Signal acquisition method and device based on blind source separation in multipath scenario, electronic equipment and storage medium

Abstract

The application provides a signal acquisition method and device based on blind source separation in a multipath scene, electronic equipment and a storage medium, and relates to the technical field of communication. The method comprises the following steps: constructing an observation vector matrix of a plurality of subcarriers, and performing blind source separation to obtain a separation signal matrix and a mixing matrix of each subcarrier; reordering each separation signal of the separation signal matrix based on the mixing matrix to obtain a target frequency domain component; correlating the target frequency domain component of each subcarrier with a local frequency domain ZC sequence to determine an effective time delay window, and taking the correlation peak position in the effective time delay window as a signal acquisition position, thereby improving the anti-interference capability of the signal acquisition based on blind source separation in the multipath scene. The application fundamentally adapts to the actual situation that the mixing characteristics of each subcarrier in the frequency selective multipath signal are different, avoids the arrangement ambiguity problem of the separation signal, and further improves the anti-interference capability of the signal acquisition based on blind source separation in the multipath scene.

Description

Technical Field

[0001] This invention relates to the field of communication technology, and in particular to a signal acquisition method, apparatus, electronic device, and storage medium based on blind source separation in multipath scenarios. Background Technology

[0002] In systems such as wireless communication, radar, and navigation, receivers need to detect desired signals and determine their arrival time against noise and interference backgrounds; this process is called signal acquisition or signal detection. In multi-antenna receiving systems, the signals received by multiple antennas are mixed versions of the source signal after propagation through different spatial channels, equivalent to different weighted combinations of the source signal in the spatial dimension. Therefore, the spatial information of multiple antennas can be used to improve reception performance. With the widespread application of broadband multicarrier systems such as OFDM and the increasing complexity of multipath propagation environments, it is necessary to utilize the spatial information of multiple antennas to achieve reliable acquisition of desired signals in frequency-selective channels.

[0003] Existing signal acquisition methods in multipath scenarios involve down-converting and analog-to-digital converting the received signals from each antenna element to obtain multiple digital baseband signals. Channel estimation is used to obtain the spatial steering vector or channel coefficients of the desired signal. Weighting coefficients for each element are calculated based on criteria such as minimum variance distortionless response and maximum signal-to-noise ratio. The signals from each element are weighted and summed to synthesize a single signal, which is equivalent to forming a main lobe pointing in the direction of the desired signal in space and suppressing the signal in the direction of interference. Finally, the synthesized single signal is subjected to sliding correlation with the local ZC sequence to detect the correlation peak and achieve acquisition.

[0004] In a multipath propagation environment, the received signal is a superposition of multiple replicas arriving at the receiving array via different paths, with different time delays and angles of arrival. Multipath propagation in broadband OFDM systems leads to frequency selectivity in the channel, meaning that the channel frequency response differs on different subcarriers. Since the spacing between antenna elements is fixed, electromagnetic waves of different frequencies correspond to different phase differences. Even at the same physical angle of arrival, the array steering vectors on different subcarriers are different. Using a fixed set of beam weights cannot form the correct beam pointing for all subcarriers, resulting in beam pointing mismatch and null shift on most subcarriers, a sharp drop in the output signal-to-noise ratio, and poor anti-interference capability of signal acquisition. Summary of the Invention

[0005] This invention provides a signal acquisition method, apparatus, electronic device, and storage medium based on blind source separation in multipath scenarios, in order to solve the defect of poor anti-interference capability of signal acquisition in multipath scenarios in the prior art, and to improve the anti-interference capability of signal acquisition in multipath scenarios.

[0006] This invention provides a signal acquisition method based on blind source separation in multipath scenarios, comprising: Based on multiple received signals from multiple receiving antennas, an observation vector matrix for multiple subcarriers is constructed. Blind source separation is performed on the observation vector matrix of each subcarrier to obtain the separated signal matrix and the mixing matrix of each subcarrier; The subcarrier-based hybrid matrix is ​​used to reorder the individual separated signals of the subcarrier's separated signal matrix to obtain the target frequency domain component of the subcarrier. The target frequency domain component of each subcarrier is multiplied by the ZC sequence in the local frequency domain. The inverse Fourier transform of the multiplication result is then performed to obtain the correlation result in the time domain. The correlation result in the time domain is then normalized to obtain the normalized result in the time domain. The effective delay window is determined based on the time point when the normalized result in the time domain first exceeds a preset threshold. The position of the relevant peak within the effective delay window is used as the signal capture position.

[0007] According to the signal acquisition method based on blind source separation in multipath scenarios provided by the present invention, blind source separation is performed on the observation vector matrix of each subcarrier to obtain the separated signal matrix and the mixing matrix of each subcarrier, including: Calculate the sample covariance matrix of the observation vector matrix for each subcarrier and perform eigenvalue decomposition; Based on the eigenvalue decomposition results of each subcarrier, a whitening matrix for each subcarrier is constructed; the eigenvalue decomposition results include the decomposed eigenvalues ​​and their eigenvectors. Based on the eigenvalue decomposition results of each subcarrier and the observation vector matrix of each subcarrier, the whitening data of each subcarrier is calculated. Calculate the separation matrix of each subcarrier based on the whitened data of each subcarrier; Based on the separation matrix and whitening data of each subcarrier, calculate the separation signal matrix of each subcarrier; Based on the separation matrix and whitening matrix of each subcarrier, the mixing matrix of each subcarrier is calculated.

[0008] According to the signal acquisition method based on blind source separation in multipath scenarios provided by the present invention, the subcarrier separation matrix is ​​calculated in the following manner: S1. Use the multiple initial unit norm vectors of the initial separation matrix of the subcarrier as multiple unit norm vectors for the current iteration; S2. Based on the unit norm vector of each current iteration and the whitened data of the subcarrier, calculate the unit norm vector of each optimization. S3. Perform order reduction orthogonalization and normalization operations on each optimized unit norm vector to obtain each updated unit norm vector. S4. Take the multiple updated unit norm vectors as the multiple unit norm vectors of the current iteration, and repeat steps S2-S4 to iteratively update the unit norm vectors until the difference between the unit norm vectors of two adjacent iterations tends to stabilize, or until the preset number of iterations is reached; construct the subcarrier separation matrix based on the multiple last updated unit norm vectors.

[0009] The signal acquisition method based on blind source separation in a multipath scenario provided by the present invention reorders the separated signals of the subcarrier's separated signal matrix based on the subcarrier's mixing matrix to obtain the target frequency domain component of the subcarrier, including: Calculate the norm of each column of the mixing matrix of the subcarriers; the norm of each column is proportional to the source power of each column; The separated signals in each column of the separated signal matrix are reordered based on the norm of each column; Based on all the reordered separated signals, the target frequency domain component of the subcarrier is determined.

[0010] The signal acquisition method based on blind source separation in a multipath scenario provided by the present invention determines the effective delay window based on the time point when the normalized result in the time domain first exceeds a preset threshold, including: The point in time when the normalized result first exceeds the preset threshold is taken as the starting point of the effective delay window; The effective delay window is determined based on the starting point and the preset delay window duration.

[0011] According to the signal acquisition method based on blind source separation in multipath scenarios provided by the present invention, the position of the correlation peak within the effective time delay window is determined in the following manner: Within the effective time delay window, the position of the maximum value of the normalized result is taken as the position of the correlation peak.

[0012] The signal acquisition method based on blind source separation in multipath scenarios provided by the present invention receives a signal comprising N time-domain sampling points, where N≥2. Based on multiple received signals from multiple receiving antennas, an observation vector matrix of multiple subcarriers is constructed, including: Perform a fast Fourier transform on the N time-domain sampling points of each receiving antenna to output the N frequency-domain points of each receiving antenna; In the frequency domain points of all receiving antennas, the frequency domain points of consecutive OFDM symbols are organized into a matrix according to the subcarrier index, resulting in an observation vector matrix of multiple subcarriers, which is greater than the number of receiving antennas.

[0013] The present invention also provides a signal acquisition device based on blind source separation in multipath scenarios, comprising: The observation vector matrix construction module is used to construct an observation vector matrix for multiple subcarriers based on multiple received signals from multiple receiving antennas. The blind source separation module is used to perform blind source separation on the observation vector matrix of each subcarrier to obtain the separated signal matrix and the mixing matrix of each subcarrier; based on the mixing matrix of the subcarrier, the separated signals of the separated signal matrix of the subcarrier are reordered to obtain the target frequency domain component of the subcarrier. The acquisition result determination module is used to perform dot product between the target frequency domain component of each subcarrier and the ZC sequence of the local frequency domain, perform inverse Fourier transform on the dot product result to obtain the correlation result in the time domain, normalize the correlation result in the time domain to obtain the normalized result in the time domain, determine the effective delay window based on the time point when the normalized result in the time domain first exceeds a preset threshold, and use the correlation peak position within the effective delay window as the signal acquisition position.

[0014] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements a signal acquisition method based on blind source separation in any of the multipath scenarios described above.

[0015] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a signal acquisition method based on blind source separation in any of the multipath scenarios described above.

[0016] This invention provides a signal acquisition method, apparatus, electronic device, and storage medium based on blind source separation in multipath scenarios. It constructs observation vector matrices for multiple subcarriers based on multiple received signals from multiple receiving antennas; performs blind source separation on the observation vector matrices of each subcarrier to obtain the separated signal matrix and mixing matrix of each subcarrier; reorders the separated signals of the separated signal matrix of the subcarrier based on the mixing matrix of the subcarrier to obtain the target frequency domain component of the subcarrier; performs a dot product between the target frequency domain component of each subcarrier and the ZC sequence of the local frequency domain, and performs an inverse Fourier transform on the dot product result to obtain the correlation result in the time domain; normalizes the correlation result in the time domain to obtain the normalized result in the time domain; determines the effective delay window based on the time point when the normalized result in the time domain first exceeds a preset threshold; and uses the correlation peak position within the effective delay window as the signal acquisition position, thereby improving the anti-interference capability of signal acquisition based on blind source separation in multipath scenarios. This invention improves the anti-interference capability of signal acquisition based on blind source separation in multipath scenarios by independently separating the observation vector matrices of each subcarrier, thus adapting to the actual situation where the mixing characteristics of each subcarrier are different in frequency-selective multipath signals. Furthermore, this invention reorders the separated signals in the separated signal matrix according to the mixing matrix, avoiding the problem of ambiguous arrangement of separated signals, thereby improving the anti-interference capability of signal acquisition based on blind source separation in multipath scenarios. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0018] Figure 1 This is one of the flowcharts illustrating the signal acquisition method based on blind source separation in multipath scenarios provided by the present invention.

[0019] Figure 2 This is the second flowchart of the signal acquisition method based on blind source separation in multipath scenarios provided by the present invention.

[0020] Figure 3 This is a flowchart illustrating the process of calculating the subcarrier separation matrix provided by the present invention.

[0021] Figure 4 This is a schematic diagram of the effective delay window provided by the present invention.

[0022] Figure 5 These are the ROC curves for three methods with a signal-to-dryness ratio of 10 dB in the simulation experiment.

[0023] Figure 6 These are the ROC curves for three methods with a signal-to-dryness ratio of 20dB in the simulation experiment.

[0024] Figure 7 This is the normalized amplitude distribution of the correlation peaks after blind source separation in this invention, with a signal-to-interference ratio of 20dB, in a simulation experiment.

[0025] Figure 8 This is the normalized amplitude distribution of the correlation peak of a single antenna with an interference-to-signal ratio of 20dB in the simulation experiment.

[0026] Figure 9 This is a schematic diagram of the signal acquisition device based on blind source separation in a multipath scenario provided by the present invention.

[0027] Figure 10 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0028] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0029] The following is combined Figures 1-10 This invention describes a signal acquisition method, apparatus, electronic device, and storage medium based on blind source separation in multipath scenarios.

[0030] Figure 1 This is one of the flowcharts illustrating the signal acquisition method based on blind source separation in multipath scenarios provided by this invention, such as... Figure 1 As shown, the signal acquisition method based on blind source separation in multipath scenarios includes steps S100 to S600, and the specific steps are as follows.

[0031] S100: Based on multiple received signals from multiple receiving antennas, construct an observation vector matrix for multiple subcarriers.

[0032] The present invention has M receiving antennas, M≥2, and the desired signal contains a local ZC sequence of length N, which is mapped to N subcarriers of an Orthogonal Frequency-Division Multiplexing (OFDM) symbol.

[0033] Each received signal is acquired from each receiving antenna. The received signal consists of multiple time-domain sampling points. A Fast Fourier Transform (FFT) is performed on the received signal to convert the multiple time-domain sampling points into multiple frequency-domain data. Each frequency-domain data is treated as an observation vector. The frequency-domain data from all receiving antennas are organized by subcarrier and OFDM symbol to construct an observation vector matrix for multiple subcarriers.

[0034] S200: Perform blind source separation on the observation vector matrix of each subcarrier to obtain the separated signal matrix and the mixing matrix of each subcarrier.

[0035] Blind source separation refers to the technique of recovering the original signals from multiple mixed signals by utilizing only the statistical properties of the observed signals (observation vector matrix) themselves, in the case of unknown channel and source signals.

[0036] For each subcarrier, perform Fast Independent Component Analysis (FastICA) on the observation vector matrix to estimate the separation matrix. Then, perform blind source separation on the observation vector matrix of each subcarrier to obtain the separated signal matrix and the mixing matrix of each subcarrier. The inverse of the mixing matrix is ​​the separation matrix.

[0037] S300: Based on the subcarrier hybrid matrix, the individual separated signals of the subcarrier separated signal matrix are reordered to obtain the target frequency domain component of the subcarrier.

[0038] Blind source separation is solved independently for each subcarrier, and the separated signals in the separation signal matrix of each subcarrier exhibit arrangement ambiguity. To eliminate this ambiguity, the separated signals in the separation signal matrix need to be reordered. Based on the reordered separated signals of the subcarriers, the target frequency domain components of the subcarriers are obtained.

[0039] S400: Perform a dot product between the target frequency domain component of each subcarrier and the ZC sequence in the local frequency domain, perform an inverse Fourier transform on the dot product result to obtain the correlation result in the time domain; normalize the correlation result in the time domain to obtain the normalized result.

[0040] S500: Determine the effective delay window based on the time point when the normalized result in the time domain first exceeds the preset threshold.

[0041] S600: Use the position of the relevant peak within the effective delay window as the signal capture position.

[0042] A local ZC sequence refers to a ZC reference sequence with identical parameters to that of the transmitter, which is pre-stored or locally generated at the receiver and used as a reference copy for related operations.

[0043] The target frequency domain component of each subcarrier and the frequency domain conjugate of the local ZC sequence are multiplied point by point; the results of each point-by-point multiplication are subjected to inverse fast Fourier transform at each frequency domain point and the modulus is taken to obtain the correlation results in each time domain; the correlation results in the time domain are normalized to obtain the normalized results.

[0044] The target frequency domain component and the frequency domain conjugate of the local ZC sequence are multiplied point-by-point. The result of the point-by-point multiplication is then subjected to an inverse fast Fourier transform (IFFT) at each frequency domain point, and the modulus is taken to obtain the correlation results in each time domain. The formula for calculating the correlation results in the time domain is as follows.

[0045] ; in, For the target frequency domain components, This is the frequency domain conjugate vector of the local ZC sequence. For time domain The relevant results This involves multiplying each point individually.

[0046] The correlation results in the time domain are normalized to obtain the normalized results in the time domain. The formula for calculating the normalized results is as follows.

[0047] ; in, For time domain The relevant results For time domain The normalized result.

[0048] The effective delay window is determined based on the time point when the normalized result in the time domain first exceeds a preset threshold. For example, the time point when the normalized result in the time domain first exceeds the preset threshold can be used as the starting point or intermediate point of the effective delay window. The effective delay window is then determined based on this starting point or intermediate point and the preset delay window.

[0049] If the time-domain normalized result exceeds a preset threshold, signal acquisition is considered successful. The effective delay window represents a coarse location for signal acquisition. The location of the correlation peak within the effective delay window is used as the precise signal acquisition location.

[0050] This invention organically combines spatial diversity of multiple receiving antennas with OFDM frequency domain multi-subcarrier structure. Through a complete processing chain of spatial separation, eigenvalue sorting and reconstruction, and time domain correlation, it achieves robust acquisition without channel information under frequency-selective multipath and strong interference conditions.

[0051] If the normalization results in the time domain are all less than the preset threshold, then the signal acquisition is deemed unsuccessful.

[0052] The signal acquisition method based on blind source separation in multipath scenarios provided by this invention constructs observation vector matrices for multiple subcarriers based on multiple received signals from multiple receiving antennas; performs blind source separation on the observation vector matrices of each subcarrier to obtain the separated signal matrix and mixing matrix of each subcarrier; reorders the separated signals of the separated signal matrix of the subcarrier based on the mixing matrix of the subcarrier to obtain the target frequency domain component of the subcarrier; performs a dot product of the target frequency domain component of each subcarrier with the ZC sequence of the local frequency domain, and performs an inverse Fourier transform on the dot product result to obtain the correlation result in the time domain; normalizes the correlation result in the time domain to obtain the normalized result in the time domain; determines the effective delay window based on the time point when the normalized result in the time domain first exceeds a preset threshold; and uses the correlation peak position within the effective delay window as the signal acquisition position, thereby improving the anti-interference capability of signal acquisition based on blind source separation in multipath scenarios. This invention, by independently separating the observation vector matrix of each subcarrier, fundamentally adapts to the actual situation where the mixing characteristics of each subcarrier in frequency-selective multipath signals are different, thus improving the anti-interference capability of signal acquisition based on blind source separation in multipath scenarios. This invention reorders the individual signals of the separation signal matrix according to the mixing matrix, avoiding the problem of ambiguous arrangement of the separation signals, thereby improving the anti-interference capability of signal acquisition based on blind source separation in multipath scenarios.

[0053] Based on the above embodiments, the position of the correlation peak within the effective delay window is determined as follows: within the effective delay window, the position of the maximum value of the normalized result is taken as the position of the correlation peak.

[0054] The formula for calculating the position of the relevant peak is as follows.

[0055] ; in, For time domain The normalization result, This indicates the location of the relevant peak.

[0056] This invention determines the position of the correlation peak based on the position of the maximum value of the normalization result, which can pinpoint the signal acquisition position to a specific time delay point and improve the accuracy of signal acquisition.

[0057] In one embodiment, the position of the relevant peak within the effective delay window can also be the position where the normalized result within the effective delay window is greater than a preset threshold.

[0058] Based on the above embodiments, the received signal includes N time-domain sampling points, where N≥2. Based on the multiple received signals from multiple receiving antennas, an observation vector matrix of multiple subcarriers is constructed, including the following steps: Perform a fast Fourier transform on the N time-domain sampling points of each receiving antenna to output the N frequency-domain points of each receiving antenna; At all frequency domain points of the receiving antennas, there will be continuous The frequency domain points of each OFDM symbol are organized into a matrix according to the subcarrier index, resulting in an observation vector matrix for multiple subcarriers. The number is greater than the number of receiving antennas.

[0059] The wireless signal reaches the receiving antenna through multiple different paths, each with different time delays, amplitudes, and angles of arrival, resulting in the received signal being a superposition of multiple copies.

[0060] The received signal from each receiving antenna consists of N time-domain sampling points. A Fast Fourier Transform is performed on the N time-domain sampling points of each receiving antenna to output N frequency-domain points for each receiving antenna.

[0061] At all frequency domain points of the receiving antennas, there will be continuous The frequency domain points of each OFDM symbol are organized into a matrix according to the subcarrier index, resulting in an observation vector matrix for multiple subcarriers. Assume the number of receiving antennas (receiving channels) is... The dimension of the observation vector matrix of the subcarrier is... The expression for the observation vector matrix is ​​as follows.

[0062] ; in, For the first The observation vector matrix for each subcarrier is constructed, and the observation vector matrix is ​​generated for each subcarrier. For the first The subcarrier at the _ ... The first symbol on the The observation vector of the root receiving antenna (calculated based on the corresponding frequency domain points). For subcarrier index, , The total number of OFDM symbols. This represents the total number of receiving antennas.

[0063] This invention utilizes multiple OFDM symbols to provide statistical samples for each subcarrier. For each subcarrier, it uses data from... The observation vectors of different OFDM symbols are used as statistical samples for blind source separation. The channel variations between different OFDM symbols ensure that these samples have good statistical independence, satisfying the premise assumptions of the blind source separation algorithm and significantly improving the separation accuracy.

[0064] Based on the above embodiments, blind source separation is performed on the observation vector matrix of each subcarrier to obtain the separated signal matrix and mixing matrix of each subcarrier, including the following steps: Calculate the sample covariance matrix of the observation vector matrix for each subcarrier and perform eigenvalue decomposition; Based on the eigenvalue decomposition results of each subcarrier, a whitening matrix for each subcarrier is constructed; the eigenvalue decomposition results include the decomposed eigenvalues ​​and their eigenvectors. Based on the eigenvalue decomposition results of each subcarrier and the observation vector matrix of each subcarrier, the whitening data of each subcarrier is calculated. Calculate the separation matrix of each subcarrier based on the whitened data of each subcarrier; Based on the separation matrix and whitening data of each subcarrier, calculate the separation signal matrix of each subcarrier; Based on the separation matrix and whitening matrix of each subcarrier, the mixing matrix of each subcarrier is calculated.

[0065] The whitening matrix is ​​a linear transformation matrix constructed from the sample covariance matrix through eigenvalue decomposition. It is used to perform coordinate transformations on the observation vectors for decorrelation and variance normalization, while also achieving dimensionality reduction and noise resistance by preserving principal eigenvalues. The whitening matrix is ​​also used to reduce the unknown complex mixture matrix (observation vector matrix) to a unitary matrix, thereby limiting subsequent complex-valued blind source separation to orthogonal constraints.

[0066] Whitened data is a new signal obtained by transforming the observation vector matrix with a whitening matrix. Whitened data has eliminated the second-order correlation between subcarriers and completely normalized the second-order statistical information. Therefore, the subsequent complex FastICA algorithm only needs to search in the unitary space based on higher-order statistics (non-Gaussianity) to recover the independent sources.

[0067] Based on the observation vector matrix of each subcarrier and the total number of OFDM symbols, the sample covariance matrix of the observation vector matrix of each subcarrier is calculated, and eigenvalue decomposition is performed to obtain the eigenvalue decomposition results for each subcarrier. The eigenvalue decomposition results include the decomposed eigenvalues ​​and their eigenvectors. The expressions for the sample covariance matrix of the subcarrier and the eigenvalue decomposition results are as follows.

[0068] ; ; in, For the first The observation vector matrix of each subcarrier, For the first The sample covariance matrix of each subcarrier, Let be the eigenvector matrix, where for 1 eigenvector for 1 eigenvalue, It is the eigenvalue matrix.

[0069] Based on the eigenvalue decomposition results of each subcarrier, a whitening matrix for each subcarrier is constructed. The formula for calculating the whitening matrix is ​​as follows.

[0070] ; in, For the first The whitening matrix of each subcarrier.

[0071] Based on the eigenvalue decomposition results and the observation vector matrix of each subcarrier, the whitening data of each subcarrier is calculated. The formula for calculating the whitening data is as follows.

[0072] ; in, For the first Whitened data of each subcarrier, For the first The observation vector matrix of each subcarrier, For the first Whitening matrix of subcarriers, The total number of OFDM symbols. This represents the total number of receiving antennas.

[0073] Estimate the separation matrix of each subcarrier based on the whitening data of each subcarrier. Calculate the separated signal matrix of each subcarrier based on its separation matrix and whitening data. Calculate the mixing matrix of each subcarrier based on its separation matrix and whitening matrix. The formulas for calculating the separated signal matrix and mixing matrix are as follows.

[0074] ; in, For the first Separated signal matrix of subcarriers, For the first Separation matrix of subcarriers, For the first Whitened data of each subcarrier, For the first A hybrid matrix of subcarriers, For the first The whitening matrix of each subcarrier.

[0075] This invention estimates and separates the separation matrix independently on each subcarrier, instead of using a common separation matrix. This allows the separation matrix to vary with the subcarrier, fundamentally adapting to the actual situation where the mixing characteristics of each subcarrier are different in frequency-selective multipath channels, and improving the accuracy of blind source separation.

[0076] Based on the above embodiments, such as Figure 3 As shown, the subcarrier separation matrix is ​​calculated in the following way: S1. Use the initial unit norm vectors of the initial separation matrix of the subcarriers as the unit norm vectors of the current iteration; S2. Based on the unit norm vector of each current iteration and the whitened data of the subcarrier, calculate the unit norm vector of each optimization. S3. Perform order reduction orthogonalization and normalization operations on each optimized unit norm vector to obtain each updated unit norm vector. S4. Take the multiple updated unit norm vectors as the multiple unit norm vectors of the current iteration, and repeat steps S2-S4 to iteratively update the unit norm vectors until the difference between the unit norm vectors of two adjacent iterations tends to stabilize, or until the preset number of iterations is reached; construct the subcarrier separation matrix based on the multiple last updated unit norm vectors.

[0077] Multiple initial unit norm vectors are randomly set and used as unit norm vectors for multiple current iterations.

[0078] Based on the unit norm vectors of each current iteration and the whitened data of the subcarriers, complex FastICA iterative updates are performed with the objective of maximizing the peak non-Gaussianity, and the optimized unit norm vectors are calculated. The formulas for calculating the optimized unit norm vectors are as follows.

[0079] ; in, For the first An optimized unit norm vector, For the first An initial unit norm vector, For the first Whitened data of each subcarrier, This represents the total number of receiving antennas.

[0080] Perform deflation (orthogonalization) and normalization operations on each optimized unit norm vector to obtain updated unit norm vectors. Arrange all optimized unit norm vectors sequentially to form the optimized initial separation matrix.

[0081] In the optimized initial separation matrix, the optimized unit norm vector is used to subtract the projection onto the converged components of the optimized unit norm vector to obtain the reduced-order orthogonalized unit norm vector.

[0082] The reduced-order orthogonalized unit norm vector is determined based on the orthogonality result of the optimized unit norm vector and its predecessor.

[0083] Calculate the reduced-order orthogonalized unit norm vectors. Normalize each reduced-order orthogonalized unit norm vector to obtain the updated unit norm vectors. The formula for calculating the updated unit norm vectors is as follows.

[0084] ; in, In the optimized initial separation matrix, the first... The optimized unit norm vector is the previous optimized unit norm vector. For the first A reduced-order orthogonalized unit norm vector For the first An updated unit norm vector, For the first An optimized unit norm vector, For the first The projection onto the converged components of an optimized unit norm vector.

[0085] Multiple updated unit norm vectors are used as multiple unit norm vectors for the current iteration, and steps S2-S4 are repeated to iteratively update the unit norm vectors until the difference between the unit norm vectors of two adjacent iterations tends to stabilize, or until a preset number of iterations is reached (for example, the preset number of iterations is 200), at which point the iteration is terminated; a subcarrier separation matrix is ​​constructed based on multiple last updated unit norm vectors.

[0086] In one embodiment, the iteration termination condition is that the difference between the unit norm vectors of two adjacent iterations tends to stabilize. The unit norm vectors of two adjacent iterations include the unit norm vector of the current iteration and the unit norm vector of the previous iteration. If the difference between the conjugate product of the unit norm vectors of the current iteration and the unit norm vectors of the previous iteration and 1 is less than a difference threshold (e.g., a difference threshold of 10). -6 When the difference between the unit norm vectors of two adjacent iterations tends to stabilize, the specific calculation formula is as follows.

[0087] ; in, Let be the unit norm vector of the current iteration. It is the unit norm vector of the previous iteration.

[0088] This invention achieves preliminary optimization of the initial separation matrix using the unit norm vector and whitened data in the current iteration, and achieves accurate calculation of the separation matrix through order reduction orthogonalization operation, normalization operation and iterative update.

[0089] Based on the above embodiments, the separation signals of the separation signal matrix of the subcarrier are reordered based on the subcarrier's hybrid matrix to obtain the target frequency domain components of the subcarrier, including the following steps: Calculate the norm of each column of the mixing matrix of the subcarriers; the norm of each column is proportional to the source power of each column; The separated signals in each column of the separated signal matrix are reordered based on the norm of each column; Based on all the reordered separated signals, the target frequency domain component of the subcarrier is determined.

[0090] The norm of each column of the subcarrier mixing matrix (e.g., the first column). The subcarrier norm of a column ), reflecting the individual discrete signals of the subcarrier (e.g., the first ), The subcarrier Separated signals The equivalent spatial gain vector from the source to the receiving antenna.

[0091] The separated signals in each column of the separated signal matrix are reordered based on the norm of each column. For example, the separated signals in each column are reordered in descending order of the norm of each column. All reordered separated signals are then based on the subcarrier. Determine the target frequency domain component of the subcarrier. .

[0092] This invention relies on differences in source power and does not require assumptions about training sequences, pilots, or cross-subcarrier consistency. The resulting target frequency domain components retain only unknown complex scaling factors, which have no substantial impact on the positions of subsequent correlation peaks, thus eliminating the need for scaling correction.

[0093] In the blind source separation process, this invention decomposes the sample covariance matrix into eigenvalues ​​and sorts them uniformly from largest to smallest, establishing a correspondence between the row vectors of the separation matrix and the signal energy strength. Since the desired signal usually has high energy, the separation output corresponding to the largest eigenvalue corresponds to the desired signal on all subcarriers. The complete frequency domain sequence of the desired signal can be reconstructed by directly extracting the frequency domain value of this path. This invention's signal acquisition method based on blind source separation in multipath scenarios utilizes the inherent eigenvalue sorting mechanism of the separation algorithm, eliminating the need for channel identification on a subcarrier-by-subcarrier basis using a local ZC sequence, simplifying the processing flow and avoiding the problem of arrangement ambiguity.

[0094] Based on the above embodiments, the effective delay window is determined based on the time point when the normalized result in the time domain first exceeds a preset threshold, including the following steps: The point in time when the normalized result first exceeds the preset threshold is taken as the starting point of the effective delay window; The effective delay window is determined based on the starting point and the preset delay window duration.

[0095] like Figure 4 As shown, the time point when the normalized result first exceeds a preset threshold is taken as the starting point of the effective delay window. The preset delay window duration is the length of the effective delay window. Based on the starting point and the preset delay window duration, the effective delay window is determined.

[0096] This invention determines the effective delay window based on the time point when the normalized result in the time domain first exceeds a preset threshold. The position of the relevant peak within the effective delay window is used as the signal acquisition position. This improves the acquisition reliability, effectively suppresses false alarms and interference, and narrows the subsequent processing range, enabling fast, robust, and interference-resistant signal acquisition.

[0097] like Figure 2 As shown, the signal acquisition method based on blind source separation in multipath scenarios provided by the present invention includes the following steps.

[0098] Step 1: Acquire time-domain signals through the first analog-to-digital (AD) conversion channel, the second analog-to-digital (AD) conversion channel, ..., the Mth analog-to-digital (AD) conversion channel.

[0099] Step 2: Perform a Fast Fourier Transform on the N time-domain sampling points acquired by each analog-to-digital conversion channel to output the N frequency-domain points of each channel.

[0100] Step 3: At all frequency domain points, organize the frequency domain data of T consecutive OFDM symbols into a matrix according to the subcarrier index to obtain the observation vector matrix of each subcarrier. .

[0101] Step 4: Perform complex FastICA operations on each subcarrier to obtain the separate signal matrix and the mixing matrix of each subcarrier.

[0102] Step 5: Reorder the separated signals in each column of the separated signal matrix according to the norm of each column of the mixing matrix to eliminate permutation ambiguity and obtain the target frequency domain components of each subcarrier.

[0103] Step 6: Perform point-by-point multiplication of the target frequency domain component and the frequency domain conjugate of the local ZC sequence for each subcarrier.

[0104] Step 7: Perform an inverse fast Fourier transform on each frequency domain point of the result of each point-by-point multiplication and take the modulus to obtain the correlation results in the time domain. Normalize the correlation results in the time domain to obtain the normalized results in the time domain.

[0105] Step 8: Determine the effective delay window based on the time point when the normalized result in the time domain first exceeds the preset threshold.

[0106] Step 9: Use the position of the relevant peak within the effective delay window as the signal capture position.

[0107] To objectively verify the beneficial effects of the signal acquisition method based on blind source separation in multipath scenarios of the present invention, a complete Monte Carlo simulation platform was built for comparative verification.

[0108] Simulation scenario setup: Receiver configuration 1 uniform linear array antenna (with a spacing of 1 / 2) ); exists in the environment There are 1 independent information sources, where source 1 is the square root of the symbol. The ZC preamble signal, source 2 is co-channel interference; OFDM subcarrier number Statistical estimation of the number of symbols Each source of information The signal arrives via multiple paths, with each path's angle of arrival being independent, uniform, and random within the range of -60° to 60°. The time delay is random within the range of 0 to 4 sampling intervals. The interference-to-signal ratio (ISR) is 10 dB (the interference power is 10 times that of the useful signal). In the case of 20 dB, the useful signal's daily line signal-to-noise ratio (SNR) is 10 dB.

[0109] Comparison Method: Under identical channel and interference conditions, three methods were run simultaneously: the signal acquisition method based on blind source separation in the multipath scenario of this invention, single-antenna direct correlation, and equal-gain dual-antenna combined correlation. All three used the same correlation structure and decision statistics to ensure a fair comparison.

[0110] Two types of hypothetical data generation: Two sets of data are generated for each experiment—H1 hypothesis (useful signal + interference + noise) is used to calculate the detection rate, and H0 hypothesis (only interference + noise, useful signal missing) is used to calculate the false alarm rate.

[0111] Detection statistics and evaluation indicators: the magnitude of the normalized result in the time domain. As a detection statistic, a total of 2000 independent Monte Carlo trials were conducted. Each trial randomly generated a channel and interference, and the detection rate was obtained by scanning the decision threshold. False alarm rate With false alarm rate Plot the Receiver Operating Characteristic (ROC) curve and compare it with the curve that satisfies the following conditions. The interference-to-signal ratio of the capture capability. Simulation results are as follows: Figure 5 As shown, when the signal-to-interference ratio is 10dB, the three methods have the same anti-interference capability. Figure 6 As shown, when the signal-to-interference ratio is 20dB, the anti-interference acquisition capability of the signal acquisition method based on blind source separation in the multipath scenario of the present invention can still meet the requirements. Therefore, the signal acquisition method based on blind source separation in multipath scenarios of the present invention can improve the anti-interference capability by 10dB compared to the signal-to-interference ratio.

[0112] Based on the above simulation experiments, the signal acquisition method based on blind source separation in multipath scenarios of the present invention has the following beneficial effects.

[0113] (1) The signal acquisition method based on blind source separation in the multipath scenario of the present invention does not require angle estimation or special synchronization header design. In the simulation, none of the three methods used any channel prior information. The present invention only utilizes the statistical independence of the received data itself to complete interference separation and acquisition. Even in the single-path scenario, angle estimation is not required, which reduces the computational complexity and the link overhead of special synchronization header design.

[0114] (2) The signal acquisition method based on blind source separation in the multipath scenario of this invention is naturally adapted to frequency-selective multipath channels. The simulated channel consists of three multipath paths with random delays and random angles of arrival. The mixing matrices of each subcarrier are different, and the frequency selectivity is significant. This invention independently separates the signals on each subcarrier, which is equivalent to estimating the spatial separation matrix for each subcarrier separately, thus fully absorbing the channel frequency selectivity. Simulation results show that under this frequency-selective multipath condition, the signal acquisition method based on blind source separation in the multipath scenario of this invention still obtains a sharp correlation peak, verifying its natural adaptability to frequency selectivity.

[0115] (3) such as Figure 7 and Figure 8 As shown, the signal acquisition method based on blind source separation in multipath scenarios of this invention can effectively suppress strong interference. Simulation results show that, under the constraint of ISR=20dB, the output sidelobes of single-antenna correlation and equal-gain merging correlation are significantly amplified by strong interference, and the amplitude distribution of the normalized result in the time domain overlaps severely under H1 and H0, resulting in a high false alarm rate. This invention separates interference from the spatial dimension through subcarrier-by-subcarrier separation, highlighting the main correlation peak, shifting the distribution of correlation peak amplitude to the right, and significantly widening the distribution interval of the normalized result amplitude under the two assumptions. The false alarm rate is significantly lower than that of the two comparative methods, and the ROC curve of the signal acquisition method based on blind source separation in multipath scenarios of this invention is significantly closer to the lower left corner. The anti-interference capability is improved by nearly 10dB in the interference-to-signal ratio.

[0116] (4) The signal acquisition method based on blind source separation in multipath scenarios of the present invention can overcome the frequency selectivity failure problem of broadband beamforming. In the comparative simulation, equal-gain combining represents a simple spatial combining method that does not rely on directional information, and its performance improvement is limited under frequency-selective channels and strong interference. The present invention separates independently on subcarriers and achieves spatial optimal separation on all subcarriers, which is equivalent to realizing adaptive frequency-selective spatial filtering without directional prior. The performance in the simulation is significantly better than that of equal-gain combining.

[0117] The signal acquisition device based on blind source separation in multipath scenarios provided by the present invention is described below. The signal acquisition device based on blind source separation in multipath scenarios described below can be referred to in correspondence with the signal acquisition method based on blind source separation in multipath scenarios described above.

[0118] like Figure 9 As shown, a signal acquisition device based on blind source separation in a multipath scenario includes: The observation vector matrix construction module 901 is used to construct an observation vector matrix for multiple subcarriers based on multiple received signals from multiple receiving antennas. The blind source separation module 902 is used to perform blind source separation on the observation vector matrix of each subcarrier to obtain the separated signal matrix and the mixing matrix of each subcarrier; based on the mixing matrix of the subcarrier, the separated signals of the separated signal matrix of the subcarrier are reordered to obtain the target frequency domain component of the subcarrier. The acquisition result determination module 903 is used to perform dot product between the target frequency domain component of each subcarrier and the ZC sequence of the local frequency domain, perform inverse Fourier transform on the dot product result to obtain the correlation result in the time domain, normalize the correlation result in the time domain to obtain the normalized result in the time domain, determine the effective delay window based on the time point when the normalized result in the time domain first exceeds a preset threshold, and use the correlation peak position within the effective delay window as the signal acquisition position.

[0119] The signal acquisition device based on blind source separation in multipath scenarios provided in this invention constructs observation vector matrices for multiple subcarriers based on multiple received signals from multiple receiving antennas; performs blind source separation on the observation vector matrices of each subcarrier to obtain the separated signal matrix and mixing matrix of each subcarrier; reorders the separated signals of the separated signal matrix of the subcarrier based on the mixing matrix of the subcarrier to obtain the target frequency domain component of the subcarrier; performs a dot product between the target frequency domain component of each subcarrier and the ZC sequence of the local frequency domain, and performs an inverse Fourier transform on the dot product result to obtain the correlation result in the time domain; normalizes the correlation result in the time domain to obtain the normalized result in the time domain; determines the effective delay window based on the time point when the normalized result in the time domain first exceeds a preset threshold; and uses the correlation peak position within the effective delay window as the signal acquisition position, thereby improving the anti-interference capability of signal acquisition based on blind source separation in multipath scenarios. This invention, by independently separating the observation vector matrices of each subcarrier, fundamentally adapts to the actual situation where the mixing characteristics of each subcarrier in frequency-selective multipath signals are different, thus improving the anti-interference capability of signal acquisition based on blind source separation in multipath scenarios. This invention reorders the individual signals of the separation signal matrix according to the mixing matrix, avoiding the problem of ambiguous arrangement of the separation signals, thereby improving the anti-interference capability of signal acquisition based on blind source separation in multipath scenarios.

[0120] All relevant content of each step involved in the above method embodiments can be referenced from the functional description of the corresponding functional module, and will not be repeated here.

[0121] Figure 10 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 10As shown, the electronic device may include: a processor 1010, a communications interface 1020, a memory 1030, and a communications bus 1040, wherein the processor 1010, the communications interface 1020, and the memory 1030 communicate with each other through the communications bus 1040. The processor 1010 can call logic instructions in the memory 1030 to execute a signal acquisition method based on blind source separation in a multipath scenario. This method includes: constructing observation vector matrices for multiple subcarriers based on multiple received signals from multiple receiving antennas; performing blind source separation on the observation vector matrices of each subcarrier to obtain a separated signal matrix and a mixing matrix for each subcarrier; reordering the separated signals of the separated signal matrix of the subcarrier based on the mixing matrix of the subcarrier to obtain the target frequency domain component of the subcarrier; performing a dot product between the target frequency domain component of each subcarrier and the ZC sequence in the local frequency domain, and performing an inverse Fourier transform on the dot product result to obtain the correlation result in the time domain; normalizing the correlation result in the time domain to obtain a normalized result in the time domain; determining the effective delay window based on the time point when the normalized result in the time domain first exceeds a preset threshold; and using the correlation peak position within the effective delay window as the signal acquisition position.

[0122] Furthermore, the logical instructions in the aforementioned memory 1030 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0123] In another aspect, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon. When executed by a processor, the computer program implements the signal acquisition method based on blind source separation in multipath scenarios provided by the above methods. The method includes: constructing observation vector matrices of multiple subcarriers based on multiple received signals from multiple receiving antennas; performing blind source separation on the observation vector matrices of each subcarrier to obtain a separated signal matrix and a mixing matrix of each subcarrier; reordering the separated signals of the separated signal matrix of the subcarrier based on the mixing matrix of the subcarrier to obtain the target frequency domain component of the subcarrier; performing a dot product of the target frequency domain component of each subcarrier with the ZC sequence of the local frequency domain, and performing an inverse Fourier transform on the dot product result to obtain the correlation result in the time domain; normalizing the correlation result in the time domain to obtain a normalized result in the time domain; determining an effective delay window based on the time point when the normalized result in the time domain first exceeds a preset threshold; and using the position of the correlation peak within the effective delay window as the signal acquisition position.

[0124] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0125] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0126] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A signal acquisition method based on blind source separation in multipath scenarios, characterized in that, include: Based on multiple received signals from multiple receiving antennas, an observation vector matrix for multiple subcarriers is constructed. Blind source separation is performed on the observation vector matrix of each subcarrier to obtain the separated signal matrix and the mixing matrix of each subcarrier; The subcarrier-based hybrid matrix is ​​used to reorder the individual separated signals of the subcarrier's separated signal matrix to obtain the target frequency domain component of the subcarrier. The target frequency domain component of each subcarrier is multiplied by the ZC sequence in the local frequency domain, and the inverse Fourier transform is performed on the multiplication result to obtain the correlation result in the time domain. The correlation results in the time domain are normalized to obtain the normalized results in the time domain. The effective delay window is determined based on the time point when the normalized result in the time domain first exceeds a preset threshold. The position of the relevant peak within the effective delay window is used as the signal capture position.

2. The signal acquisition method based on blind source separation in multipath scenarios according to claim 1, characterized in that, The step of performing blind source separation on the observation vector matrices of each subcarrier to obtain the separated signal matrix and mixing matrix of each subcarrier includes: Calculate the sample covariance matrix of the observation vector matrix of each subcarrier, and perform eigenvalue decomposition; Based on the eigenvalue decomposition results of each subcarrier, a whitening matrix for each subcarrier is constructed; the eigenvalue decomposition results include the decomposed eigenvalues ​​and their eigenvectors. Based on the eigenvalue decomposition results of each subcarrier and the observation vector matrix of each subcarrier, the whitening data of each subcarrier is calculated. The separation matrix of each subcarrier is calculated based on the whitened data of each subcarrier; Based on the separation matrix and whitening data of each subcarrier, calculate the separation signal matrix of each subcarrier; Based on the separation matrix and whitening matrix of each subcarrier, the mixing matrix of each subcarrier is calculated.

3. The signal acquisition method based on blind source separation in multipath scenarios according to claim 2, characterized in that, The separation matrix of the subcarriers is calculated in the following manner: S1. The initial unit norm vectors of the initial separation matrix of the subcarrier are used as the unit norm vectors of the current iteration. S2. Based on the unit norm vector of each current iteration and the whitening data of the subcarrier, calculate each optimized unit norm vector; S3. Perform order reduction orthogonalization and normalization operations on each of the optimized unit norm vectors to obtain each updated unit norm vector. S4. Using multiple updated unit norm vectors as the unit norm vectors of the multiple current iterations, repeat steps S2-S4 to iteratively update the unit norm vectors until the difference between the unit norm vectors of two adjacent iterations tends to stabilize, or until a preset number of iterations is reached; construct the subcarrier separation matrix based on the multiple last updated unit norm vectors.

4. The signal acquisition method based on blind source separation in multipath scenarios according to claim 1, characterized in that, The subcarrier-based hybrid matrix reorders the separated signals of the subcarrier's separated signal matrix to obtain the target frequency domain components of the subcarrier, including: Calculate the norm of each column of the mixing matrix of the subcarrier; the norm of each column is proportional to the source power of each column; The separated signals in each column of the separated signal matrix are reordered based on the norm of each column; Based on all the reordered separated signals, the target frequency domain component of the subcarrier is determined.

5. The signal acquisition method based on blind source separation in multipath scenarios according to claim 1, characterized in that, The effective delay window is determined at the time point when the time-domain normalization result first exceeds a preset threshold, including: The time point at which the normalized result first exceeds a preset threshold is taken as the starting point of the effective delay window; The effective delay window is determined based on the starting point and the preset delay window duration.

6. The signal acquisition method based on blind source separation in multipath scenarios according to claim 1, characterized in that, The position of the relevant peak within the effective delay window is determined based on the following method: Within the effective time delay window, the position of the maximum value of the normalized result is taken as the position of the correlation peak.

7. The signal acquisition method based on blind source separation in multipath scenarios according to claim 1, characterized in that, The received signal includes N time-domain sampling points, where N≥2. The observation vector matrix for constructing multiple subcarriers based on the multiple received signals from multiple receiving antennas includes: Perform a fast Fourier transform on the N time-domain sampling points of each receiving antenna to output the N frequency-domain points of each receiving antenna; At all frequency domain points of the receiving antennas, there will be continuous The frequency domain points of each OFDM symbol are organized into a matrix according to the subcarrier index, thus obtaining the observation vector matrix of the multiple subcarriers. The number is greater than the number of receiving antennas.

8. A signal acquisition device based on blind source separation in a multipath scenario, characterized in that, include: The observation vector matrix construction module is used to construct an observation vector matrix for multiple subcarriers based on multiple received signals from multiple receiving antennas. The blind source separation module is used to perform blind source separation on the observation vector matrix of each subcarrier to obtain the separated signal matrix and the mixing matrix of each subcarrier; based on the mixing matrix of the subcarrier, the separated signals of the separated signal matrix of the subcarrier are reordered to obtain the target frequency domain component of the subcarrier. The capture result determination module is used to perform a dot product between the target frequency domain component of each subcarrier and the ZC sequence of the local frequency domain, and perform an inverse Fourier transform on the dot product result to obtain the correlation result in the time domain. The relevant results in the time domain are normalized to obtain the normalized results in the time domain; the effective delay window is determined based on the time point when the normalized results in the time domain first exceed a preset threshold. The position of the relevant peak within the effective delay window is used as the signal capture position.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the signal acquisition method based on blind source separation in a multipath scenario as described in any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the signal acquisition method based on blind source separation in multipath scenarios as described in any one of claims 1 to 7.

Images