A network performance monitoring method and system based on a 5G antenna

Abstract

The application relates to the technical field of network monitoring, and discloses a network performance monitoring method and system based on a 5G antenna, wherein the network performance monitoring method based on the 5G antenna comprises the following steps: step S1, linear signal power and linear noise power are calculated; step S2, coverage energy weight is calculated; step S3, spatial barycenter coordinates are calculated; step S4, frequency space slope is calculated; step S5, structure tilt angle is calculated; step S6, endpoint coordinates are calculated; and step S7, pointing deviation angle is calculated. According to the application, the weak coverage spatial barycenter is calculated in frequency subbands, the coverage difference between different subbands can be identified, the structure tilt angle is inversed through the frequency space slope, the coverage drift caused by pure beam oblique vision and antenna structure deformation can be quantitatively distinguished, and the pointing deviation is calculated by comprehensively considering the physical line segment length and the structure tilt angle, so that accurate basis is provided for network optimization and antenna maintenance.

Description

Technical Field

[0001] This invention relates to the field of network monitoring technology, and more specifically, to a network performance monitoring method and system based on a 5G antenna. Background Technology

[0002] 5G NR (Fifth Generation New Radio) technology achieves beam management and physical layer measurement through SSB (Synchronization Block) and CSI-RS (Channel State Information Reference Signal). Existing network performance monitoring systems support statistics on RSRP (Reference Signal Received Power), SINR (Signal-to-Interference-plus-Noise Ratio), beamswitch counts, and MU-MIMO (Multi-User Multiple-Input Multiple-Output) layers by cell or beam. Base station antennas commonly employ multi-port phased array structures, enabling flexible beam scanning and pointing adjustments, and are widely used in macro and micro base station deployments.

[0003] Existing network performance monitoring methods primarily rely on averaged KPIs (Key Performance Indicators) for coverage assessment, typically attributing weak coverage to factors such as insufficient transmit power, excessive inter-site spacing, or obstruction. These methods have limitations in addressing coverage issues caused by antenna structure and frequency coupling; they cannot distinguish coverage differences between different frequency sub-bands, nor can they establish a correlation between network performance degradation and the physical state of the antenna structure.

[0004] The root cause of the aforementioned defects lies in the fact that existing methods do not consider the coupling effect between the inherent beam slant of broadband phased arrays and the micro-deflection of the antenna panel. In specific scenarios such as high-rise street valleys, minute antenna structural deformations amplify frequency-related beam pointing offsets, causing the same nominal beam to spatially drift in weak coverage areas of different subbands. Averaged KPIs mask these spatial differences between subbands, making it difficult for maintenance personnel to promptly detect and locate such hidden coverage problems. This invention aims to solve the above-mentioned technical problems by providing a network performance monitoring method capable of identifying coverage drift caused by antenna structure and frequency coupling. Summary of the Invention

[0005] This invention provides a network performance monitoring method and system based on 5G antennas, which solves the technical problems mentioned in the background.

[0006] This invention provides a network performance monitoring method based on 5G antennas, comprising the following steps: Step S1: Within the monitoring time window, collect the logarithmic power of the signal and the logarithmic power of the noise according to the beam number, frequency band number and sample location point, and convert the logarithmic power of the signal and the logarithmic power of the noise into linear signal power and linear noise power. Step S2: Divide the linear signal power by the linear noise power to generate a power ratio, and calculate the coverage energy weight based on the power ratio; Step S3: The sample location points are weighted by the coverage energy weight to generate the spatial barycenter coordinates corresponding to each frequency band number. Step S4: Calculate the average frequency value by calculating the average center frequency of the frequency band, calculate the average coordinate value by calculating the average centroid coordinates of the space, and calculate the frequency space slope based on the frequency band center frequency, the centroid coordinates of the space, the average frequency value, and the average coordinate value. Step S5: Project the frequency space slope onto the beam azimuth direction, and calculate the structure tilt angle by combining the reference frequency value, beam azimuth angle, beam elevation angle and antenna height value. Step S6: Calculate the reference centroid coordinates using the frequency space slope and the reference frequency value, and calculate the coordinates of the first endpoint and the second endpoint based on the frequency space slope, the reference centroid coordinates, the minimum boundary frequency, and the maximum boundary frequency. Step S7: Calculate the physical line segment length between the coordinates of the first endpoint and the coordinates of the second endpoint. Combine the physical line segment length, antenna height value and structural tilt angle to calculate the pointing deviation angle and output it as the monitoring result.

[0007] This invention provides a network performance monitoring system based on a 5G antenna, comprising: The data acquisition module collects the logarithmic power of the signal and the logarithmic power of the noise based on the beam number, frequency band number and sample location points within the monitoring time window, and converts the logarithmic power of the signal and the logarithmic power of the noise into linear signal power and linear noise power. The coverage energy weight calculation module divides the linear signal power by the linear noise power to generate a power ratio, and calculates the coverage energy weight based on the power ratio. The spatial barycenter coordinate calculation module calculates the spatial barycenter coordinates corresponding to each frequency band number by weighting the coordinates of the sample location points through the coverage energy weight. The frequency space slope calculation module calculates the average frequency value by calculating the average center frequency of the frequency band, and the average coordinate value by calculating the average centroid coordinates of the space. It then calculates the frequency space slope based on the center frequency of the frequency band, the centroid coordinates of the space, the average frequency value, and the average coordinate value. The structural tilt angle calculation module projects the frequency space slope onto the beam azimuth direction and calculates the structural tilt angle by combining the reference frequency value, beam azimuth angle, beam elevation angle and antenna height value. The endpoint coordinate calculation module calculates the reference centroid coordinates using the frequency space slope and the reference frequency value, and calculates the first endpoint coordinates and the second endpoint coordinates based on the frequency space slope, the reference centroid coordinates, the minimum boundary frequency, and the maximum boundary frequency. The pointing deviation angle calculation module calculates the physical line segment length between the coordinates of the first endpoint and the coordinates of the second endpoint, and calculates the pointing deviation angle by combining the physical line segment length, antenna height value and structural tilt angle, and outputs it as the monitoring result.

[0008] The beneficial effects of this invention are as follows: By calculating the spatial centroid of weak coverage through frequency sub-band division, this invention can identify coverage differences between different sub-bands, avoiding hidden coverage problems masked by averaged KPI indicators. By calculating the frequency spatial slope using the centered least squares method, it can quantitatively extract the coverage drift characteristics caused by flexural amplification of frequency slope, improving the robustness and stability of drift trend calculation. By inverting the structural tilt angle through the frequency spatial slope, it can establish a quantitative correlation between network performance measurement results and the antenna physical structure state, quantitatively distinguishing between coverage drift caused by pure beam squint and antenna structural deformation. By generating physical weak coverage drift segments, it can transform frequency domain characteristics into visible spatial domain results, facilitating engineers to quickly understand and locate coverage problems. By calculating the pointing deviation by comprehensively considering the physical segment length and structural tilt angle, it can comprehensively reflect the overall impact of antenna structure and frequency coupling on beam pointing, providing an accurate basis for network optimization and antenna maintenance. Attached Figure Description

[0009] Figure 1 This is a calculation flowchart of a network performance monitoring method based on a 5G antenna according to the present invention; Figure 2 This is a schematic diagram of a network performance monitoring system based on a 5G antenna according to the present invention. Detailed Implementation

[0010] The subject matter described herein will now be discussed with reference to exemplary embodiments. It should be understood that these embodiments are discussed only to enable those skilled in the art to better understand and implement the subject matter described herein, and changes may be made to the function and arrangement of the elements discussed without departing from the scope of this specification. Various processes or components may be omitted, substituted, or added as needed in the examples. Furthermore, features described in some examples may be combined in other examples.

[0011] It should be noted that, unless otherwise defined, the technical or scientific terms used in one or more embodiments of the present invention should have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in one or more embodiments of the present invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" indicate that the element or object preceding the term encompasses the elements or objects listed following the term and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0012] like Figures 1-2 As shown, a network performance monitoring method based on 5G antennas includes the following steps: Step S1: Within the monitoring time window, collect the logarithmic power of the signal and the logarithmic power of the noise according to the beam number, frequency band number and sample location point, and convert the logarithmic power of the signal and the logarithmic power of the noise into linear signal power and linear noise power. Step S2: Divide the linear signal power by the linear noise power to generate a power ratio, and calculate the coverage energy weight based on the power ratio; Step S3: The sample location points are weighted by the coverage energy weight to generate the spatial barycenter coordinates corresponding to each frequency band number. Step S4: Calculate the average frequency value by calculating the average center frequency of the frequency band, calculate the average coordinate value by calculating the average centroid coordinates of the space, and calculate the frequency space slope based on the frequency band center frequency, the centroid coordinates of the space, the average frequency value, and the average coordinate value. Step S5: Project the frequency space slope onto the beam azimuth direction, and calculate the structure tilt angle by combining the reference frequency value, beam azimuth angle, beam elevation angle and antenna height value. Step S6: Calculate the reference centroid coordinates using the frequency space slope and the reference frequency value, and calculate the coordinates of the first endpoint and the second endpoint based on the frequency space slope, the reference centroid coordinates, the minimum boundary frequency, and the maximum boundary frequency. Step S7: Calculate the physical line segment length between the coordinates of the first endpoint and the coordinates of the second endpoint. Combine the physical line segment length, antenna height value and structural tilt angle to calculate the pointing deviation angle and output it as the monitoring result.

[0013] In one embodiment of the present invention, the specific calculation process of step S1 includes: According to the Beam number, the Frequency band number, number The formula for calculating the linear signal power is as follows: (This is the logarithmic power of the signal at each sample location.) in Indicates the first Beam number, the Frequency band number, number Linear signal power at each sample location point Indicates the first Beam number, the Frequency band number, number Logarithmic power of the signal at each sample location point Indicates the beam number. Indicates the frequency band number. Indicates the sample location point number; According to the Beam number, the Frequency band number, number The formula for calculating the linear noise power is as follows: (Noise log power at each sample location) in Indicates the first Beam number, the Frequency band number, number Linear noise power at each sample location point Indicates the first Beam number, the Frequency band number, number The noise log power at each sample location point Indicates the beam number. Indicates the frequency band number. Indicates the sample location point number.

[0014] It should be noted that the beam sequence number is used to identify different 5G beams and is a base station configuration parameter, which can be obtained through the base station network management system. The frequency band sequence number is used to identify different frequency sub-bands within the monitoring bandwidth; it is a user-defined parameter with a value range of 1 to M, where M is the total number of frequency bands. The sample location point is used to identify the collection location of network measurement data and can be obtained through minimized drive test MDT, professional drive test equipment, or measurement grid data generated by the network management system. The signal logarithmic power represents the reference signal received power in decibels and milliwatts (dW / mW) and can be obtained through user equipment measurement and reporting or professional drive test equipment collection. The noise logarithmic power represents the interference plus noise power in decibels and milliwatts (dW / mW) and can be obtained through user equipment measurement and reporting or professional drive test equipment collection. The frequency band center frequency represents the center operating frequency of each frequency sub-band and is a base station configuration parameter, which can be obtained through the base station network management system. The beam azimuth angle represents the pointing angle of the beam in the horizontal plane and is a base station configuration parameter, which can be obtained through the base station network management system. The beam elevation angle represents the pointing angle of the beam in the vertical plane and is a base station configuration parameter, which can be obtained through the base station network management system. Antenna height represents the installation height of the base station antenna, a base station configuration parameter that can be obtained from the base station engineering parameter table. Sample location point coordinates represent the coordinates of the sample location point in a two-dimensional ground coordinate system, which can be obtained through the user equipment's Global Navigation Satellite System module or the positioning module of professional drive-test equipment. Linear signal power represents the received power of the reference signal in watts, calculated from the signal logarithmic power. Linear noise power represents the interference plus noise power in watts, calculated from the noise logarithmic power.

[0015] It should be noted that the duration of the same monitoring time window can be set according to network load and measurement accuracy requirements, typically ranging from 15 minutes to 1 hour. For example, when the network load is high (e.g., when the average number of online users in the cell exceeds 500, or the average throughput of the downlink physical shared channel exceeds 800 megabits per second), the monitoring time window can be appropriately shortened to 15 minutes to ensure the real-time nature of the measurement data and to promptly detect changes in network performance. When the measurement accuracy requirements are high (e.g., when it is necessary to detect antenna structure tilt angles of less than 0.02 radians, or when the accuracy requirement for locating weak coverage areas is better than 5 meters), the monitoring time window can be appropriately extended to 1 hour to increase the sample size and improve the reliability of the statistical results. Measurement samples can come from the Minimum Drive Test MDT, measurement grid data generated by professional drive test equipment, or network management systems. Specifically, Minimum Drive Test (MDT) data consists of measurement data automatically reported by user equipment during normal communication. This data includes reference signal received power, interference plus noise power, and location information. No additional testing equipment is required, enabling large-scale, all-weather network measurements. Professional drive test equipment, via vehicle-mounted or handheld methods, measures the target area, collecting high-precision data suitable for detailed testing in key areas. Measurement grid data generated by the network management system divides the target area into several equally sized grids and calculates the average measurement data within each grid, suitable for macro-level network performance analysis. Furthermore, all measurement samples are anonymized, eliminating the need to collect user identity information and protecting user privacy.

[0016] It should be noted that minimizing MDT (Multi-Target Data) data from road tests typically employs a periodic reporting method, with a reporting cycle set to 5 to 30 seconds; the sampling interval of professional road test equipment is usually 1 to 5 seconds; and the measurement grid size generated by the network management system is typically 10 meters by 10 meters to 50 meters by 50 meters. Furthermore, within the same monitoring time window, it is necessary to ensure that the measurement data of all beams and sub-bands remain aligned in time to avoid inaccurate calculation results due to time deviations.

[0017] It's important to clarify that logarithmic power is a power representation based on a logarithmic scale. The addition, subtraction, multiplication, and division operations between its values ​​have no physical meaning. For example, if two signals have logarithmic powers of 0 dBmW and 3 dBmW respectively, their corresponding linear powers are approximately 1 mW and 2 mW, respectively, with a linear power ratio of approximately 2. The difference between the two logarithmic power values ​​is 3 dB, which needs to be converted by dividing 10 by 10 to obtain an approximate 2x linear power ratio. Therefore, in subsequent weighted centering calculations, linear power should be used instead of directly using dBm values. Subsequent weighted centering calculations require linear operations on the power values ​​to reflect the link quality differences between different samples. Directly using logarithmic power would lead to calculation results that deviate from the actual physical situation and fail to accurately reflect the location of weak coverage areas. Converting logarithmic power to linear power ensures that the proportional relationship between power values ​​truly reflects the energy ratio of signal and noise, ensuring the physical consistency of subsequent calculation results and thus guaranteeing the reliability and accuracy of the entire network performance monitoring method.

[0018] In one embodiment of the present invention, the specific calculation process of step S2 includes: The calculation process for the power ratio based on the linear signal power and linear noise power is as follows: [The process is repeated here, so the translation ends here.] Beam number, the Frequency band number, number Linear signal power at each sample location As the dividend, the first Beam number, the Frequency band number, number Linear noise power at each sample location As the divisor, the ratio of the dividend to the divisor is calculated to obtain the power ratio. ; in Indicates the first Beam number, the Frequency band number, number The power ratio of each sample location. Indicates the first Beam number, the Frequency band number, number Linear signal power at each sample location point Indicates the first Beam number, the Frequency band number, number Linear noise power at each sample location point Indicates the beam number. Indicates the frequency band number. Indicates the sample location point number; The calculation process for the coverage energy weight based on the power ratio is as follows: [Calculate the coverage energy weight using the power ratio...] Add the results to obtain the weighted denominator, use 1 as the weighted numerator, and calculate the ratio of the weighted numerator to the weighted denominator to obtain the coverage energy weight. ; in Indicates the first Beam number, the Frequency band number, number Coverage energy weights for each sample location point Indicates the first Beam number, the Frequency band number, number The power ratio of each sample location.

[0019] It should be noted that the power ratio represents the ratio of linear signal power to linear noise power, obtained by dividing the linear signal power by the linear noise power. The coverage energy weight represents the degree to which the sample location point is close to a weak coverage or interference-limited state; the larger the weight, the higher the degree of weak coverage, and it is calculated from the power ratio.

[0020] It should be noted that the coverage energy weight ranges from 0 to 1. When the signal is extremely strong and the interference is extremely weak, the power ratio is large, and the coverage energy weight approaches 0, indicating that the link quality at this sample location is very good and it does not belong to a weak coverage area. When the signal is extremely weak or the interference is extremely strong, the power ratio is small, and the coverage energy weight approaches 1, indicating that the link quality at this sample location is very poor and it belongs to a weak coverage area. For example, when the linear signal power is 1 milliwatt and the linear noise power is 1 microwatt, the power ratio is 1000, and the coverage energy weight is approximately 0.001; when the linear signal power is 1 microwatt and the linear noise power is 1 milliwatt, the power ratio is 0.001, and the coverage energy weight is approximately 0.999. It should be noted that the coverage energy weight is continuous and can reflect the subtle differences in link quality between different sample locations, rather than a simple binary judgment.

[0021] It should be noted that when the calculated value of the linear noise power is less than... To avoid division by zero errors when the power is in watts, the minimum value of the linear noise power can be limited to 1. Watts. This value can be used as a preset minimum positive value to avoid calculation instability caused by division by zero or abnormally small values; in actual deployment, this minimum value can be adjusted according to the subband bandwidth, receiver noise figure and measurement system accuracy, which will not be elaborated here.

[0022] It should be further explained that the coverage energy weight is derived from the natural proportional relationship between signal and noise interference power, rather than a manually set service threshold, thus objectively reflecting the actual link quality of the samples. It should also be noted that traditional weak coverage judgment methods typically use a fixed reference signal received power threshold, such as below -110 dB / mW for weak coverage. This method has two main problems: first, it ignores the impact of interference. In cases of high interference (when the average SINR of the cell is below 5 dB, or when neighboring cell interference accounts for more than 60% of the total interference power), even with a high reference signal received power, the user's actual experience will be poor; second, the threshold setting is subjective, and different operators or different regions may use different thresholds, leading to inconsistent judgment results. The coverage energy weight of this invention is calculated based on the actual measured signal power and noise interference power, simultaneously considering the impact of signal strength and interference level, objectively reflecting the actual link quality of each sample location point, and is not affected by manually set thresholds. In other words, it weights and amplifies samples in weak coverage areas, improving the sensitivity of subsequent spatial centroid calculations to performance-degraded areas, and more accurately locating the core position of weak coverage areas.

[0023] In one embodiment of the present invention, the specific calculation process of step S3 includes: Based on the coverage energy weight and the coordinates of the sample location points, the formula for calculating the spatial centroid coordinates is as follows: in Indicates the first Beam number, the The spatial barycenter coordinates corresponding to each frequency band number. Indicates the first Beam number, the Frequency band number, number Coverage energy weights for each sample location point Indicates the first The x-coordinates of each sample location point Indicates the first The ordinates of each sample location point Indicates the beam number. Indicates the frequency band number. Indicates the sample location point number. Indicates participation in the beam number and the The summation of all sample location points calculated from each frequency band index. Indicates the first The transposed coordinates of the coordinates of each sample location point.

[0024] It should be noted that the horizontal weighting value represents the product of the sample location's x-coordinate and the coverage energy weight. The vertical weighting value represents the product of the sample location's y-coordinate and the coverage energy weight. The horizontal weighted sum represents the sum of all horizontal weighting values ​​within the same beam and frequency band. The vertical weighted sum represents the sum of all vertical weighting values ​​within the same beam and frequency band. The weighted sum represents the sum of all coverage energy weights within the same beam and frequency band. The centroid x-coordinate represents the horizontal coordinate of the weak coverage spatial centroid, obtained by dividing the horizontal weighted sum by the weighted sum. The centroid y-coordinate represents the second horizontal coordinate, or vertical coordinate, of the weak coverage spatial centroid in the ground two-dimensional coordinate system, obtained by dividing the vertical weighted sum by the weighted sum. The spatial centroid coordinate represents the ground location that best represents the impact of weak coverage within the same beam and frequency band, composed of the centroid x-coordinate and centroid y-coordinate.

[0025] It should be noted that the coordinates of the sample location points are either local Cartesian coordinates or transformed from the WGS84 (World Geodetic System 1984) coordinate system to ensure consistency of coordinate data from different measurement sources. The local Cartesian coordinate system is a plane coordinate system established based on the local geographical location, suitable for measurements of small areas, and its coordinate values ​​directly reflect distances and positional relationships on the ground. The WGS84 coordinate system is a globally used geodetic coordinate system, widely used in global navigation satellite systems; its coordinate values ​​are latitude and longitude, and require projection transformation to obtain plane coordinates. In practical applications, minimizing drive test MDT data typically uses the WGS84 coordinate system, while professional drive test equipment can be set to either the local Cartesian coordinate system or the WGS84 coordinate system as needed. To ensure that coordinate data from different measurement sources can be used for unified calculations, all coordinates in the WGS84 coordinate system need to be converted to the local Cartesian coordinate system before calculating the spatial centroid. For example, for a certain area, the Gauss-Kruger projection is used to convert the WGS84 latitude and longitude coordinates into plane coordinates. The converted coordinate units are meters, which can be directly used for distance and location calculations.

[0026] It should be noted that when the sum of weights under the same beam and frequency band is less than If the signal is invalid, it indicates that the beam and sub-band do not have any effective weak coverage samples within the current monitoring time window, and the spatial centroid coordinates cannot be calculated. In this case, the subsequent calculation of the beam and sub-band can be skipped, or it can be marked as having no weak coverage.

[0027] It should be further explained that this invention uses coverage energy weights to weight the sample coordinates, rather than an arithmetic mean, so that the calculation results are more concentrated in the areas where network performance degradation is most significant. It should be noted that the arithmetic mean treats all samples with equal weights, and the calculated centroid position is the average of all sample positions, which cannot distinguish the link quality differences between different samples. In actual network coverage, the distribution of samples is usually uneven. In areas with better coverage, there are more users and more samples; in areas with poorer coverage, there are fewer users and fewer samples. If the arithmetic mean is used to calculate the spatial centroid, the calculation results will be biased towards areas with more samples and better coverage, and will not accurately reflect the core location of weak coverage areas. This invention uses coverage energy weights to weight the sample coordinates, with samples with poorer link quality having a larger weight and occupying a larger proportion in the calculation of the spatial centroid. In this way, even if the number of samples in weak coverage areas is small, it can still have a significant impact on the location of the spatial centroid, making the calculated spatial centroid coordinates more concentrated in the areas where network performance degradation is most significant, thus providing reliable spatial data for subsequent frequency drift analysis.

[0028] In one embodiment of the present invention, the specific calculation process of step S4 includes: The formula for calculating the average frequency value based on the center frequency and the number of frequency bands is as follows: in This represents the average frequency value. This indicates the number of frequency bands involved in the calculation. Indicates the first Each frequency band number corresponds to the center frequency of that frequency band. Indicates the frequency band number; The formula for calculating the average coordinate value based on the spatial centroid coordinates and the number of frequency bands is as follows: in Indicates the first The average coordinate value corresponding to each beam number Indicates the first Beam number, the The spatial barycenter coordinates corresponding to each frequency band number. This indicates the number of frequency bands involved in the calculation. Indicates the beam number. Indicates the frequency band number; The formula for calculating the frequency spatial slope is as follows, based on the frequency band center frequency, spatial centroid coordinates, average frequency value, and average coordinate value: in Indicates the first The frequency space slope corresponding to each beam number Indicates the first The frequency difference between the center frequency and the average frequency of each frequency band Indicates the first Beam number, the The coordinate difference between the spatial barycenter coordinates and the average coordinates corresponding to each frequency band number. This indicates the sequence number of the first frequency band up to the second. Sum of frequency band indices Indicates the beam number. Indicates the frequency band number. This indicates the number of frequency bands involved in the calculation.

[0029] It should be noted that the average frequency value represents the arithmetic mean of the center frequencies of all frequency bands involved in the calculation, obtained by summing the center frequencies of all frequency bands and dividing by the number of frequency bands. The average coordinate value represents the arithmetic mean of the spatial barycenter coordinates corresponding to all frequency bands under the same beam, obtained by summing the spatial barycenter coordinates of all frequency bands under the same beam and dividing by the number of frequency bands. The frequency difference value represents the difference between the center frequency of a single frequency band and the average frequency value. The coordinate difference value represents the difference between the spatial barycenter coordinates corresponding to a single frequency band and the average coordinate value. The slope product represents the product of the frequency difference and the coordinate difference. The slope numerator represents the sum of all slope products. The frequency square value represents the square of the frequency difference. The slope denominator represents the sum of all frequency square values. The frequency spatial slope represents the two-dimensional rate of change of the weak coverage barycenter with frequency; its direction represents the drift direction, and its magnitude represents the displacement caused by a unit frequency change, obtained by dividing the slope numerator by the slope denominator.

[0030] It should be noted that the number of frequency bands can be set according to the monitoring bandwidth and frequency resolution requirements, generally not less than three, to ensure the effectiveness and accuracy of linear fitting. It should also be noted that the selection of the number of frequency bands needs to balance frequency resolution and computational efficiency. When the monitoring bandwidth is wide, the number of frequency bands can be increased to improve frequency resolution and more accurately reflect the trend of weak coverage centroid changing with frequency. When the monitoring bandwidth is narrow, the number of frequency bands can be appropriately reduced, but not less than three, otherwise effective linear fitting cannot be performed. For example, for a 100 MHz monitoring bandwidth, it can be divided into 10 10 MHz sub-bands, with a frequency band count of 10; for a 20 MHz monitoring bandwidth, it can be divided into 4 5 MHz sub-bands, with a frequency band count of 4. If the number of effective frequency bands is less than three, for example, only two, although the slope can be calculated from the two spatial centroid coordinates, it is impossible to reliably evaluate the linear trend and goodness of fit of the weak coverage centroid changing with frequency. Therefore, this embodiment preferably requires that the number of effective frequency bands be no less than three. If there are too many frequency bands (for example, when the subband width is less than 1 MHz, or the number of frequency bands exceeds 20), such as dividing a 100 MHz monitoring bandwidth into 100 1 MHz subbands, it will greatly increase the amount of calculation, reduce monitoring efficiency, and also increase the impact of measurement errors on the calculation results.

[0031] It should be noted that when calculating the frequency space slope, subbands without valid spatial centroid coordinates must be skipped. If the number of valid subbands is less than three, linear fitting cannot be performed, and the beam is marked as unable to calculate the frequency space slope. Furthermore, the goodness of linear fitting can be evaluated by calculating the coefficient of determination. When the coefficient of determination is less than 0.7, it indicates that the linear relationship between the weak coverage centroid and frequency is not significant. In this case, the calculated frequency space slope has low reliability (for example, the calculation error of the frequency space slope may exceed 30%, making it unusable for accurately inverting the structural tilt angle), and can be marked as a poor fitting state, which will not be elaborated further here.

[0032] It should be further explained that this invention employs a centered least squares method to calculate the frequency space slope, avoiding reliance on the measurement error of any single sub-band and improving the robustness and stability of drift trend calculation. The centered least squares method, based on the least squares method, subtracts the average value from all data points before performing linear fitting. This method eliminates the influence of data offset on the fitting results, making the fitting results more stable. In actual network measurements, the measurement data of each sub-band may contain certain errors. For example, due to the influence of multipath effects, shadow fading, and other factors, the spatial centroid coordinates of a certain sub-band may have a large deviation. Using the centered least squares method can eliminate the influence of mean offset on slope calculation, improving the numerical stability of slope calculation. For individual abnormal sub-band measurement points, they can be removed or marked before fitting through outlier filtering, goodness-of-fit judgment, or robust fitting methods to reduce the influence of abnormal data points on the frequency space slope.

[0033] In one embodiment of the present invention, the specific calculation process of step S5 includes: The formula for calculating the azimuth unit coordinates based on the beam azimuth angle is as follows: in Indicates the first The azimuth coordinates corresponding to each beam number. Indicates the first The beam azimuth angle corresponding to each beam number Indicates the beam number. Indicates coordinate transpose; The formula for calculating the squint geometry value based on the beam elevation angle is as follows: in Indicates the first The squint geometry corresponding to each beam number Indicates the first The beam elevation angle corresponding to each beam number Indicates the beam number; The formula for calculating the deflection geometry value based on the beam pitch angle is as follows: in Indicates the first The deflection geometry value corresponding to each beam number. Indicates the first The beam elevation angle corresponding to each beam number Indicates the beam number; Based on the frequency space slope, azimuth unit coordinates, reference frequency value, antenna height value, oblique geometry value, and deflection geometry value, the formula for calculating the structural tilt angle is as follows: in Indicates the first The structural tilt angle corresponding to each beam number. Indicates the reference frequency value. Indicates the first The frequency space slope corresponding to each beam number Indicates the first The azimuth coordinates corresponding to each beam number. This represents the projection slope of the frequency space slope onto the beam azimuth direction. Indicates the first The antenna height value corresponding to each beam number. Indicates the first The squint geometry corresponding to each beam number Indicates the first The deflection geometry value corresponding to each beam number. Indicates the beam number.

[0034] It should be noted that the azimuth unit coordinates represent the unit vector in the beam azimuth direction, composed of the cosine and sine values ​​of the beam azimuth. The projection slope represents the projection of the frequency space slope onto the beam azimuth direction, obtained by multiplying the frequency space slope by the azimuth unit coordinates. The squint geometry value represents the geometric contribution of pure broadband beam squint to ground coverage drift, calculated using a trigonometric function of the beam elevation angle. The deflection factor represents the amplification factor of the structural tilt contribution to coverage drift, calculated using a trigonometric function of the beam elevation angle. The deflection geometry value represents the geometric coefficient of the combined effect of structural tilt and beam elevation angle, calculated using a trigonometric function of the beam elevation angle. The reference frequency value represents the center operating frequency of the monitored beam, a base station configuration parameter obtainable through the base station network management system. The tilt numerator represents the intermediate value used to calculate the structural tilt angle. The structural tilt angle represents the slight tilt of the antenna panel or equivalent array in that beam direction relative to the beam pointing direction, obtained by dividing the tilt numerator by the deflection geometry value.

[0035] It should be noted that the reference frequency value is directly obtained from the base station configuration parameters and is the nominal center operating frequency of the monitored beam. The reference frequency value is the fundamental parameter for calculating the impact of broadband beam squint and antenna structure tilt on coverage drift, and it must be consistent with the actual operating frequency of the beam; otherwise, it will lead to significant deviations in the calculation results. For example, for a 5G beam operating in the 3.5 GHz band, its nominal center operating frequency is 3500 MHz, and the reference frequency value should be set to 3500 MHz. If the reference frequency value is set to 3400 MHz, the calculated beam squint and structure tilt angles will be incorrect and cannot accurately reflect the actual state of the antenna. In practical applications, the base station's operating frequency may be dynamically adjusted according to network load and interference conditions. Therefore, when monitoring network performance, it is necessary to obtain the current operating frequency of the base station in real time to ensure the accuracy of the reference frequency value.

[0036] It should be noted that the reasonable range for the beam pitch angle is 0 to 60 degrees, which covers the commonly used downtilt angles of 5G base station beams. When the beam pitch angle exceeds this range, the calculation model of this invention may produce a large error, and it should be corrected according to the actual situation.

[0037] It should be further explained that this invention establishes a quantitative correlation between network performance measurement results and the physical structure state of the antenna by inverting the structural tilt angle using frequency space slope, thereby pinpointing the structural causes of performance degradation. Traditional network performance monitoring methods can only detect network coverage problems but cannot determine the specific causes. For example, when a low reference signal reception power is found in a certain area, possible causes include insufficient transmit power, excessive inter-site spacing, severe obstruction, and antenna pointing deviation. Maintenance personnel need to conduct extensive on-site testing and troubleshooting to determine the cause of the coverage problem, which is inefficient. This invention is based on the flexural amplification of frequency slant coverage drift characteristics (referring to the slight physical flexure of the antenna panel of a 5G multi-port phased array antenna under external forces (wind load, solar temperature difference) in the actual deployment environment; this slight flexure amplifies the inherent frequency slant effect of the broadband beam, that is, within the same beam, there is an inherent difference in beam pointing between the low-frequency sub-band and the high-frequency sub-band; ultimately causing the weak coverage areas corresponding to different frequency sub-bands to undergo reverse spatial drift along a specific direction). By measuring the frequency spatial slope and inverting the structural tilt angle, the invention can directly correlate network performance degradation with the physical structure state of the antenna; for example, when the calculated structure... When the structural tilt angle is large (e.g., when the structural tilt angle exceeds 0.05 radians), it indicates that there is significant micro-deflection of the antenna panel, causing beam pointing deviation, requiring antenna calibration or maintenance. When the structural tilt angle is small (e.g., when the structural tilt angle is less than 0.02 radians), it indicates that the coverage drift is mainly caused by broadband beam squint, which is a normal phenomenon and does not require additional processing. In other words, quantitatively distinguishing between pure beam squint and coverage drift caused by antenna structural deformation provides a direct basis for antenna status diagnosis, helps maintenance personnel quickly determine the cause of coverage problems, and improves network maintenance efficiency.

[0038] In one embodiment of the present invention, the specific calculation process of step S6 includes: Based on the average coordinate value, frequency space slope, reference frequency value, and average frequency value, the formula for calculating the reference centroid coordinates is as follows: in Indicates the first The reference barycenter coordinates corresponding to each beam number Indicates the first The average coordinate value corresponding to each beam number Indicates the first The frequency space slope corresponding to each beam number Indicates the reference frequency value. This represents the average frequency value. Indicates the beam number; Based on the reference centroid coordinates, frequency space slope, minimum boundary frequency, and reference frequency value, the formula for calculating the coordinates of the first endpoint is as follows: in Indicates the first The coordinates of the first endpoint corresponding to each beam number Indicates the first The reference barycenter coordinates corresponding to each beam number Indicates the first The frequency space slope corresponding to each beam number Indicates the minimum boundary frequency. Indicates the reference frequency value. Indicates the beam number; Based on the reference centroid coordinates, frequency space slope, maximum boundary frequency, and reference frequency value, the formula for calculating the coordinates of the second endpoint is as follows: in Indicates the first The coordinates of the second endpoint corresponding to each beam number Indicates the first The reference barycenter coordinates corresponding to each beam number Indicates the first The frequency space slope corresponding to each beam number Indicates the maximum boundary frequency. Indicates the reference frequency value. Indicates the beam number.

[0039] It should be noted that: The reference frequency difference represents the difference between the reference frequency value and the average frequency value. The reference coordinate difference represents the product of the reference frequency difference and the frequency space slope. The reference centroid coordinates represent the weak coverage reference position at the center frequency, obtained by adding the average coordinate value and the reference coordinate difference. The first boundary frequency difference represents the difference between the minimum boundary frequency and the reference frequency value. The first endpoint difference represents the product of the first boundary frequency difference and the frequency space slope. The first endpoint coordinates represent the weak coverage position corresponding to the lowest sub-band, obtained by adding the reference centroid coordinates and the first endpoint difference. The second boundary frequency difference represents the difference between the maximum boundary frequency and the reference frequency value. The second endpoint difference represents the product of the second boundary frequency difference and the frequency space slope. The second endpoint coordinates represent the weak coverage position corresponding to the highest sub-band, obtained by adding the reference centroid coordinates and the second endpoint difference.

[0040] It should be noted that the minimum boundary frequency is the lower boundary frequency of the monitoring bandwidth, and the maximum boundary frequency is the upper boundary frequency of the monitoring bandwidth, both directly obtained from the base station configuration parameters. The minimum and maximum boundary frequencies determine the range of the monitoring bandwidth and are fundamental parameters for calculating the endpoints of weak coverage drift segments. For example, in a scenario with a center frequency of 3500 MHz and a monitoring bandwidth of 100 MHz, the lower boundary frequency of the monitoring bandwidth is 3450 MHz, and the upper boundary frequency is 3550 MHz. If this 100 MHz bandwidth is divided into 10 sub-bands of 10 MHz each, then the lowest sub-band center frequency is 3455 MHz, and the highest sub-band center frequency is 3545 MHz. In practical applications, the range of the monitoring bandwidth may be adjusted according to network configuration and service requirements. Therefore, when monitoring network performance, it is necessary to obtain the current monitoring bandwidth and sub-band division of the base station in real time to ensure the accuracy of the minimum and maximum boundary frequencies.

[0041] It should be noted that when the magnitude of the frequency space slope is less than At a frequency of 0.1 m / MHz, the weak coverage center of gravity hardly shifts with frequency. In this case, the length of the weak coverage drift segment approaches zero, which can be marked as a state with no significant coverage drift. For example, when the frequency space slope is 0.1 m / MHz, and the upper and lower boundary frequencies of the monitoring bandwidth are used as endpoint frequencies, with a frequency span of 100 MHz, the length of the weak coverage drift segment is 10 meters. If the lowest and highest sub-band center frequencies are used as endpoint frequencies, and the 100 MHz bandwidth is divided into 10 10 MHz sub-bands, then the center frequency span is 90 MHz, corresponding to a segment length of 9 meters.

[0042] It should be further explained that this invention transforms the abstract frequency drift trend into physical line segments on the ground, intuitively presenting the spatial distribution range, length, and drift direction of weak coverage bands. It should be noted that the frequency space slope only represents the rate of change of the weak coverage center of gravity with frequency, and cannot intuitively present the actual distribution of weak coverage bands on the ground. This invention generates physical weak coverage drift line segments by calculating the weak coverage positions corresponding to the lowest and highest sub-bands. This line segment is a concrete spatial entity with clearly defined start and end positions, length, and direction, intuitively presenting the distribution of weak coverage bands on the ground; for example, when the frequency space slope is 0.1 meters per MHz and the monitoring bandwidth is 100 MHz, the length of the weak coverage drift line segment is 10 meters, and its direction is consistent with the direction of the frequency space slope. Maintenance personnel can quickly determine the location and range of weak coverage bands based on this line segment and take corresponding optimization measures.

[0043] In one embodiment of the present invention, the specific calculation process of step S7 includes: The formula for calculating the length of the physical line segment based on the coordinates of the first and second endpoints is as follows: in Indicates the first The physical line segment length corresponding to each beam number Indicates the first The coordinates of the second endpoint corresponding to each beam number Indicates the first The coordinates of the first endpoint corresponding to each beam number The norm of the two-dimensional coordinate difference. Indicates the beam number; The formula for calculating the pointing deviation angle, based on the physical line segment length, antenna height, and structural tilt angle, is as follows: in Indicates the first The pointing deviation angle corresponding to each beam number Indicates the first The physical line segment length corresponding to each beam number Indicates the first The antenna height value corresponding to each beam number. Indicates the first The structural tilt angle corresponding to each beam number. Represents the absolute value of the structural tilt angle. Indicates the beam number.

[0044] It should be noted that the endpoint coordinate difference represents the difference between the coordinates of the second endpoint and the first endpoint. The lateral square value represents the square of the lateral component of the endpoint coordinate difference. The longitudinal square value represents the square of the longitudinal component of the endpoint coordinate difference. The physical segment length represents the ground length of the weak coverage drift segment, obtained by taking the square root of the sum of the lateral and longitudinal square values. The height multiple represents twice the antenna height. The angle ratio represents the ratio of the physical segment length to the height multiple. The segment angle represents the half-angle deviation projected from the physical segment length onto the antenna height, obtained by the arctangent of the angle ratio. The absolute tilt value represents the absolute value of the structural tilt angle. The pointing deviation angle represents the equivalent beam pointing deviation that combines the frequency span and structural tilt, obtained by adding the segment angle and the absolute tilt value.

[0045] It should be noted that monitoring results can be output as text reports, GIS (Geographic Information System) map annotations, or database records, including the start and end coordinates, length, and equivalent pointing deviation angle of the weak coverage drift segment, facilitating subsequent network optimization and antenna maintenance. Different output formats are suitable for different application scenarios. For example, text reports contain detailed monitoring data and analysis results, suitable for archiving and written reporting; GIS map annotations overlay weak coverage drift segments onto an electronic map, visually displaying their geographical location and extent, suitable for on-site operation and maintenance and network planning; database records store monitoring results in a database for subsequent querying, statistics, and analysis. Monitoring results can be generated periodically, such as daily reports, weekly reports, and monthly reports, for long-term network performance analysis and trend prediction; they can also be generated in real time when abnormal coverage drift is detected, such as when the length of a weak coverage drift segment exceeds a preset threshold (e.g., when the length of the weak coverage drift segment exceeds 20 meters), immediately generating alarm information and a detailed monitoring report to notify maintenance personnel for timely handling.

[0046] It should be noted that, in one implementation, the equivalent pointing deviation angle threshold can be set to 0.05 radians; when the pointing deviation angle exceeds this threshold, it is marked as a suspected abnormal coverage drift state, and an on-site inspection or antenna calibration process is triggered. This threshold can be adjusted according to the site type, frequency band, beamwidth, coverage target, and operator maintenance strategy, which will not be elaborated here.

[0047] It should be further explained that the pointing deviation is calculated by combining the physical segment length and the structural tilt angle to comprehensively reflect the overall impact of antenna structure and frequency coupling on beam pointing. It should be noted that beam pointing deviation is mainly caused by two factors: first, broadband beam squint, where different frequency sub-bands have different beam pointing, causing weak coverage bands to drift with frequency; second, antenna structural deformation, where the antenna panel undergoes slight deflection under the influence of wind load, solar radiation, and temperature differences, resulting in an overall shift in beam pointing. The physical segment length reflects the beam pointing deviation caused by broadband beam squint, while the structural tilt angle reflects the beam pointing deviation caused by antenna structural deformation. Using either parameter alone cannot fully reflect the overall beam pointing deviation; for example, when the physical segment length is 10 meters and the antenna height is 30 meters, the half-angle deviation caused by broadband beam squint is approximately 0.167 radians; when the structural tilt angle is 0.1 radians, the beam pointing deviation caused by antenna structural deformation is 0.1 radians. This invention combines the two to calculate the pointing deviation angle, which can comprehensively reflect the overall impact of antenna structure and frequency coupling on beam pointing, providing an accurate basis for antenna calibration and network optimization.

[0048] In one embodiment of the present invention, such as Figure 2As shown, a network performance monitoring system based on a 5G antenna includes: The data acquisition module is used to collect the logarithmic power of the signal and the logarithmic power of the noise according to the beam number, frequency band number and sample location point within the monitoring time window, and convert the logarithmic power of the signal and the logarithmic power of the noise into linear signal power and linear noise power. The coverage energy weight calculation module is used to divide the linear signal power by the linear noise power to generate a power ratio, and to calculate the coverage energy weight based on the power ratio. The spatial barycenter coordinate calculation module is used to perform coordinate weighted calculation of sample location points by covering energy weight, and generate the spatial barycenter coordinates corresponding to each frequency band number. The frequency space slope calculation module is used to calculate the average frequency value by calculating the average center frequency of the frequency band, the average coordinate value by calculating the average centroid coordinates of the space, and to calculate the frequency space slope based on the center frequency of the frequency band, the centroid coordinates of the space, the average frequency value, and the average coordinate value. The structural tilt angle calculation module is used to project the frequency space slope onto the beam azimuth direction and calculate the structural tilt angle by combining the reference frequency value, beam azimuth angle, beam elevation angle and antenna height value. The endpoint coordinate calculation module is used to calculate the reference centroid coordinates using the frequency space slope and the reference frequency value, and to calculate the first endpoint coordinates and the second endpoint coordinates based on the frequency space slope, the reference centroid coordinates, the minimum boundary frequency, and the maximum boundary frequency. The pointing deviation angle calculation module is used to calculate the physical line segment length between the coordinates of the first endpoint and the coordinates of the second endpoint. It combines the physical line segment length, antenna height value and structural tilt angle to calculate the pointing deviation angle and output it as the monitoring result.

[0049] Specifically, this invention can be deployed in the network management system of a 5G network as a functional module for network performance monitoring. During deployment, the computational program of this invention needs to be integrated into the data analysis platform of the network management system, and an interface with the base station network management system needs to be configured to obtain base station configuration parameters and measurement data. The computational program can be deployed in a distributed architecture on a server cluster to support parallel monitoring of a large number of base stations.

[0050] In actual operation, the system periodically obtains beam configuration parameters, subband division information, antenna height and measurement sample data from the base station network management system according to the preset monitoring time window. The measurement sample data includes signal logarithmic power, noise logarithmic power and sample location point coordinates. The system preprocesses the acquired data, including coordinate transformation, data alignment and outlier filtering, and then performs calculations in sequence according to the steps (modules) of this invention to finally generate the monitoring results for each beam.

[0051] For example, for a 5G macro base station operating in the 3.5 GHz band, with a monitoring bandwidth of 100 MHz divided into 10 10 MHz sub-bands, an antenna height of 30 meters, an azimuth angle of 90 degrees, and an elevation angle of 10 degrees for a certain beam, the system acquires 1000 valid measurement samples within a 1-hour monitoring window. Calculations show that the frequency space slope vector of this beam is (0.120, -0.0023) meters per MHz, its mode length is approximately 0.120 meters per MHz, and its projection in the beam azimuth direction is approximately -0.0023 meters per MHz. Based on a reference frequency of 3500 MHz, an antenna height of 30 meters, and a beam elevation angle of 10 degrees, the structural tilt angle is calculated to be approximately 0.08 radians. Using the upper and lower boundary frequencies of the monitoring bandwidth as endpoint frequencies, the start and end coordinates of the weak coverage drift segment are (500, 300) and (512.0, 299.77), respectively, with a length of approximately 12.0 meters and an equivalent pointing deviation angle of approximately 0.28 radians. The system outputs these results in the form of GIS map annotations, allowing maintenance personnel to visually see the location and range of the weak coverage drift segment on an electronic map and determine the need for antenna calibration based on the equivalent pointing deviation angle.

[0052] It should be noted that the interval and threshold sizes are set for ease of comparison. The size of the threshold depends on the amount of sample data and the base number set by those skilled in the art for each set of sample data, as long as it does not affect the proportional relationship between the parameter and the quantized value. Furthermore, the above formulas are all dimensionless calculations, and the formulas are derived from software simulations using a large amount of collected data to obtain a formula that is closest to the real situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0053] The content of this embodiment has been described above, but this embodiment is not limited to the specific implementation methods described above. The specific implementation methods described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms based on the guidance of this embodiment, all of which are within the protection scope of this embodiment.

Claims

1. A network performance monitoring method based on 5G antennas, characterized in that, Includes the following steps: Step S1: Within the monitoring time window, collect the logarithmic power of the signal and the logarithmic power of the noise according to the beam number, frequency band number and sample location point, and convert the logarithmic power of the signal and the logarithmic power of the noise into linear signal power and linear noise power. Step S2: Divide the linear signal power by the linear noise power to generate a power ratio, and calculate the coverage energy weight based on the power ratio; Step S3: The sample location points are weighted by the coverage energy weight to generate the spatial barycenter coordinates corresponding to each frequency band number. Step S4: Calculate the average frequency value by calculating the average center frequency of the frequency band, calculate the average coordinate value by calculating the average centroid coordinates of the space, and calculate the frequency space slope based on the frequency band center frequency, the centroid coordinates of the space, the average frequency value, and the average coordinate value. Step S5: Project the frequency space slope onto the beam azimuth direction, and calculate the structure tilt angle by combining the reference frequency value, beam azimuth angle, beam elevation angle, and antenna height value, including: Step S501: Generate azimuth unit coordinates based on beam azimuth angle, and multiply the frequency space slope with the azimuth unit coordinates to obtain the projection slope. Step S502: Calculate the cosine value of the beam elevation angle, multiply the cosine value by itself and take the reciprocal, then multiply the reciprocal by the tangent value of the beam elevation angle to obtain the oblique look-out geometric value. Step S503: Multiply the beam elevation tangent by itself, then multiply by three, and add the product to one to obtain the deflection ratio value; Step S504: Multiply the reciprocal of the cosine value of the beam elevation angle by itself and the deflection ratio value to obtain the deflection geometry value; Step S505: Multiply the reference frequency value by the projection slope, divide the product by the antenna height value and take the opposite number, then subtract the oblique geometry value from the opposite number to obtain the tilt numerator. Step S506: Divide the tilted molecule by the flexural geometry value to obtain the structural tilt angle; Step S6: Calculate the reference centroid coordinates using the frequency space slope and the reference frequency value, and calculate the coordinates of the first endpoint and the second endpoint based on the frequency space slope, the reference centroid coordinates, the minimum boundary frequency, and the maximum boundary frequency. Step S7: Calculate the physical line segment length between the coordinates of the first endpoint and the second endpoint. Combine the physical line segment length, antenna height, and structural tilt angle to calculate the pointing deviation angle and output it as the monitoring result, including: Step S801: Multiply the antenna height value by two to obtain the height multiple, and divide the physical line segment length by the height multiple to obtain the angle ratio. Step S802: Calculate the arctangent of the angle ratio to obtain the line segment angle; Step S803: Take the absolute value of the structural tilt angle to obtain the absolute value of the tilt. Step S804: Add the line segment angle to the absolute value of the tilt to obtain the pointing deviation angle, and output the coordinates of the first endpoint, the coordinates of the second endpoint, the physical line segment length, and the pointing deviation angle as monitoring results.

2. The network performance monitoring method based on a 5G antenna according to claim 1, characterized in that, Within the monitoring time window, the logarithmic power of the signal and the logarithmic power of the noise are collected based on the beam number, frequency band number, and sample location points. The logarithmic power of the signal and the logarithmic power of the noise are then converted into linear signal power and linear noise power, specifically including the following steps: Step S101: Within the same monitoring time window, read the signal logarithmic power, noise logarithmic power, frequency band center frequency, beam azimuth angle, beam elevation angle, antenna height value, and sample location point coordinates according to the beam number, frequency band number, and sample location point. Step S102: Divide the signal logarithmic power by the 30 dBmW conversion constant, and divide the difference by 10 to obtain the signal exponent value; Step S103: Using ten as the base and the signal exponent value as the exponent, obtain the linear signal power; Step S104: Subtract the noise logarithmic power from the 30 dBmW conversion constant, and divide the difference by 10 to obtain the noise index value; Step S105: Using ten as the base and the noise index value as the exponent, obtain the linear noise power.

3. The network performance monitoring method based on a 5G antenna according to claim 1, characterized in that, The power ratio is generated by dividing the linear signal power by the linear noise power, and the coverage energy weight is calculated based on the power ratio. The specific steps include: Step S201: Using the linear signal power as the dividend and the linear noise power as the divisor, calculate the ratio of the dividend to the divisor to obtain the power ratio. Step S202: Add the power ratio to obtain the weighted denominator; Step S203: Take one as the weight numerator, calculate the ratio of the weight numerator to the weight denominator, and obtain the coverage energy weight.

4. The network performance monitoring method based on a 5G antenna according to claim 1, characterized in that, By weighting the sample location points using coverage energy, the spatial barycenter coordinates corresponding to each frequency band number are generated through coordinate weighting calculation. The specific steps include: Step S301: For the same beam number and the same frequency band number, multiply the horizontal coordinate of each sample location point by the coverage energy weight to obtain the horizontal weighting value, and multiply the vertical coordinate of each sample location point by the coverage energy weight to obtain the vertical weighting value. Step S302: Sum all horizontally weighted values ​​to obtain a horizontally weighted sum, and sum all vertically weighted values ​​to obtain a vertically weighted sum; Step S303: Add up all the coverage energy weights to obtain the weight sum value; Step S304: Divide the horizontal weighted sum by the weighted sum to obtain the horizontal coordinate of the centroid, and divide the vertical weighted sum by the weighted sum to obtain the vertical coordinate of the centroid. Step S305: Combine the horizontal and vertical coordinates of the centroid to form the spatial centroid coordinates.

5. The network performance monitoring method based on a 5G antenna according to claim 1, characterized in that, The average frequency value is obtained by calculating the average center frequency of the frequency band, and the average coordinate value is obtained by calculating the average centroid coordinates of the space. The frequency space slope is then calculated based on the center frequency of the frequency band, the centroid coordinates of the space, the average frequency value, and the average coordinate value. The specific steps include: Step S401: Add the center frequencies of all frequency bands involved in the calculation, and divide the sum by the number of frequency bands to obtain the average frequency value; Step S402: Add up all the spatial barycenter coordinates corresponding to the same beam number, and divide the sum by the number of frequency bands to obtain the average coordinate value; Step S403: Subtract the center frequency of each frequency band from the average frequency value to obtain the frequency difference value, and subtract the centroid coordinate of each space from the average coordinate value to obtain the coordinate difference value. Step S404: Multiply the frequency difference by the coordinate difference to obtain the slope product, and add all the slope products together to obtain the slope numerator; Step S405: Multiply the frequency difference by itself to obtain the frequency square value, and add up all the frequency square values ​​to obtain the slope denominator; Step S406: Divide the slope numerator by the slope denominator to obtain the frequency space slope.

6. The network performance monitoring method based on a 5G antenna according to claim 1, characterized in that, The reference centroid coordinates are calculated using the frequency space slope and the reference frequency value. The coordinates of the first and second endpoints are then calculated based on the frequency space slope, the reference centroid coordinates, the minimum boundary frequency, and the maximum boundary frequency. The specific steps include: Step S601: Subtract the reference frequency value from the average frequency value to obtain the reference frequency difference; Step S602: Multiply the reference frequency difference by the frequency space slope to obtain the reference coordinate difference, and add the average coordinate value to the reference coordinate difference to obtain the reference centroid coordinates; Step S603: Subtract the minimum boundary frequency from the reference frequency value to obtain the first boundary frequency difference; Step S604: Multiply the first boundary frequency difference by the frequency space slope to obtain the first endpoint difference, and add the reference centroid coordinates to the first endpoint difference to obtain the first endpoint coordinates; Step S605: Subtract the maximum boundary frequency from the reference frequency value to obtain the second boundary frequency difference; Step S606: Multiply the second boundary frequency difference by the frequency space slope to obtain the second endpoint difference, and add the reference centroid coordinates to the second endpoint difference to obtain the second endpoint coordinates.

7. The network performance monitoring method based on a 5G antenna according to claim 1, characterized in that, Calculating the physical line segment length between the coordinates of the first endpoint and the coordinates of the second endpoint involves the following steps: Step S701: Subtract the coordinates of the second endpoint from the coordinates of the first endpoint to obtain the endpoint coordinate difference; Step S702: Multiply the horizontal difference in the endpoint coordinate difference by itself to obtain the horizontal square value, multiply the vertical difference in the endpoint coordinate difference by itself to obtain the vertical square value, add the horizontal square value and the vertical square value and take the square root to obtain the physical line segment length.

8. A network performance monitoring system based on a 5G antenna, characterized in that, The method for monitoring network performance based on a 5G antenna as described in any one of claims 1 to 7 includes: The data acquisition module collects the logarithmic power of the signal and the logarithmic power of the noise based on the beam number, frequency band number and sample location points within the monitoring time window, and converts the logarithmic power of the signal and the logarithmic power of the noise into linear signal power and linear noise power. The coverage energy weight calculation module divides the linear signal power by the linear noise power to generate a power ratio, and calculates the coverage energy weight based on the power ratio. The spatial barycenter coordinate calculation module calculates the spatial barycenter coordinates corresponding to each frequency band number by weighting the coordinates of the sample location points through the coverage energy weight. The frequency space slope calculation module calculates the average frequency value by calculating the average center frequency of the frequency band, and the average coordinate value by calculating the average centroid coordinates of the space. It then calculates the frequency space slope based on the center frequency of the frequency band, the centroid coordinates of the space, the average frequency value, and the average coordinate value. The structural tilt angle calculation module projects the frequency space slope onto the beam azimuth direction and calculates the structural tilt angle by combining the reference frequency value, beam azimuth angle, beam elevation angle and antenna height value. The endpoint coordinate calculation module calculates the reference centroid coordinates using the frequency space slope and the reference frequency value, and calculates the first endpoint coordinates and the second endpoint coordinates based on the frequency space slope, the reference centroid coordinates, the minimum boundary frequency, and the maximum boundary frequency. The pointing deviation angle calculation module calculates the physical line segment length between the coordinates of the first endpoint and the coordinates of the second endpoint, and calculates the pointing deviation angle by combining the physical line segment length, antenna height value and structural tilt angle, and outputs it as the monitoring result.

Images