carrier frequency of f. c1 =40kHz, signal bandwidth is 16kHz, up / down sampling factor is 16. To avoid inter-symbol crosstalk, a root-raised cosine filter with a roll-off factor of 1 is used for shaping filtering before up-conversion, and the same filter is used for matched filtering at the receiver. The interference signal is set as a Gaussian narrowband noise signal with a center frequency of f. c2 =36kHz, signal bandwidth is 4kHz. The power ratio of the two signals, i.e., the signal-to-interference ratio, is set to -30dB. It is obvious that the spectrum of the interfering signal is within the operating bandwidth of the communication signal, and the power is much greater than that of the communication signal, so this is a typical partial bandgap jamming scenario.

[0116] The channel mixing matrix is ​​set to A = [0.64598 0.93228; 0.93647 0.1598], the receiver sampling rate is set to 160kHz, and the signal after channel noise addition is E. b N0 reaches 7dB, which is the demodulation threshold for speech signals transmitted over an additive white Gaussian noise (AWGN) channel. The reference signal is set as a cosine pulse function with the same carrier frequency as the target signal, i.e., r(t) = cos(2πf). c t), which is easily generated in actual communication systems.

[0117] Since digitally modulated communication signals are all super-Gaussian signals, a non-quadratic function is chosen as... The proximity metric uses the mean squared error function, i.e. The feature parameter threshold is set to The Lagrange multipliers and penalty factor are initialized to... , The learning rate for the iterative calculation of the separation vector is set to... .

[0118] During the iteration process, the following formula is used to determine whether the separation vector has reached the convergence criterion: In simulation Set to 10 -3 .

[0119] Finally, the robustness of the algorithm in separating and extracting the target signal is measured by the percentage of times the source information bit sequence is effectively transmitted in N transmission trials (accuracy). The accuracy is defined as follows:

[0120] ;

[0121] in , and These represent the total number of tests, the number of times the bit error rate did not exceed the theoretical value, and the number of times the bit error rate exceeded the theoretical value, respectively. The value range is between 0 and 1. The larger the value, the better the stability of the ICA-R algorithm. According to the theoretical bit error rate calculation formula for BPSK modulated signals transmitted over a Gaussian white noise channel, E... b The bit error rate at N0=7dB is 7.73×10⁻⁶. -4 This means that the number of error bits transmitted in 10,000 information bits should not exceed 8. Therefore, in the simulation experiment, we use the number of error bits transmitted in 10,000 information bits not exceeding 8 as the statistical threshold for the bit error rate not exceeding the theoretical value. The value is 50000.

[0122] The threshold for the feature parameter PSI is set to 0.26.

[0123] like Figure 6 It can be seen that when setting the threshold using the power spectral density estimation symmetry coefficient (PSI) as a characteristic parameter, regardless of the separation vector ω p Regardless of the initialization method used, the accuracy of the ICA-R algorithm in extracting the target communication signal and the number of algorithm reruns both exhibit an S-shaped exponential function relationship. Furthermore, as... Figure 6 As shown in the diagram on the left, the separation vector ω p When initializing using a random generation method, setting the maximum number of algorithm reruns to 20 or more can improve the accuracy from 15.11% per run to over 90%, an improvement of more than 75 percentage points. The separating vector ω... p When initializing using a non-random generation method, if the maximum number of times the algorithm is rerun is set to more than 16, the accuracy can be improved from 95.76% in a single run to more than 99%, an improvement of more than 3 percentage points.

[0124] Figure 6 The experimental results show that the signal extraction method based on the ICA-R algorithm provided by this invention can improve the robustness of the algorithm.

[0125] To solve the above technical problems, such as Figure 7 As shown, the present invention provides a signal extraction device based on the ICA-R algorithm, comprising:

[0126] Memory 71 is used to store computer programs;

[0127] The processor 72 is used to implement the steps of the signal extraction method based on the ICA-R algorithm described above when executing a computer program. For a description of the signal extraction device based on the ICA-R algorithm, please refer to the above embodiments; further details will not be repeated here.

[0128] To address the aforementioned technical problems, this invention provides a non-volatile storage medium storing a computer program. When executed by a processor, the computer program implements the steps of the signal extraction method based on the ICA-R algorithm described above. For a description of the non-volatile storage medium, please refer to the above embodiments; further details are omitted here.

[0129] It should also be noted that, in this specification, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0130] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A signal extraction method based on the ICA-R algorithm, characterized in that, include: Receive input signals, the input signals including target signals and interference signals; the input signals are communication signals. A reference signal is generated based on the carrier frequency of the target signal, wherein the reference signal is a cosine pulse signal with the same carrier frequency as the target signal; Calculate feature parameters based on prior information of the target signal, and set threshold values ​​for the feature parameters; The estimated signal of the target signal is extracted from the input signal according to the ICA-R algorithm and the reference signal; Extract the feature parameters from the estimated signal, and determine whether the estimated signal meets the preset requirements based on the feature parameters and the feature parameter threshold. If the condition is not met, the process will re-enter the step of extracting the estimated signal of the target signal from the input signal according to the ICA-R algorithm and the reference signal, until the estimated signal meets the preset requirements. Calculate feature parameters based on prior information of the target signal, including: Perform a Fourier transform on the target signal to obtain a frequency domain signal; Calculate the power spectral density of the frequency domain signal; Calculate the power spectral density symmetry coefficients based on the power spectral density; Calculate feature parameters based on prior information of the target signal, including: Perform a Fourier transform on the target signal to obtain a frequency domain signal; Calculate the power spectral density of the frequency domain signal; Calculate the first power value within the signal bandwidth based on the distribution of the power spectral density; Calculate the second power value within the same bandwidth outside the signal bandwidth based on the distribution of the power spectral density; Calculate the ratio between the first power value and the second power value to obtain the in-band and out-of-band power ratio.

2. The signal extraction method based on the ICA-R algorithm as described in claim 1, characterized in that, The expression for the power spectral density symmetry coefficient is: Wherein, PSI is the power spectral density symmetry coefficient, M is the number of sampling points in the power spectral density that deviate from the center frequency, k represents the k-th deviation sampling point, and Y... L (k) represents the power spectral density value of the k-th sampling point in the portion of the target signal where the power spectral density is less than the center frequency, Y R (k) represents the power spectral density value of the k-th sampling point in the portion of the target signal where the power spectral density is greater than the center frequency. (k) is Y R (k) horizontal flip, Represents the modulus of a vector.

3. The signal extraction method based on the ICA-R algorithm as described in claim 1, characterized in that, The expression for the power spectral density symmetry coefficient is: Wherein, PSI is the power spectral density symmetry coefficient, M is the number of sampling points in the power spectral density that deviate from the center frequency, k represents the k-th deviation sampling point, and Y... L (k) represents the power spectral density value of the k-th sampling point in the portion of the target signal where the power spectral density is less than the center frequency, Y R (k) represents the power spectral density value of the k-th sampling point in the portion of the target signal where the power spectral density is greater than the center frequency. (k) is Y R (k) horizontal flip, The modulus of a vector. This represents the second-order normal form.

4. The signal extraction method based on the ICA-R algorithm as described in claim 1, characterized in that, The characteristic parameter is a power spectral density symmetry coefficient. Determining whether the estimated signal meets preset requirements based on the characteristic parameter and the characteristic parameter threshold includes: Determine whether the power spectral density coefficient in the estimated signal is greater than the power spectral density coefficient threshold; If the power spectral density coefficient is greater than the threshold value, then the estimated signal is determined to meet the preset requirement. If the power spectral density coefficient is not greater than the threshold value, then the estimated signal is determined not to meet the preset requirements.

5. The signal extraction method based on the ICA-R algorithm as described in any one of claims 1-4, characterized in that, After determining that the preset requirements are not met, the process also includes: Determine whether the number of times the step of extracting the estimated signal of the target signal from the input signal according to the ICA-R algorithm and the reference signal is performed is greater than the set maximum number of optimizations; If the number of optimization attempts exceeds the set maximum number, then the estimated signal is determined to be the target signal. If the number of optimization attempts is not greater than the set maximum number of attempts, then the process re-enters the step of extracting the estimated signal of the target signal from the input signal according to the ICA-R algorithm and the reference signal.

6. The signal extraction method based on the ICA-R algorithm as described in claim 5, characterized in that, The ICA-R algorithm includes a built-in number of iterations. After determining that the number of times the step of extracting the estimated signal of the target signal from the input signal according to the ICA-R algorithm and the reference signal is performed is not greater than the set maximum number of optimizations, it further includes: The built-in iteration count of the ICA-R algorithm is reset to 0, and the process re-enters the step of extracting the estimated signal of the target signal from the input signal according to the ICA-R algorithm and the reference signal.

7. A signal extraction device based on the ICA-R algorithm, characterized in that, include: Memory, used to store computer programs; A processor, configured to, when executing a computer program, implement the steps of the signal extraction method based on the ICA-R algorithm as described in any one of claims 1-6.

8. A non-volatile storage medium, characterized in that, The non-volatile storage medium stores a computer program, which, when executed by a processor, implements the steps of the signal extraction method based on the ICA-R algorithm as described in any one of claims 1-6.

Images