Digital signal processing method based on 5G repeater interference signal automatic shielding

Abstract

The invention discloses a digital signal processing method based on 5G repeater interference signal automatic shielding, relates to the technical field of digital signal processing, and solves the technical problems of insufficient utilization of interference space features and poor interference cooperative suppression effect. The method comprises the following steps: according to historical data of a repeater area and a base station load dynamic calculation threshold, adapting to different scenes such as a city core area and a suburban area, reducing interference misjudgment and missed judgment rates, breaking through single time domain power threshold judgment, extracting spatial features such as DOA, a polarization angle and an inclination angle of interference, and utilizing a DBSCAN clustering algorithm to cluster an interference group based on spatial feature dimensions, so as to realize the interference suppression of the interference. According to the method, multipath interference, coherent interference and isolated interference are distinguished, a suppression system of'interference group space correlation graph + cooperative beam forming + real-time updating 'is constructed, strong correlation interference sub-graphs are divided by using a graph segmentation algorithm, unified shielding parameters are designed for the sub-graphs, and a suppression strategy is more targeted.

Description

Technical Field

[0001] The present invention relates to the technical field of digital signal processing, and specifically to a digital signal processing method based on automatic shielding of 5G repeater interference signals. Background Art

[0002] In wireless communications, especially 5G communications, repeaters and other equipment, while enhancing signal coverage, can also introduce various interference signals, such as drone interference and co-channel interference from neighboring cells. This interference can affect communication quality, leading to signal misjudgment and abnormal base station loads.

[0003] Patent application publication number CN107357169B discloses an automated control model for shielding interference signals. The model includes a temperature preprocessing block, an alarm block, a comparison block, and a 5-second on-delay timer. The temperature preprocessing block is sequentially connected to the alarm block and the on-delay timer, which are then connected to the comparison block. The temperature preprocessing block preprocesses the temperature signal to obtain a temperature signal value. The comparison block compares the preprocessed temperature signal value with a set value and then executes the model according to different program control models. The invention ensures accurate and reliable signals, ensures timely and effective interlocking, and simultaneously identifies and shields signal mutations caused by interference. In actual field applications, the shielding rate has reached 100%.

[0004] Although existing technologies can perform simple signal power threshold judgment and interference classification, they have the following shortcomings when faced with complex scenarios: Classifying interference based solely on time-domain power thresholds makes it difficult to accurately identify interference signals in complex scenarios and is prone to misjudgment. For the classified interference, there is a lack of fine clustering from the spatial feature dimension, and it is impossible to distinguish "interference groups", resulting in insufficient targeting of the suppression strategy; Interference suppression often uses simple shielding or fixed beamforming, without taking into account the spatial correlation characteristics of the interference group for collaborative optimization. This results in poor suppression effect and insufficient real-time performance, making it difficult to adapt to the needs of 5G dynamic scenarios. Summary of the Invention

[0005] In response to the shortcomings of the existing technology, the present invention provides a digital signal processing method based on automatic shielding of 5G repeater interference signals, which solves the problems of insufficient utilization of interference spatial characteristics and poor interference collaborative suppression effect.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: a digital signal processing method based on automatic shielding of 5G repeater interference signals, the method specifically comprising the following steps: Calculate the power integral of the data signal and determine the preset threshold based on historical communication data. Compare the two and filter out interference signals and normal signals. Interference signals are shielded and their arrival directions are calculated using a subspace algorithm. Their polarization characteristics are then acquired. The correlation between interference signals is analyzed based on the spatial coherence coefficient, and correlated and uncorrelated interference signals are classified. The DBSCAN clustering algorithm is used to construct a data set with DOA and polarization features as coordinates. The number of other points in the neighborhood of each point in the data set is calculated according to the Euclidean formula, and the same interference group is classified according to the number. The same interference group is collaboratively suppressed, an optimization objective function is constructed, and the beamforming weight vector is calculated. It is used as a standard for suppression and combined with the interference correlation graph for comprehensive shielding.

[0007] As a further solution of the present invention, the specific method of screening interference signals and normal signals is: The digital signal is labeled r, and r=1, 2, ..., f, where f represents the type of digital signal. According to the formula Calculate the power integral Ei corresponding to the digital signal i, where t1 and t2 are time windows. Calculate the instantaneous power of the signal r(t); Collect historical communication data in the area where the repeater is located and build a threshold calculation model , calculate the preset threshold Ey, where L is the current load rate of the base station, E avg is the historical normal signal average power integral, and is the scene coefficient, and the two are compared to screen interference signals and normal signals.

[0008] As a further solution of the present invention, the specific method of comparing the two to screen the interference signal and the normal signal is: If the power integral E r If the power integral Ey is greater than the preset threshold, the corresponding digital signal is classified as an interference signal and is labeled as i, and i=1, 2, ..., j, where j represents the type of interference signal. r If the value is less than the preset threshold Ey, the corresponding numerical signal is classified as a normal signal.

[0009] As a further solution of the present invention, the specific method of shielding the interference signal is: Utilize 5G base station MIMO antenna arrays to synchronously collect multi-channel signals and form a spatiotemporal data set , where N is the number of antennas and T is the number of sampling points. The subspace algorithm is used to calculate the arrival direction of the interference signal and the signal polarization characteristics, which include polarization angle and tilt angle. Calculate the sample covariance matrix for spatiotemporal data X: , where XH is the conjugate transpose of X, T is the number of snapshots, and through eigendecomposition Separate signal subspace U s With the noise subspace U n , then construct the spatial spectrum function: ,in is the array steering vector, =[1, ,…, ] T , spectrum peak search obtains the interference signal arrival direction DOA= , where j is the number of interference signals; Polarization angle : , where E x and E y are the horizontal and vertical polarization components, respectively, and the inclination : .

[0010] As a further solution of the present invention, the specific method of classifying to obtain the relevant interference signals and the unrelated interference signals is: According to the formula Calculate the spatial coherence coefficient between interference signals, where a i is the steering vector of the i-th interference signal, for a i The conjugate transpose of Compare with judgment threshold; If the spatial coherence coefficient > judgment threshold, the two corresponding interference signals are marked as correlated interference signals. Conversely, if the spatial coherence coefficient < judgment threshold, it means that the two interference signals are spatially irrelevant and are recorded as uncorrelated interference signals.

[0011] As a further solution of the present invention, the specific method of classifying the same interference group according to quantity is: The DOA and polarization angle of each interference signal and inclination Combining them into a feature vector, the feature vector of the i-th interference signal is represented as x i =[DOA i , , ], where i=1, 2, ..., j, and further obtain the data set X=[x1, x2, ..., x j ]; For each point x in the dataset i Calculate its radius Other points in the neighborhood are calculated based on the Euclidean distance, and point x i [DOA i , , ] and x j [DOA j , , ]Substitute into the formula to calculate the Euclidean distance d(x i , x j ), and d(x i , x j ) = , and the resulting Euclidean distance d(x i , x j ) and radius The neighborhood is compared to determine the quantity.

[0012] As a further solution of the present invention, the specific method of comparing the obtained Euclidean distance with the neighborhood to determine the quantity is: If d(x i , x j )≤ , then it is considered that point x i and x j In the neighborhood, get the radius at the same time The minimum number of points MinPts corresponding to the neighborhood, then for each point x i Perform classification and identification; If point x i of If the number of points in the neighborhood is greater than or equal to the minimum number of points MinPts, then the point is recorded as the core point. i It is not a core point, but it falls on a certain core point If the point x is within the neighborhood, then the point is recorded as a boundary point. i If a point is neither a core point nor a boundary point, it will be recorded as a noise point.

[0013] As a further solution of the present invention, the specific method of collaboratively suppressing the same interference group is: Constructing the optimization objective function , where w is the beamforming weight vector, R is the signal covariance matrix, is the desired signal steering vector, w H Rw is the beam output power, w H =1 is the constraint condition; Based on the above optimization objective function, construct the Lagrangian function ,in is the carrier wavelength, taking the derivative with respect to w and setting the derivative to 0 gives , combined with the constraint w H =1, the combined solution is , and perform suppression processing according to the obtained beamforming weight vector w; At the same time, the interference sources are regarded as graph nodes, the spatial correlation is used as the edge weight, and an interference correlation graph is constructed. The graph segmentation algorithm is used to divide the strongly correlated interference into subgraphs, and a unified shielding parameter is designed for each subgraph.

[0014] The present invention provides a digital signal processing method based on automatic shielding of 5G repeater interference signals. Compared with the existing technology, it has the following advantages: The present invention combines the "time domain power integration + scenario-based threshold model" and dynamically calculates the threshold based on the historical data of the repeater area and the base station load, adapts to different scenarios such as urban core areas and suburbs, reduces the interference misjudgment and missed judgment rates, breaks through the single time domain power threshold judgment, extracts the spatial characteristics of interference such as DOA, polarization angle, and inclination angle, and uses the DBSCAN clustering algorithm to cluster "interference groups" based on the spatial feature dimension to distinguish multipath interference, coherent interference and isolated interference.

[0015] The present invention constructs a suppression system of "interference group spatial correlation graph + collaborative beamforming + real-time update", uses a graph segmentation algorithm to divide strongly correlated interference subgraphs, designs unified shielding parameters for the subgraphs, and makes the suppression strategy more targeted. It uses recursive least squares to update the covariance matrix in real time to adapt to 5G dynamic scenarios and ensure that the suppression effect is sustained and effective. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a diagram of the steps of the present invention. DETAILED DESCRIPTION

[0017] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0018] See also Figure 1 , this application provides a digital signal processing method based on automatic shielding of 5G repeater interference signals, the method specifically comprising the following steps: Step S1: Get the digital signal corresponding to the repeater and label the digital signal as r, where r=1, 2, ..., f, where f represents the type of digital signal. At the same time, calculate the power integral of the digital signal r in the time domain. According to the formula Calculate the power integral Ei corresponding to the digital signal i, where t1 and t2 are time windows. Calculate the instantaneous power of the signal r(t), and integrate the power E r The value is compared with the preset threshold Ey, and the specific calculation method of the preset threshold Ey is as follows: By collecting historical communication data in the area where the repeater is located, calculating the normal signal power integral range, and combining it with the current base station load, a threshold calculation model is constructed. , calculate the preset threshold Ey, where L is the current load rate of the base station, E avg is the historical normal signal average power integral, and is the scenario coefficient, such as the urban core area =0.8, =0.2, suburban =0.9, =0.1; If the power integral E r If the power integral Ey is greater than the preset threshold, the corresponding digital signal is classified as an interference signal and is labeled as i, and i=1, 2, ..., j, where j represents the type of interference signal. r If the value is less than the preset threshold Ey, the corresponding numerical signal is classified as a normal signal.

[0019] Take a 5G repeater station of an operator in the city core area as an example: during the evening peak of weekdays (18:00-20:00), the repeater covers an area with dense office buildings, the base station load rate reaches 75%, and the time window [t1, t2] = 10ms is set. The digital signal is collected and labeled to calculate the power integral. The historical normal signal average power integral E avg =5, according to the formula The preset threshold Ey=4.15 is calculated; The power integral E of a digital signal r is 6. Since 6>4.15, the digital signal is an interference signal.

[0020] Step S2: Shield the classified interference signals and use the 5G base station MIMO antenna array (such as 64T64R) to synchronously collect multi-channel signals to form a time-space domain data set. , where N is the number of antennas and T is the number of sampling points. For example, in a drone interference scenario, a uniform linear array (ULA) is used to collect signals, and the phase difference between antennas is used to extract spatial angle information. A subspace algorithm (such as MUSIC and ESPRIT) is used to calculate the direction of arrival (DOA) of the interference signal and the signal polarization characteristics, which include polarization angle and tilt angle. The specific processing method is as follows: Taking drone jamming as an example, the ULA array is preferred, taking advantage of its high resolution in the horizontal dimension. The array parameters are set as follows: Array element spacing d= / 2, is the carrier wavelength, for example, 3.5 GHz corresponds to d≈4.3 cm; Array length L = (N-1)d; Calculate the sample covariance matrix for spatiotemporal data X: , where X H is the conjugate transpose of X, T is the number of snapshots, and satisfies T≥2N; By eigendecomposition Separate signal subspace U s With the noise subspace U n , then construct the spatial spectrum function: ,in is the array steering vector, specifically in the ULA (Uniform Linear Array) scenario, =[1, ,…, ] T , spectrum peak search obtains the interference signal arrival direction DOA= , where j is the number of interference signals; Using dual-polarization array elements (such as Polarization), analyze the polarization state of the interference signal: Polarization angle : describes the rotation direction of the polarization ellipse, , where E x and E y are the horizontal and vertical polarization components respectively; inclination : describes the axis ratio of the polarization ellipse, ; For the received signal of the dual-polarization array element, the polarization correlation coefficient matrix is ​​calculated, and the polarization ellipse parameters are fitted by combining the maximum likelihood estimation (MLE) to achieve the polarization angle and inclination estimates; By performing the inner product operation of the steering vector, the spatial correlation between the interference signals is quantified. According to the formula Calculate the spatial coherence coefficient between interference signals, where a i is the steering vector of the i-th interference signal, for a i The conjugate transpose of Compared with the judgment threshold, if the spatial coherence coefficient > judgment threshold, the two corresponding interference signals are marked as correlated interference signals. Conversely, if the spatial coherence coefficient < judgment threshold, it means that the two interference signals are spatially irrelevant and are recorded as uncorrelated interference signals; Step S3: Use the DBSCAN clustering algorithm to classify the related interference signals into the same "interference group" based on DOA and polarization characteristics. The specific processing method is as follows: The DOA and polarization angle of each interference signal and inclination Combining them into a feature vector, the feature vector of the i-th interference signal is represented as x i =[DOA i , , ], where i=1, 2, ..., j, and further obtain the data set X=[x1, x2, ..., x j ]; For each point x in the dataset i Calculate its radius Other points in the neighborhood are calculated based on the Euclidean distance, and point x i [DOA i , , ] and x j [DOA j , , ]Substitute into the formula to calculate the Euclidean distance d(x i , x j ), and d(x i , x j ) = , and the resulting Euclidean distance d(x i , x j ) and radius Make comparisons; If d(x i , x j )≤ , then it is considered that point x i and x j In the neighborhood, get the radius at the same time The minimum number of points MinPts corresponding to the neighborhood is specifically expressed in a point The minimum number of points that must be included in the neighborhood to be considered a core point, and then for each point x i Perform classification and identification; If point x i of If the number of points in the neighborhood is greater than or equal to (≥) the minimum number of points MinPts, then the point is recorded as a core point. i It is not a core point, but it falls on a certain core point If the point x is within the neighborhood, then the point is recorded as a boundary point. iIf it is neither a core point nor a boundary point, it will be recorded as a noise point, and so on to obtain all the same "interference groups".

[0021] Combined with actual analysis, such as 5G millimeter wave base stations (25GHz, =10.7mm), using a 64-element ULA (Uniform Linear Array) to cover three-dimensional spatial interference, including drone interference (multipath scenario), which generates three main interference paths, as shown in the following table: For each interference path, calculate the normalized feature: Path 1: x1 = [0.667, 0.5, 0.333] T , Path 2: x2=[0.678, 0.522, 0.356] T , Path 3: x3=[0.833, 0.667, 0.5] T ; Calculate the normalized distance between path 1 and path 2, d(x1, x2) = = 0.031 (small distance, highly similar features), the normalized distance d(x1, x3) between path 1 and path 3 = =0.285 (large distance, significant feature difference); Path 1 Neighborhood: contains path 2 (distance 0.031≤0.05), and the number of points in the neighborhood = 2, which is less than the minimum number of points 3, and is not determined as a core point; Path 2 Neighborhood: contains path 1 (distance 0.031≤0.05), and the number of points in the neighborhood = 2, which is less than the minimum number of points 3, and is not determined as a core point; Path 3 Neighborhood: There are no other interference points (distance path 1 / 2> 0.05), the number of points in the neighborhood = 0, and it is judged as a noise point.

[0022] Since drone multipath interference usually has a continuous density distribution, after collecting more snapshot data, the number of points in the neighborhood of path 1 / 2 reaches the minimum number of points MinPts=3, so the number of points in the neighborhood of path 1 / 2 is If the number of points in the neighborhood is ≥3, it will be upgraded to a core point. Path 3 is still isolated and is determined to be a noise point.

[0023] Step S4: perform collaborative suppression on the obtained same “interference group” and construct an optimization objective function: , where w is the beamforming weight vector (N×1), R is the signal covariance matrix (N×N), is the desired signal steering vector (N×1), w H Rw is the beam output power (scalar), wH =1 is the constraint condition; Based on the above optimization objective function, the Lagrangian function is constructed: ,in is the carrier wavelength, take the derivative with respect to w and set the reciprocal to 0 to obtain , combined with the constraint w H =1, and the combined solutions are: , and suppression is performed based on the obtained beamforming weight vector w. Meanwhile, the interference sources are considered as graph nodes, and spatial correlations are used as edge weights to construct an interference correlation graph. A graph segmentation algorithm (such as spectral clustering) is used to partition strongly correlated interference into subgraphs, and unified shielding parameters (such as filter coefficients and beam null direction) are designed for each subgraph. Then, recursive least squares (RLS) is used to update the covariance matrix R with a period of 10ms. ,in is the forgetting factor, and The value is 0.98.

[0024] Some of the data in the above formulas are calculated based on their numerical values ​​and are not substituted into parameter units for calculation. At the same time, the contents not described in detail in this specification belong to the existing technology known to those skilled in the art.

[0025] The above embodiments are only used to illustrate the technical method of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A digital signal processing method based on automatic shielding of 5G repeater interference signals, characterized in that: The method specifically comprises the following steps: Calculate the power integral of the data signal and determine the preset threshold based on historical communication data. Compare the two and filter out interference signals and normal signals. Interference signals are shielded and their arrival directions are calculated using a subspace algorithm. Their polarization characteristics are then acquired. The correlation between interference signals is analyzed based on the spatial coherence coefficient, and correlated and uncorrelated interference signals are classified. The DBSCAN clustering algorithm is used to construct a data set with DOA and polarization features as coordinates. The number of other points in the neighborhood of each point in the data set is calculated according to the Euclidean formula, and the same interference group is classified according to the number. The same interference group is collaboratively suppressed, an optimization objective function is constructed, and the beamforming weight vector is calculated. It is used as a standard for suppression and combined with the interference correlation graph for comprehensive shielding.

2. The digital signal processing method based on automatic shielding of 5G repeater interference signals according to claim 1 is characterized in that: The specific method of screening interference signals and normal signals is: The digital signal is labeled r, and r=1, 2, ..., f, where f represents the type of digital signal. According to the formula Calculate the power integral Ei corresponding to the digital signal i, where t1 and t2 are time windows. Calculate the instantaneous power of the signal r(t); Collect historical communication data in the area where the repeater is located and build a threshold calculation model , calculate the preset threshold Ey, where L is the current load rate of the base station, E avg is the historical normal signal average power integral, and is the scene coefficient, and the two are compared to screen interference signals and normal signals.

3. The digital signal processing method based on automatic shielding of 5G repeater interference signals according to claim 2 is characterized in that: The specific method of comparing the two to screen interference signals and normal signals is: If the power integral E r If the power integral Ey is greater than the preset threshold, the corresponding digital signal is classified as an interference signal and is labeled as i, and i=1, 2, ..., j, where j represents the type of interference signal. r If the value is less than the preset threshold Ey, the corresponding numerical signal is classified as a normal signal.

4. The digital signal processing method based on automatic shielding of 5G repeater interference signals according to claim 1 is characterized in that: The specific method of shielding the interference signal is as follows: Utilize 5G base station MIMO antenna arrays to synchronously collect multi-channel signals and form a spatiotemporal data set , where N is the number of antennas and T is the number of sampling points. The subspace algorithm is used to calculate the arrival direction of the interference signal and the signal polarization characteristics, which include polarization angle and tilt angle. Calculate the sample covariance matrix for spatiotemporal data X: , where X H is the conjugate transpose of X, T is the number of snapshots, and through eigendecomposition Separate signal subspace U s With the noise subspace U n , then construct the spatial spectrum function: ,in is the array steering vector, =[1, ,…, ] T , spectrum peak search obtains the interference signal arrival direction DOA= , where j is the number of interference signals; Polarization angle : , where E x and E y are the horizontal and vertical polarization components, respectively, and the inclination : .

5. The digital signal processing method based on automatic shielding of 5G repeater interference signals according to claim 1 is characterized in that: The specific method of obtaining the relevant interference signals and the unrelated interference signals by classification is: According to the formula Calculate the spatial coherence coefficient between interference signals, where a i is the steering vector of the i-th interference signal, for a i The conjugate transpose of Compare with judgment threshold; If the spatial coherence coefficient > judgment threshold, the two corresponding interference signals are marked as correlated interference signals. Conversely, if the spatial coherence coefficient < judgment threshold, it means that the two interference signals are spatially irrelevant and are recorded as uncorrelated interference signals.

6. The digital signal processing method based on automatic shielding of 5G repeater interference signals according to claim 1 is characterized in that: The specific method of classifying the same interference group according to quantity is as follows: The DOA and polarization angle of each interference signal and inclination Combining them into a feature vector, the feature vector of the i-th interference signal is represented as x i =[DOA i , , ], where i=1, 2, ..., j, and further obtain the data set X=[x1, x2, ..., x j ]; For each point x in the dataset i Calculate its radius Other points in the neighborhood are calculated based on the Euclidean distance, and point x i [DOA i , , ] and x j [DOA j , , ]Substitute into the formula to calculate the Euclidean distance d(x i , x j ), and d(x i , x j ) = , and the resulting Euclidean distance d(x i , x j ) and radius The neighborhood is compared to determine the quantity.

7. The digital signal processing method based on automatic shielding of 5G repeater interference signals according to claim 6 is characterized in that: The specific method of comparing the obtained Euclidean distance with the neighborhood to determine the quantity is: If d(x i , x j )≤ , then it is considered that point x i and x j In the neighborhood, get the radius at the same time The minimum number of points MinPts corresponding to the neighborhood, then for each point x i Perform classification and identification; If point x i of If the number of points in the neighborhood is greater than or equal to the minimum number of points MinPts, then the point is recorded as the core point. i It is not a core point, but it falls on a certain core point If the point x is within the neighborhood, then the point is recorded as a boundary point. i If a point is neither a core point nor a boundary point, it will be recorded as a noise point.

8. The digital signal processing method based on automatic shielding of 5G repeater interference signals according to claim 1 is characterized in that: The specific method of collaboratively suppressing the same interference group is: Constructing the optimization objective function , where w is the beamforming weight vector, R is the signal covariance matrix, is the desired signal steering vector, w H Rw is the beam output power, w H =1 is the constraint condition; Based on the above optimization objective function, construct the Lagrangian function ,in is the carrier wavelength, taking the derivative with respect to w and setting the derivative to 0 gives , combined with the constraint w H =1, the combined solution is , and perform suppression processing according to the obtained beamforming weight vector w; At the same time, the interference sources are regarded as graph nodes, the spatial correlation is used as the edge weight, and an interference correlation graph is constructed. The graph segmentation algorithm is used to divide the strongly correlated interference into subgraphs, and a unified shielding parameter is designed for each subgraph.

Images