A 5G base station signal dynamic optimization method and system

Abstract

The application relates to the technical field of wireless communication, in particular to a 5G base station signal dynamic optimization method and system. The method comprises the following steps: acquiring multi-dimensional running state data of each radio frequency channel of a base station; performing correlation analysis on the multi-dimensional running state data based on a preset combination judgment rule; when the correlation analysis simultaneously satisfies the combination condition, determining that the radio frequency channel has suspected reciprocity failure; performing a controlled power down test on the radio frequency channel determined as the suspected reciprocity failure, monitoring the index change trend in real time, and distinguishing the fault root cause according to a preset response relationship; if it is determined that the fault root cause is temperature-induced reciprocity failure, injecting two downlink probe signals to obtain a real phase offset value, and generating a temperature phase compensation corresponding relationship table; matching the phase compensation value in real time, and outputting an optimized beamforming result. The method has the advantages that the uplink and downlink channel reciprocity failure can be effectively identified and solved, and the beamforming precision and user service quality are significantly improved.

Description

Technical Field

[0001] This application relates to the field of wireless communication technology, and more specifically, to a method and system for dynamic optimization of 5G base station signals. Background Technology

[0002] In modern 5G base station TDD systems, beamforming technology is crucial for improving wireless resource utilization and user experience. This technology concentrates energy on the target user by precisely adjusting the signal transmission direction and shape of the antenna array, thereby increasing signal strength and reducing interference to other users. Beamforming typically relies on the uplink-downlink channel reciprocity principle, which estimates the channel state using the received uplink signal and optimizes the downlink beam accordingly. However, in actual operation, the electrical characteristics of the base station's internal hardware can dynamically change due to environmental factors. Especially under specific service modes, localized heat accumulation and potential hardware defects can work together to invalidate the uplink-downlink channel reciprocity assumption, thus affecting beamforming accuracy and ultimately degrading the quality of communication services.

[0003] To address the aforementioned issues, existing technologies urgently need improvement. Summary of the Invention

[0004] The purpose of this application is to provide a method and system for dynamic optimization of 5G base station signals, which has the advantages of effectively identifying and solving the problem of uplink and downlink channel reciprocity failure caused by temperature or hardware failure in 5G base stations, and significantly improving beamforming accuracy and user service quality.

[0005] This application provides a method for dynamic optimization of 5G base station signals, including:

[0006] Acquire multi-dimensional operational status data for each radio frequency channel of the base station. The multi-dimensional operational status data should include at least the service load intensity, power amplifier output power, local temperature change rate, and user service quality indicators.

[0007] Based on preset combination judgment rules, correlation analysis is performed on multi-dimensional operation status data;

[0008] When the correlation analysis simultaneously meets the combined conditions of high service load, high power output, continuous rise in local temperature and decline in user service quality, the radio frequency channel is determined to have a suspected reciprocity failure.

[0009] For radio frequency channels suspected of reciprocity failure, a controlled power downsampling test is performed. Local temperature change trends, uplink and downlink phase offset change trends, and user service quality index change trends are monitored in real time. The root cause of the failure is identified according to the preset response relationship.

[0010] If the root cause of the fault is determined to be reciprocity failure caused by temperature, two downlink detection signals are injected and the corresponding composite phase offset is obtained by loopback measurement. Based on the two composite phase offsets, the power amplifier memory effect component is removed to obtain the true phase offset value. The current local temperature and the true phase offset value are recorded in the temperature phase compensation correspondence table.

[0011] The system reads the current local temperature in real time, obtains the matching phase compensation value by interpolation or table lookup based on the temperature phase compensation correspondence table, applies the phase compensation value to the calculation of downlink beamforming weight, and outputs the optimized beamforming result.

[0012] If the root cause of the fault is determined to be a hardware failure not caused by temperature factors, skip phase compensation and output a targeted maintenance alarm.

[0013] The above solution can dynamically identify and distinguish the root causes of reciprocity failures (temperature or hardware failures) in base station radio frequency channels, and perform targeted phase compensation or alarms, effectively improving beamforming accuracy and user service quality, and solving the problem that existing technologies cannot effectively diagnose and compensate for dynamic reciprocity failures.

[0014] Furthermore, the root causes of failures are distinguished according to preset response relationships, including:

[0015] The local temperature change trend, uplink and downlink phase offset change trend and user service quality index change trend collected during the controlled power downgrade test are normalized to generate corresponding trend feature data.

[0016] Based on trend feature data, the trend direction, change magnitude and response time sequence information of each indicator are extracted to construct a response feature vector;

[0017] The response feature vector is compared with a preset temperature-dominant response relationship template and a preset non-temperature-dominant fault response template. If the response feature vector is consistent with the temperature-dominant response relationship, the root cause of the fault is determined to be a reciprocity failure caused by temperature. If the response feature vector is consistent with the non-temperature-dominant fault response relationship, the root cause of the fault is determined to be a hardware fault caused by non-temperature factors.

[0018] Furthermore, by eliminating the power amplifier memory effect component based on the two composite phase offsets, the true phase offset value is obtained, including:

[0019] Two downlink probe signals are injected into the target radio frequency channel within a preset interval time to obtain the corresponding two sets of composite phase offset measurement data;

[0020] Calculate the phase response difference caused by the change in power amplifier output power in the two sets of measurement data, and extract the corresponding memory effect component;

[0021] The memory effect component is removed from the composite phase offset to obtain the true phase offset value related to temperature change.

[0022] Furthermore, the matching phase compensation value is obtained by interpolation or table lookup based on the temperature phase compensation correspondence table, including:

[0023] If the current local temperature matches the preset temperature point in the temperature phase compensation correspondence table perfectly, the corresponding phase compensation value is obtained by looking up the table.

[0024] Otherwise, select the upper and lower temperature points adjacent to the current local temperature and their corresponding adjacent temperature points from the temperature phase compensation correspondence table, and calculate the phase compensation value corresponding to the current local temperature using a linear interpolation algorithm based on the phase compensation values ​​of the adjacent temperature points.

[0025] Furthermore, the current local temperature and the actual phase offset value are recorded in a temperature phase compensation correspondence table, including:

[0026] Multiple consecutive data pairs of the current local temperature and the corresponding true phase offset value are obtained to form a new set of measurement samples;

[0027] Calculate the dispersion of the true phase shift value at each temperature point in the measurement sample to determine whether the measurement sample is stable.

[0028] When the dispersion is lower than the preset stability threshold, the weighted average method is used to fuse the true phase offset value corresponding to the temperature point, and the phase offset value associated with the corresponding temperature point is updated.

[0029] Furthermore, before updating the phase offset value associated with the corresponding temperature point, the process also includes:

[0030] The test context information for obtaining the true phase offset value after weighted average fusion includes the traffic load level, output power level and power stability at the time of measurement.

[0031] Based on the test context information and the preset credibility evaluation rules, the credibility level of the group of true phase offset values ​​is evaluated.

[0032] When the confidence level is higher than the preset confidence threshold, the set of real phase offset values ​​is used to update the temperature phase compensation correspondence table.

[0033] Furthermore, the controlled power downsampling test includes:

[0034] According to the preset power reduction curve, the power amplifier output power is reduced in stages for radio frequency channels that are suspected of reciprocity failure.

[0035] During each power reduction phase, corresponding local temperatures, uplink and downlink phase offsets, and user service quality indicators are collected synchronously to form multiple sets of power response data. Based on these multiple sets of power response data, the response sensitivity of multi-dimensional operating status data under different power conditions is evaluated to help distinguish between temperature-dominated reciprocity failures and hardware failures caused by non-temperature factors.

[0036] Furthermore, correlation analysis is performed on multi-dimensional operational status data, including:

[0037] Within a preset monitoring period, if the business load intensity exceeds the preset business threshold, the power amplifier output power exceeds the preset power threshold, the local temperature change rate continues to rise, and the user service quality indicators continue to decline, it is determined that the user service quality has deteriorated, triggering a correlation analysis of multi-dimensional operating status data.

[0038] Furthermore, the phase compensation value is applied to the calculation of the downlink beamforming weight, including:

[0039] Based on a preset beam direction configuration strategy, construct the initial weights for downlink beamforming;

[0040] The phase compensation value is superimposed on the phase factor of the corresponding RF channel in the initial weight of downlink beamforming to generate the temperature-compensated beamforming weight.

[0041] The temperature-compensated beamforming weights are used for beamforming calculations of the downlink signal.

[0042] The above scheme can superimpose precise phase compensation values ​​into the downlink beamforming weights to generate temperature-compensated beamforming weights, thereby effectively correcting phase errors caused by temperature and significantly improving downlink beamforming accuracy.

[0043] Furthermore, this application also discloses a 5G base station signal dynamic optimization system, including:

[0044] The data acquisition module is used to acquire multi-dimensional operating status data of each radio frequency channel of the base station. The multi-dimensional operating status data includes at least the service load intensity, power amplifier output power, local temperature change rate and user service quality indicators.

[0045] The correlation analysis module is used to perform correlation analysis on multi-dimensional operational status data based on preset combination judgment rules;

[0046] The determination module is used to determine that the radio frequency channel has a suspected reciprocity failure when the correlation analysis simultaneously meets the combined conditions of high service load, high power output, continuous rise in local temperature and decline in user service quality.

[0047] The fault analysis module is used to perform controlled power downsampling tests on radio frequency channels suspected of reciprocity failure, monitor local temperature change trends, uplink and downlink phase offset change trends, and user service quality index change trends in real time, and distinguish the root cause of the fault according to the preset response relationship.

[0048] The test and analysis module is used to inject two downlink detection signals and obtain the corresponding composite phase offset by loopback measurement if the root cause of the fault is determined to be reciprocity failure caused by temperature. Based on the two composite phase offsets, the power amplifier memory effect component is removed to obtain the true phase offset value. The current local temperature and the true phase offset value are recorded in the temperature phase compensation correspondence table.

[0049] The adjustment module is used to read the current local temperature in real time, obtain the matching phase compensation value by interpolation or table lookup according to the temperature phase compensation correspondence table, apply the phase compensation value to the calculation of downlink beamforming weight, and output the optimized beamforming result.

[0050] The alarm module is used to skip phase compensation and output a targeted maintenance alarm if the root cause of the fault is determined to be a hardware failure that is not caused by temperature factors.

[0051] The above scheme provides a system entity to implement the dynamic optimization method. Through modular design, it can efficiently perform functions such as data acquisition, correlation analysis, fault determination, root cause differentiation, phase compensation, and alarm, providing hardware and software support for the stable operation and performance optimization of 5G base stations.

[0052] As can be seen from the above, the 5G base station signal dynamic optimization method and system provided in this application can effectively improve beamforming accuracy and user service quality by dynamically identifying and distinguishing the root causes of reciprocity failure (temperature or hardware failure) of the base station radio frequency channel and performing targeted phase compensation or alarm. It has the advantages of effectively identifying and solving the problem of uplink and downlink channel reciprocity failure caused by temperature or hardware failure in 5G base stations, and significantly improving beamforming accuracy and user service quality. Attached Figure Description

[0053] Figure 1 This is a flowchart illustrating a method for dynamic optimization of 5G base station signals provided in this application.

[0054] Figure 2 A flowchart of a 5G base station signal dynamic optimization system provided in this application.

[0055] In the diagram: 1. Data acquisition module; 2. Correlation analysis module; 3. Judgment module; 4. Fault analysis module; 5. Test analysis module; 6. Adjustment module; 7. Alarm module. Detailed Implementation

[0056] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0057] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0058] Reference Figure 1 This application proposes a method for dynamic optimization of 5G base station signals, which specifically includes the following steps:

[0059] S1000: Acquire multi-dimensional operational status data of each radio frequency channel of the base station. The multi-dimensional operational status data includes at least the service load intensity, power amplifier output power, local temperature change rate, and user service quality indicators.

[0060] S2000: Based on preset combination judgment rules, perform correlation analysis on multi-dimensional operation status data;

[0061] S3000: When the correlation analysis simultaneously meets the combined conditions of high service load, high power output, continuous rise in local temperature and decline in user service quality, the radio frequency channel is determined to have a suspected reciprocity failure.

[0062] S4000: Performs controlled power downsampling test on RF channels suspected of reciprocity failure, monitors local temperature change trends, uplink and downlink phase offset change trends and user service quality index change trends in real time, and distinguishes the root cause of the fault according to the preset response relationship.

[0063] S5000: If the root cause of the fault is determined to be reciprocity failure caused by temperature, inject two downlink detection signals and use loopback measurement to obtain the corresponding composite phase offset. Based on the two composite phase offsets, remove the power amplifier memory effect component to obtain the true phase offset value. Record the current local temperature and the true phase offset value in the temperature phase compensation correspondence table.

[0064] S6000: Reads the current local temperature in real time, obtains the matching phase compensation value by interpolation or table lookup according to the temperature phase compensation correspondence table, applies the phase compensation value to the calculation of downlink beamforming weight, and outputs the optimized beamforming result.

[0065] S7000: If the root cause of the fault is determined to be a hardware fault not caused by temperature factors, skip phase compensation and output a targeted maintenance alarm.

[0066] In this embodiment, multi-dimensional operational status data refers to a collection of various types of data reflecting the operational status of the base station's radio frequency channel. This data may include, but is not limited to, service load intensity, power amplifier output power, local temperature change rate, and user service quality indicators. For example, it may also include signal-to-noise ratio, bit error rate, and channel utilization. Its main purpose is to comprehensively monitor the base station's operational status and provide a data foundation for subsequent anomaly identification. Preset combination judgment rules refer to a series of logical conditions or threshold combinations pre-set before or during system operation. These can be implemented using rule sets based on expert experience, classifiers trained by machine learning models, or statistical analysis models. For example, a logical judgment such as "when the service load intensity is higher than X, the power amplifier output power is higher than Y, the local temperature change rate is consistently higher than Z, and the user service quality indicator is lower than W" is used. Its main purpose is to identify specific patterns indicating potential reciprocity failures from complex multi-dimensional data.

[0067] Suspected reciprocity failure refers to a preliminary judgment state in which the uplink and downlink channel characteristics of a base station radio frequency channel may no longer meet the reciprocity assumption under specific operating conditions. It can be manifested as asymmetry in the phase, amplitude, or delay response of the uplink and downlink. It is mainly to indicate that the system needs to perform further diagnosis and confirmation of the channel.

[0068] Controlled power reduction testing is a diagnostic operation that involves gradually reducing the output power of the power amplifier of an RF channel suspected of reciprocity failure under controlled conditions. It can be achieved by using methods such as phased power reduction, step-by-step power reduction, or continuous smooth power reduction. Its main purpose is to observe the response changes of various indicators of the channel under different power levels, thereby helping to distinguish the root cause of the fault.

[0069] The memory effect component of a power amplifier refers to the nonlinear phase or amplitude response of its output signal, which depends not only on the current input signal but also on its historical input signals or operating state. This nonlinearity can be identified and quantified using methods such as two-tone testing, impulse response measurement, or nonlinear modeling based on machine learning. Its main purpose is to accurately isolate errors caused by the power amplifier's own nonlinear characteristics when measuring composite phase offsets, thus obtaining a more realistic, temperature-dependent phase offset. A temperature-phase compensation mapping table is a data structure that stores the mapping relationship between different local temperature values ​​and their corresponding true phase offset values. It can be constructed using lookup tables, functional relationships, or polynomial fitting models. Its main purpose is to provide data for obtaining accurate phase compensation values ​​based on real-time temperature. Interpolation or lookup table methods refer to two common methods for obtaining matching phase compensation values ​​from the temperature-phase compensation mapping table based on the current local temperature value. Lookup table is suitable when the temperature value perfectly matches a preset point in the table, while interpolation (e.g., linear interpolation, spline interpolation) is suitable when the current temperature value falls between preset points in the table. Its main purpose is to ensure accurate phase compensation values ​​can be obtained under various temperature conditions. The application of phase compensation value in the calculation of downlink beamforming weights refers to introducing the phase compensation value obtained based on real-time temperature as a correction factor into the downlink beamforming algorithm to adjust the direction and shape of beamforming. This can be achieved by directly superimposing it onto the phase factor, multiplicative correction, or adjusting the weight matrix based on the compensation value. Its main purpose is to correct the uplink and downlink channel phase offset caused by temperature changes, thereby restoring beamforming accuracy and ensuring that the downlink signal is accurately focused on the target user.

[0070] The core innovation of this application lies in combining the correlation analysis of multi-dimensional operating status data with controlled power downsampling tests, thereby accurately identifying and distinguishing reciprocity failures caused by temperature changes from hardware faults caused by non-temperature factors. Furthermore, by eliminating the memory effect component of the power amplifier, the true temperature-related phase offset value is obtained. Finally, this phase compensation value is dynamically applied to the calculation of downlink beamforming weights, achieving the effect of effectively compensating for dynamic reciprocity failures and maintaining beamforming performance.

[0071] This application's solution acquires multi-dimensional operational status data from each radio frequency (RF) channel of the base station. Based on preset combination judgment rules, it performs correlation analysis on this multi-dimensional operational status data. The aim is to identify patterns where a single indicator may not exceed a threshold, but when combined, they indicate potential problems. When the correlation analysis results simultaneously satisfy the combined conditions of high service load, high power output, continuous rise in local temperature, and decline in user service quality, the system determines that the RF channel has a suspected reciprocity failure. This indicates that the channel may have performance degradation caused by the coupling of multiple factors, requiring further diagnosis. Once a suspected reciprocity failure is determined, the system immediately performs a controlled power reduction test on the RF channel. During the test, the system monitors the local temperature change trend, the uplink and downlink phase offset change trend, and the user service quality indicator change trend in real time. By observing the response characteristics of these indicators during power reduction, the system can distinguish the root cause of the fault according to preset response relationships, determining whether the reciprocity failure is mainly caused by temperature changes or by hardware failures caused by other non-temperature factors. This distinction is crucial because it determines the subsequent processing path. If the root cause of the fault is determined to be a temperature-induced reciprocity failure, the system will take precise compensation measures. First, the system injects two downlink probe signals and uses loopback measurement technology to obtain the corresponding composite phase offset. To ensure the accuracy of compensation, the system uses a specific algorithm to eliminate the power amplifier memory effect component based on these two composite phase offsets, thus obtaining the true phase offset value related to temperature changes. Subsequently, the system records the current local temperature and this true phase offset value in a temperature-phase compensation correspondence table. During daily operation, the system reads the current local temperature in real time. Based on the previously established temperature-phase compensation correspondence table, the system obtains the phase compensation value matching the current local temperature through interpolation or table lookup. This compensation value is a precise correction for the uplink and downlink phase asymmetry under the current temperature conditions. After obtaining the compensation value, the system applies it to the calculation of downlink beamforming weights. This means that when generating the beamforming weights for the downlink signal, the phase error caused by temperature is considered and offset, thus outputting an optimized beamforming result. In this way, even in complex environments with dynamically changing temperatures, the downlink beam can accurately point to the target user, maintaining communication quality. Conversely, if the root cause of the fault is determined to be a hardware failure not caused by temperature factors, the system will skip the phase compensation step, as phase compensation cannot solve the underlying problem in this case. Instead, the system will output a targeted maintenance alarm, prompting maintenance personnel to perform corresponding hardware checks and repairs, thereby avoiding unnecessary compensation operations and improving the efficiency and accuracy of fault handling.

[0072] In some preferred embodiments, this application is implemented as follows: Base station operational status data can be collected through sensors and software modules integrated within the radio frequency unit. For example, service load intensity can be obtained by statistically analyzing data throughput or user connections within a specific time period; power amplifier output power can be directly read from the power amplifier's monitoring port; the rate of change of local temperature can be monitored in real time and its rate of change calculated using a thermistor or infrared sensor array; user service quality indicators can be evaluated based on network performance indicators such as signal-to-noise ratio, retransmission rate, and latency. When performing correlation analysis, an expert system based on a rule engine can be used, where preset combination judgment rules can be defined as a series of logical AND operations. For example, when it is detected that the service load intensity exceeds 80% for 5 consecutive minutes, the power amplifier output power remains above 90% of the rated power, the local temperature rises by more than 0.5 degrees Celsius per minute, and the average user download speed decreases by more than 10%, a suspected reciprocity failure judgment is triggered. When performing controlled power reduction tests, the system can reduce the power amplifier output power by 5% every 30 seconds according to a preset power reduction curve, and simultaneously collect various monitoring data. The root cause differentiation of faults can utilize machine learning classifiers, such as Support Vector Machines (SVMs) or neural networks, to learn the response characteristics of temperature-dominated and non-temperature-dominated hardware faults by training on historical fault data. If a temperature-induced reciprocal failure is determined, the downlink probe signal can use a single-tone signal or a pseudo-random sequence signal with a specific frequency and power. Loopback measurement can be performed by connecting the transmitter and receiver of the RF channel through a calibrated RF cable to form a closed loop. The removal of power amplifier memory effect components can be modeled using a Volterra series model or a neural network model, and its influence can be removed by inverting two measurement data. The temperature phase compensation correspondence table can be stored in the base station's non-volatile memory and organized in the form of key-value pairs (temperature: phase offset). When reading the current local temperature in real time and obtaining the phase compensation value, if the current temperature does not perfectly match the temperature point recorded in the table, a linear interpolation algorithm can be used to calculate the compensation value based on the two known temperature points closest to the current temperature and their corresponding phase offset values. For example, if the table records that 40℃ corresponds to a 5-degree offset and 50℃ corresponds to a 7-degree offset, then when the real-time temperature is 45℃, a 6-degree offset is calculated through linear interpolation. When applying the phase compensation value to the calculation of downlink beamforming weights, this compensation value can be directly superimposed on the phase factor of the corresponding RF channel in the beamforming algorithm. For example, if the phase factor of a certain channel in the original beamforming weights is θ, then the compensated phase factor will be θ + the phase compensation value. Finally, the optimized beamforming result will be used for downlink signal transmission at the base station to ensure accurate focusing of signal energy.

[0073] Another embodiment of this application further proposes that the sub-step of S4000: distinguishing the root cause of the fault according to a preset response relationship, includes:

[0074] S4210: Normalizes the local temperature change trend, uplink and downlink phase offset change trend and user service quality index change trend collected during the controlled power downsizing test to generate corresponding trend feature data.

[0075] S4220: Extract the trend direction, change magnitude and response time sequence information of each indicator based on trend feature data, and construct a response feature vector;

[0076] S4230: Compare the response feature vector with the preset temperature-dominant response relationship template and the preset non-temperature fault response template. If the response feature vector is consistent with the temperature-dominant response relationship, the root cause of the fault is determined to be a reciprocity failure caused by temperature. If the response feature vector is consistent with the non-temperature fault response relationship, the root cause of the fault is determined to be a hardware fault caused by non-temperature factors.

[0077] Normalization refers to converting data with different dimensions or numerical ranges to a unified scale. Methods such as min-max normalization, Z-score normalization, or decimal scaling normalization can be used. The purpose is to eliminate the influence of differences in dimensions and numerical ranges between data, so that the changing trends of different indicators can be compared and analyzed on the same scale, avoiding misjudgments caused by numerical differences.

[0078] Trend feature data refers to the normalized trends of local temperature changes, uplink and downlink phase offset changes, and user service quality indicators, aiming to provide standardized input for subsequent feature extraction and pattern recognition. Trend direction refers to the overall direction of the indicator's change with controlled power downscaling, such as rising, falling, or remaining stable, which can be determined by calculating the slope of the data sequence, the sign of the difference between the first and last values, or by using trend line fitting. Change magnitude refers to the severity or range of numerical changes in the indicator during controlled power downscaling testing, which can be quantified by calculating the difference between the maximum and minimum values, standard deviation, or mean absolute deviation of the data sequence. Response timing information refers to the speed and sequence of different indicators' responses to controlled power downscaling, which can be obtained by analyzing the time points when each indicator reaches a specific change ratio, response delay, or cross-correlation. A response feature vector is a multidimensional data representation formed by combining the trend direction, magnitude of change, and response timing information of extracted indicators. These feature values ​​can be arranged into a vector in a specific order. Its purpose is to comprehensively and quantitatively describe the overall response pattern of each indicator during controlled power reduction testing, serving as a basis for distinguishing the root cause of a fault. A preset temperature-dominant response relationship template is a pre-established reference pattern characterizing the response characteristics of each indicator under temperature-induced reciprocal failure fault modes. It can be a typical feature vector or feature range constructed through analysis of a large amount of historical fault data, expert experience summarization, or simulation. Its purpose is to provide a standardized reference for identifying temperature-related fault modes. A preset non-temperature-related fault response template is a pre-established reference pattern characterizing the response characteristics of each indicator under hardware fault modes caused by non-temperature factors. It can be a typical feature vector or feature range constructed through analysis of a large amount of historical fault data, expert experience summarization, or simulation. Its purpose is to provide a standardized reference for identifying non-temperature-related hardware fault modes.

[0079] This application's solution normalizes the local temperature change trends, uplink and downlink phase offset change trends, and user service quality index change trends collected during controlled power reduction testing. This eliminates differences in the dimensions and numerical ranges of different indicators, allowing for comparison and analysis of the change trends of different indicators on the same scale. This avoids misjudgments caused by numerical differences and lays the foundation for subsequent feature extraction and pattern recognition. Based on this, and using the normalized trend feature data, the trend direction, change magnitude, and response timing information of each indicator are further extracted to construct a response feature vector. The trend direction reflects the change trend of the indicator with power reduction, the change magnitude quantifies the severity of the indicator change, and the response timing information describes the speed and sequence of different indicators' responses to the power reduction operation. By comprehensively considering these three aspects, the response characteristics under different fault modes can be more comprehensively described, constructing a response feature vector capable of distinguishing different fault root causes. Finally, the constructed response feature vector is compared with a preset temperature-dominant response relationship template and a preset non-temperature-dominant fault response template. By comparing the extracted response feature vector with the preset templates, automatic identification and classification of fault modes are achieved. By comparing the results, it can be determined which fault mode the current response feature vector most closely matches, thus identifying the root cause of the fault. This collaborative mechanism significantly improves the accuracy and efficiency of base station fault diagnosis, avoids misjudgments and unnecessary maintenance, and ensures the stability of beamforming performance and continuous optimization of user service quality.

[0080] In some preferred embodiments, this application is implemented as follows: After performing a controlled power downscaling test on an RF channel suspected of reciprocity failure and monitoring the local temperature change trend, uplink / downlink phase offset change trend, and user service quality index change trend in real time, the system can first normalize these raw monitoring data. For example, a minimum-maximum normalization method can be used to linearly scale the data sequence of each index to the range of 0 to 1, i.e.: normalized value = (current value - minimum value) / (maximum value - minimum value). After this processing, the numerical range of different indices becomes consistent, facilitating subsequent feature extraction. Then, based on these normalized trend feature data, the system can extract the trend direction, change magnitude, and response timing information of each index. For example, for the trend direction, the linear regression slope of each index data sequence can be calculated; a positive slope indicates an upward trend, a negative slope indicates a downward trend, and a slope close to zero indicates a stable trend. For the change magnitude, the difference between the maximum and minimum values ​​of each index data sequence can be calculated to quantify its fluctuation range. For response timing information, the time required for each indicator to reach a steady state from the start of power reduction can be recorded, or the peak delay of the cross-correlation between different indicator change curves can be calculated to reflect the order of their responses. These extracted feature values, such as temperature trend slope, phase offset range, and user quality response time, can be combined into a response feature vector. For example, a response feature vector can be represented as [temperature slope, phase offset range, user quality response time, ...]. Finally, the constructed response feature vector is compared with a preset template. The preset template can be stored as multiple typical feature vectors, each representing a fault mode (e.g., temperature-dominated reciprocity failure or non-temperature-related hardware failure). During the comparison, the Euclidean distance or cosine similarity between the current response feature vector and each preset template vector can be calculated. If the current response feature vector is closest to or has the highest similarity to the preset temperature-dominated response relationship template, the root cause of the fault is determined to be a temperature-induced reciprocity failure. Conversely, if it is closest to or has the highest similarity to the preset non-temperature-related fault response template, the root cause of the fault is determined to be a non-temperature-related hardware failure. These templates can be generated and optimized in advance by training a large amount of historical failure data with machine learning (e.g., support vector machines, neural networks, or decision trees) to ensure their ability to distinguish different failure modes.

[0081] Another embodiment of this application further proposes that the sub-step of S5000: based on the two composite phase offsets, the power amplifier memory effect component is eliminated to obtain the true phase offset value, including:

[0082] S5210: Inject two downlink probe signals into the target RF channel within a preset interval time to obtain the corresponding two sets of composite phase offset measurement data;

[0083] S5220: Calculate the phase response difference caused by the change in power amplifier output power in two sets of measurement data, and extract the corresponding memory effect component;

[0084] S5230: Remove the memory effect component from the composite phase offset to obtain the true phase offset value related to temperature change.

[0085] The preset interval time refers to the time length between two downlink probe signal injections. Its purpose is to ensure that the power amplifier can fully demonstrate its memory effect between the two probes, while avoiding excessive interference from other non-memory effect factors due to excessive time. The composite phase offset refers to the sum of all phase responses of the RF channel obtained through loopback measurement. Its purpose is to comprehensively reflect the phase characteristics of the RF channel. The phase response difference caused by the change in power amplifier output power refers to the difference in phase response caused by the change in power amplifier output power during the two downlink probe signal injections. Its purpose is to capture the direct manifestation of the power amplifier memory effect. The memory effect component refers to the portion of the phase offset in the composite phase offset that is independently contributed by the power amplifier memory effect. Its purpose is to identify and eliminate the part that interferes with the true phase offset value. The true phase offset value refers to the phase offset caused only by temperature change after removing the memory effect component from the composite phase offset. Its purpose is to provide an accurate basis for subsequent temperature phase compensation.

[0086] This scheme injects two downlink probe signals into the target RF channel within a preset interval to obtain the measurement data of the corresponding two sets of composite phase offsets, thereby capturing the phase change of the power amplifier due to the memory effect.

[0087] In some preferred embodiments: after determining that there is a temperature-induced reciprocity failure in the RF channel, a dedicated phase measurement process is initiated. Specifically, the baseband processing unit can control the RF front-end to continuously inject two downlink probe signals into the target RF channel at a preset interval of, for example, 10 milliseconds. These probe signals can be short pulse signals with specific frequencies and powers. After each injection, the corresponding composite phase offset is acquired in real time through the loopback path of the RF channel using a high-precision RF measurement module. Thus, two sets of measurement data can be obtained, for example, the first set of composite phase offset is Φ1, and the second set is Φ2. Further, a digital signal processing unit can receive these two sets of measurement data. This digital signal processing unit can pre-store a memory effect characteristic model of the target power amplifier, or monitor the actual output power change of the power amplifier during the two probe signal injections in real time. For example, if the power amplifier output power is P1 during the first probe and P2 during the second, and it is known that the power amplifier will introduce a phase response difference of ΔΦ_PA under the power change from P1 to P2, then this ΔΦ_PA is the memory effect component. The digital signal processing unit can calculate the difference between Φ1 and Φ2, and, combined with the change in power amplifier output power, accurately extract the phase response difference caused by the change in power amplifier output power, i.e., the memory effect component, through table lookup or model calculation. Finally, the digital signal processing unit removes the extracted memory effect component from any set of composite phase offsets. For example, subtracting the memory effect component from Φ1 yields the true phase offset value Φ_true related to temperature changes. This Φ_true value can then be used to update the temperature phase compensation correspondence table, thereby ensuring that subsequent beamforming weight calculations are based on accurate temperature-related phase information.

[0088] Another embodiment of this application further proposes that the sub-step of S6000: obtaining the matching phase compensation value according to the temperature phase compensation correspondence table by interpolation or table lookup includes:

[0089] S6210: If the current local temperature matches the preset temperature point in the temperature phase compensation correspondence table perfectly, look up the table to obtain the corresponding phase compensation value.

[0090] S6220: Otherwise, select the upper limit temperature point and lower limit temperature point adjacent to the current local temperature and their corresponding adjacent temperature points from the temperature phase compensation correspondence table, and calculate the phase compensation value corresponding to the current local temperature using a linear interpolation algorithm based on the phase compensation value of the adjacent temperature points.

[0091] Among them, the preset temperature point refers to a specific temperature value with a corresponding phase compensation value that is pre-recorded in the temperature phase compensation correspondence table. It can be a discrete temperature sampling point obtained by measurement or calibration during the operation of the base station. The upper limit temperature point and the lower limit temperature point refer to the two recorded temperature points that are closest to the current local temperature in the temperature phase compensation correspondence table. One temperature point is higher than the current local temperature, and the other temperature point is lower than the current local temperature.

[0092] In some preferred embodiments, this application is implemented as follows: Assume the temperature phase compensation correspondence table is stored as an ordered list of temperature-phase value pairs, for example, with temperature as the key and phase compensation value as the value, arranged in ascending order of temperature. When the system needs to obtain the phase compensation value corresponding to the current local temperature, a processing unit first reads the current local temperature. The processing unit traverses or uses binary search or other methods to search the temperature phase compensation correspondence table for a preset temperature point that is completely consistent with the current local temperature. For example, if the current local temperature is 50.0 degrees Celsius, and there happens to be a temperature point with a record of 50.0 degrees Celsius in the table, then the processing unit will directly extract the phase compensation value corresponding to 50.0 degrees Celsius. If the processing unit fails to find a preset temperature point that completely matches the current local temperature in the table, for example, if the current local temperature is 50.5 degrees Celsius, but the table only contains records of 50.0 degrees Celsius and 51.0 degrees Celsius, then the processing unit will identify 50.0 degrees Celsius as the lower limit temperature point and 51.0 degrees Celsius as the upper limit temperature point. The processing unit then obtains the phase compensation values ​​corresponding to each of the two temperature points. For example, suppose the phase compensation value for 50.0 degrees Celsius is P1, and the phase compensation value for 51.0 degrees Celsius is P2. The processing unit will use the linear interpolation formula: Compensation value = P1 + (P2 - P1) The phase compensation value corresponding to the current local temperature of 50.5 degrees Celsius is calculated using the formula: ((current local temperature - lower limit temperature point) / (upper limit temperature point - lower limit temperature point)). This method ensures a smooth and continuous phase compensation value even if the current temperature is not at a preset discrete point, thus guaranteeing the accuracy of the compensation.

[0093] Another embodiment of this application further proposes a sub-step of S5000: recording the current local temperature and the true phase offset value into a temperature phase compensation correspondence table, including:

[0094] S5310: Acquire multiple consecutive data pairs of the current local temperature and the corresponding true phase offset value to form a new set of measurement samples;

[0095] S5320: Calculate the dispersion of the true phase shift value of each temperature point in the measurement sample to determine whether the measurement sample is stable;

[0096] S5330: When the dispersion is lower than the preset stability threshold, the weighted average method is used to fuse the true phase offset value corresponding to the temperature point, and the phase offset value associated with the corresponding temperature point is updated.

[0097] The measurement sample refers to a set of data continuously acquired over a period of time, showing the current local temperature and its corresponding true phase offset. Its purpose is to provide sufficient data for statistical analysis, thereby reducing random errors that may be introduced by a single measurement. Dispersion refers to the range of fluctuation of the true phase offset values ​​in the measurement sample around its mean or median. This can be measured using statistical measures such as standard deviation, variance, or mean absolute deviation, and its purpose is to quantify the stability of the measurement data. The preset stability threshold is a predetermined value used to determine whether the measurement sample meets the stability conditions. It can be set according to the accuracy requirements of the actual application scenario, and its purpose is to screen out reliable measurement data. Weighted averaging fusion refers to a method of averaging the true phase offset values ​​corresponding to each temperature point in the measurement sample by assigning different weights based on the reliability or importance of each data point. Specifically, it can use algorithms that allocate weights based on factors such as measurement time, signal strength, or data quality, and its purpose is to further reduce the impact of noise and improve the accuracy of the fusion results during the data fusion process.

[0098] The proposed solution optimizes the construction process of the temperature phase compensation correspondence table, ensuring the accuracy and reliability of the recorded data. This improvement not only solves the data instability problem that may be caused by the original recording method, but also further enhances the robustness and performance of the entire beamforming optimization method, enabling the base station to more accurately compensate for reciprocity failures caused by temperature and maintain high-quality communication services.

[0099] In some preferred embodiments, this application is implemented as follows: When the system determines that the root cause of the fault is a reciprocity failure caused by temperature, when recording the current local temperature and the true phase offset value into the temperature phase compensation correspondence table, the system first continuously acquires, for example, 10 sets of data pairs of the current local temperature and the corresponding true phase offset value. These data pairs are collected within a short time (e.g., within 1 minute) and together constitute a measurement sample. For example, at a temperature of 55.0°C, the continuously measured true phase offset values ​​may be 7.1°, 7.0°, 7.2°, 7.0°, 7.1°, 7.0°, 7.1°, 7.2°, 7.0°, and 7.1°. Subsequently, the system calculates the dispersion of the true phase offset values ​​in the measurement sample; for example, it can calculate the standard deviation of these 10 phase offset values. If the calculated standard deviation is lower than a preset stability threshold (e.g., 0.1°), the measurement sample is determined to be stable. When it is determined to be stable, the system uses a weighted average method to fuse the true phase offset values ​​corresponding to the temperature point. For example, an Exponentially Weighted Moving Average (EWMA) algorithm can be used to fuse these 10 phase offset values, where newer measurements are given higher weights to reflect the current state. The fused result (e.g., 7.09°) will be used to update the phase offset value associated with 55.0°C in the temperature phase compensation correspondence table. This method ensures that the data recorded in the compensation table has undergone multi-point verification and optimization, thereby improving the accuracy and reliability of the data.

[0100] Another embodiment of this application further proposes that, before updating the phase offset value associated with the corresponding temperature point, the sub-step of S5330 includes:

[0101] S5331: Test context information for obtaining the true phase offset value after weighted average fusion. The test context information includes the traffic load level, output power level and power stability at the time of measurement.

[0102] S5332: Based on test context information and preset confidence evaluation rules, evaluate the confidence level of the set of true phase offset values;

[0103] S5333: When the confidence level is higher than the preset confidence threshold, the set of real phase offset values ​​is used to update the temperature phase compensation correspondence table.

[0104] Among them, the test context information refers to the operating environment and condition data when measuring the true phase offset value, which may include the load level, output power level, and power stability at the time of measurement; the preset reliability evaluation rules refer to the set of criteria used to quantitatively evaluate the reliability of the measurement data, which can be determined by predefined weights or logical judgments based on factors such as load level, output power level, and power stability; the reliability level refers to the quantitative representation of the reliability of a set of true phase offset values, which can be a numerical score, a discrete level, or a Boolean value, and its purpose is to intuitively reflect the quality of the measurement data; the preset reliability threshold refers to the minimum standard for judging whether the true phase offset value is reliable enough to update the temperature phase compensation correspondence table, which can be a numerical threshold or a specific level, and its purpose is to screen out high-quality measurement data and avoid introducing errors.

[0105] This application's solution further improves the accuracy and reliability of the temperature phase compensation correspondence table by introducing an evaluation mechanism for the reliability of measurement samples. Specifically, before updating the true phase offset values ​​after weighted average fusion, the system first obtains the test context information related to the measurement sample. This information covers the service load level, output power level, and power stability during the measurement, factors that directly reflect the base station's operating conditions and influence the measurement results of the phase offset values. Subsequently, based on this test context information and combined with preset reliability evaluation rules, the system evaluates the reliability level of the set of true phase offset values. For example, measurement data obtained under conditions of high service load, high output power, and stable power output of the base station are usually evaluated as having a high reliability level because these conditions are closer to the typical operating state of the base station, and the data is more representative. Conversely, measurement data obtained under conditions of low service load or large output power fluctuations may be evaluated as having a low reliability level because these data may not accurately reflect the normal performance of the base station. Finally, only when the evaluated reliability level is higher than a preset reliability threshold will the set of true phase offset values ​​be used to update the temperature phase compensation correspondence table. This mechanism ensures that only measurement data obtained under representative and reliable operating conditions are included in the compensation table, thus avoiding errors introduced by low-quality data. In this way, this scheme, while ensuring the stability of the measurement samples, further filters out high-quality measurement data, enabling the temperature phase compensation correspondence table to more accurately reflect the true phase characteristics of the base station's RF channel at different temperatures, thereby improving the stability and accuracy of beamforming performance.

[0106] In some preferred embodiments, this application is implemented as follows: After the system calculates the dispersion of the true phase offset value through multiple consecutive measurement samples and determines that it is lower than a preset stability threshold, a reliability assessment process is initiated before using a weighted average method to fuse the true phase offset value corresponding to the temperature point and preparing to update the temperature phase compensation correspondence table. First, the system obtains the test context information of the current measurement moment from the base station's operation monitoring module. For example, it can obtain the current service load level, such as CPU utilization, number of user connections, or data throughput percentage; output power level, such as the average value of the actual output power of the power amplifier; and power stability, such as the fluctuation range or standard deviation of the output power over a period of time. This information can be collected and stored in real time by the base station's performance management unit. Then, a built-in reliability assessment algorithm calculates the reliability level of the set of true phase offset values ​​based on this test context information and preset reliability evaluation rules. For example, the preset reliability evaluation rule can be defined as follows: if the workload level is higher than 80%, the output power level is higher than 70% of the rated power, and the power stability fluctuation is less than 5%, the reliability level is "high"; if the workload level is between 50% and 80%, the output power level is between 50% and 70%, and the power stability fluctuation is between 5% and 10%, then all other situations are "low". The reliability level can be quantified as a score from 0 to 100. Finally, the system compares the calculated reliability level with the preset reliability threshold. For example, if the preset reliability threshold is set to "medium" or the quantified score is 70, then only when the evaluation result reaches the "high" level or the score is higher than 70 will the weighted average fused true phase offset value of that group be allowed to be used to update the temperature phase compensation correspondence table. If the reliability level is lower than the threshold, the data group will be marked as unreliable and will not be used for updating, thus avoiding the contamination of the compensation table by data measured under atypical or unstable operating conditions.

[0107] Another embodiment of this application further proposes that the steps of the controlled power downsampling test include:

[0108] A1: According to the preset power reduction curve, perform a phased power amplifier output power reduction operation on the RF channel that is suspected of reciprocity failure;

[0109] A2: During each power reduction phase, the corresponding local temperature, uplink and downlink phase offset, and user service quality indicators are collected synchronously to form multiple sets of power response data. Based on the multiple sets of power response data, the response sensitivity of multi-dimensional operating status data under different power conditions is evaluated to help distinguish between temperature-dominated reciprocity failures and hardware failures caused by non-temperature factors.

[0110] The preset power reduction curve refers to a series of power output levels and their corresponding reduction order and magnitude that are predefined by the system before the power reduction test. It can take various forms such as step reduction, linear reduction or non-linear reduction. The purpose is to perform fine and multi-level control of the power amplifier output power so as to observe the response of the RF channel at different power points.

[0111] The phased power amplifier output power reduction operation means gradually reducing the output power of the power amplifier according to a preset power reduction curve, and maintaining it at each reduced power level for a period of time to allow the system to collect data. This can be done at multiple discrete power points, with the aim of simulating the behavior of the RF channel under different workloads, thereby obtaining more comprehensive operating status data.

[0112] Multiple sets of power response data refer to the collection of local temperature, uplink / downlink phase offset, and user service quality indicators synchronously acquired at each stage of the phased power downsampling operation. This data can include multi-dimensional measurements corresponding to each power downsampling point, aiming to provide rich data samples for subsequent root cause analysis. Evaluating the response sensitivity of multi-dimensional operating status data under different power conditions means analyzing the magnitude, rate of change, and responsiveness of these data points—local temperature, uplink / downlink phase offset, and user service quality indicators—to power amplifier output power changes. This can be achieved by calculating the gradient, correlation coefficient, or rate of change of each indicator relative to power changes. Temperature-dominated reciprocity failure refers to the disruption of uplink / downlink reciprocity in the RF channel. Its main cause is an abnormal increase in local temperature leading to changes in the electrical characteristics of RF devices. This can manifest as a significant response of local temperature and uplink / downlink phase offset to power downsampling, aiming to distinguish it from failures caused by non-temperature factors. Hardware failures caused by non-temperature factors refer to reciprocity failures or performance degradation of the radio frequency channel. The cause is not due to local temperature anomalies, but rather to other hardware defects or aging factors. This can manifest as a significant response of user service quality indicators to power reduction, while the response to local temperature and phase offset is not obvious. The purpose is to make accurate fault classification.

[0113] This application's solution improves the accuracy of root cause diagnosis by performing more refined controlled power reduction tests on RF channels suspected of reciprocity failure. Specifically, the solution first performs a phased power reduction operation on the power amplifier according to a preset power decrease curve. This phased reduction method allows the system to gradually and meticulously observe the behavior of the RF channel at different power levels, avoiding insufficient information or misjudgments that may result from a single power reduction. Due to this refined power control, the system can observe at multiple stable power points, laying the foundation for subsequent data acquisition. Based on this, at each power reduction stage, the system simultaneously collects corresponding local temperatures, uplink and downlink phase offsets, and user service quality indicators, thus forming multiple sets of power response data. This multi-dimensional, synchronous acquisition method ensures that comprehensive operational status information is obtained at each power point, avoiding the random errors that may exist at a single data point, and significantly improving the reliability of data analysis. By acquiring this rich power response data, the system can more comprehensively capture the dynamic response characteristics of the RF channel under different power conditions. Ultimately, based on these multiple sets of power response data, the system evaluates the response sensitivity of multi-dimensional operational status data under different power conditions. For example, temperature-dominated reciprocity failures typically exhibit high sensitivity of local temperature and uplink / downlink phase offset to power changes; that is, when power is reduced, these indicators show obvious and predictable changes. Hardware failures caused by non-temperature factors, on the other hand, may show that user service quality indicators are more sensitive to power changes, while the response to temperature and phase offset is less pronounced. It is through this evaluation of the response sensitivity of multi-dimensional data that this solution can more accurately identify the root cause of the failure, thereby helping to distinguish between temperature-dominated reciprocity failures and hardware failures caused by non-temperature factors.

[0114] Another embodiment of this application further proposes that S2000 includes:

[0115] S2100: Within a preset monitoring period, if the service load intensity exceeds the preset service threshold, the power amplifier output power exceeds the preset power threshold, the local temperature change rate continues to rise, and the user service quality indicators continue to decline, it is determined that the user service quality has deteriorated, triggering multi-dimensional operation status data for correlation analysis.

[0116] The preset monitoring time period refers to a specific time window set by the system to observe and confirm abnormal trends of multiple operating indicators. It can be a fixed duration, such as several minutes or several hours, or a dynamically adjusted duration. Its purpose is to filter out instantaneous fluctuations and ensure that the observed abnormal state has a certain degree of continuity and stability, thereby avoiding misjudgments caused by occasional events.

[0117] Determining a deterioration in user service quality means that the system, based on a continuous downward trend in user service quality indicators and in conjunction with preset deterioration judgment logic, confirms that the user's actual experience has been negatively affected. Specifically, this can be achieved by comparing the current user service quality indicators with historical baselines or preset minimum standards, and by comprehensively evaluating factors such as the magnitude and duration of the decline. The purpose is to ensure that subsequent correlation analysis is initiated for genuine user experience issues, thus avoiding unnecessary resource consumption.

[0118] This application's solution optimizes the triggering mechanism for multi-dimensional operational status data correlation analysis by introducing multiple conditional constraints and a time dimension. Specifically, the system first continuously monitors the service load intensity, power amplifier output power, local temperature change rate, and user service quality indicators of the base station's radio frequency channel within a preset monitoring period. Only when all four conditions are met simultaneously—service load intensity exceeding a preset service threshold, power amplifier output power exceeding a preset power threshold, local temperature change rate continuously increasing, and user service quality indicators continuously decreasing—will the system further determine whether user service quality has deteriorated. Once user service quality is confirmed to have deteriorated, correlation analysis of the multi-dimensional operational status data is finally triggered. This mechanism avoids misjudgments caused by instantaneous fluctuations or occasional factors, ensuring a higher confidence level for the initiation of correlation analysis. By closely linking the triggering conditions with the actual deterioration of user experience, this solution can more accurately capture the actual impact of reciprocity failure on system performance, thereby enabling subsequent fault diagnosis and handling processes to be more precisely targeted at real problems. This, combined with the step of performing correlation analysis based directly on preset combination judgment rules in the basic scheme, makes the judgment of suspected reciprocity failure of radio frequency channels more accurate and reliable, thereby improving the efficiency and accuracy of the entire 5G base station signal dynamic optimization method and reducing unnecessary diagnosis and maintenance operations.

[0119] In some preferred embodiments, this application is implemented as follows: The system can set a preset monitoring period of 5 minutes. During these 5 minutes, the system continuously collects and analyzes the operating status data of each radio frequency channel of the base station. For example, when the service load intensity of a certain radio frequency channel is detected to be higher than a preset service threshold (e.g., exceeding 80% of the channel capacity) for 3 consecutive minutes, and the output power of the power amplifier is higher than a preset power threshold (e.g., exceeding 90% of its rated maximum output power) for 3 consecutive minutes, and at the same time, the data from the local temperature sensor shows that the local temperature change rate of the channel has been continuously increasing over the past 5 minutes, for example, increasing by more than 0.5 degrees Celsius per minute. Furthermore, user service quality indicators, such as average user throughput or retransmission rate, continuously decrease within the same 5 minutes, for example, the average throughput decreases by more than 10% or the retransmission rate increases by more than 5%. When these conditions are met simultaneously within the preset monitoring period, the system can further determine whether user service quality has deteriorated. For example, if the average user throughput is lower than a preset minimum service guarantee threshold, or the retransmission rate exceeds a preset alarm threshold, it is determined that user service quality has deteriorated. Once degradation is confirmed, the system will trigger a correlation analysis of the multi-dimensional operating status data of the radio frequency channel to initiate the subsequent fault diagnosis process.

[0120] Another embodiment of this application further proposes that the sub-step of S6000: applying the phase compensation value to the calculation of the downlink beamforming weight includes:

[0121] S6310: Construct initial weights for downlink beamforming based on a preset beam direction configuration strategy;

[0122] S6320: The phase compensation value is superimposed on the phase factor of the corresponding RF channel in the initial weight of downlink beamforming to generate the temperature-compensated beamforming weight.

[0123] S6330: Apply the temperature-compensated beamforming weights to the beamforming calculation of the downlink signal.

[0124] The preset beam orientation configuration strategy refers to the beam pointing and shape rules pre-set according to network planning, coverage area, user distribution, or specific service requirements. It can be implemented using fixed beam configuration based on sector coverage, dynamic beam configuration based on user location, or adaptive beam configuration based on channel measurement. Its purpose is to provide a basic beam pointing and energy distribution for downlink signal transmission. The phase factor refers to the complex component in beamforming weights used to adjust the signal phase. Specifically, it is the phase part of the complex weights corresponding to each antenna element in the beamforming weight vector or matrix. Its purpose is to control the phase of the transmitted signal from each antenna element, thereby achieving beam pointing and shape control.

[0125] This application's solution addresses the issue of decreased beamforming accuracy due to temperature variations by incorporating temperature-induced phase compensation values ​​into the downlink beamforming process. Specifically, firstly, based on a preset beam orientation configuration strategy, initial downlink beamforming weights are constructed, laying the foundation for subsequent phase compensation and ensuring that the basic beam pointing and shape conform to network planning. These initial weights allow for targeted adjustments. Secondly, phase compensation values ​​obtained from the temperature phase compensation correspondence table are superimposed on the phase factor of the corresponding RF channel in the initial downlink beamforming weights to generate temperature-compensated beamforming weights. This step is the core of the solution; it directly corrects for RF channel phase deviations caused by temperature changes. By adjusting the phase factor, the beamforming weights reflect the true channel state, thus avoiding beam pointing deviations. This superposition compensates for temperature-induced reciprocity failures. Finally, the temperature-compensated beamforming weights are used for downlink signal beamforming calculations. This ensures that downlink signals can utilize temperature-corrected beamforming weights during transmission, thereby achieving beam pointing and energy concentration, improving signal strength and communication quality for target users, while reducing interference to other users. Overall, this solution is closely integrated with the previous steps of obtaining the true phase offset value and establishing a temperature phase compensation correspondence table, forming a closed-loop control mechanism. This mechanism encompasses everything from identifying temperature-induced reciprocity failures, quantifying phase offsets, establishing compensation relationships, to ultimately applying the compensation values ​​to beamforming. This continuously optimizes beamforming performance under dynamic temperature conditions, ensuring the stability and reliability of 5G base station communication.

[0126] In some preferred embodiments, applying the phase compensation value to the calculation of downlink beamforming weights can be implemented as follows: First, the base station's beamforming module can generate an initial beamforming weight matrix based on a preset sector coverage pattern or user location information. For example, for a three-sector base station, each sector can have a preset main beam direction, and the initial complex weights can be calculated accordingly. Next, when the system obtains the phase compensation value for a certain RF channel from the temperature phase compensation correspondence table, for example, if the channel experiences a 5-degree phase lag due to temperature increase, the 5-degree compensation value is converted to the corresponding complex form and directly multiplied or added to the phase factor corresponding to that RF channel in the initial beamforming weight matrix. For example, if the initial phase factor is e^(j If the initial theta is the compensation value and delta_phi is the initial theta, then the new phase factor can be e^(j... (theta_initial + delta_phi)). This adjusted phase factor, along with the phase factors of other temperature-unaffected channels and the amplitude factors of all channels, forms a new, temperature-compensated beamforming weight matrix. Finally, the base station's digital baseband processing unit multiplies the downlink signal to be transmitted with this temperature-compensated beamforming weight matrix to adjust the signal's phase and amplitude. The adjusted signal is then sent to the RF front-end for up-conversion and power amplification, and finally transmitted through the antenna array to form a precisely directional, energy-concentrated downlink beam.

[0127] Reference Figure 2 Another embodiment of this application further proposes a 5G base station signal dynamic optimization system, including:

[0128] Data acquisition module 1 is used to acquire multi-dimensional operating status data of each radio frequency channel of the base station. The multi-dimensional operating status data includes at least the service load intensity, power amplifier output power, local temperature change rate and user service quality indicators.

[0129] The correlation analysis module 2 is used to perform correlation analysis on multi-dimensional operation status data based on preset combination judgment rules;

[0130] Module 3 is used to determine that the radio frequency channel has a suspected reciprocity failure when the correlation analysis simultaneously meets the combined conditions of high service load, high power output, continuous rise in local temperature and decline in user service quality.

[0131] Fault analysis module 4 is used to perform controlled power downsampling tests on radio frequency channels suspected of reciprocity failure, monitor local temperature change trends, uplink and downlink phase offset change trends and user service quality index change trends in real time, and distinguish the root cause of the fault according to the preset response relationship.

[0132] Test and analysis module 5 is used to inject two downlink detection signals and obtain the corresponding composite phase offset by loopback measurement if the root cause of the fault is determined to be reciprocity failure caused by temperature. Based on the two composite phase offsets, the power amplifier memory effect component is removed to obtain the true phase offset value. The current local temperature and the true phase offset value are recorded in the temperature phase compensation correspondence table.

[0133] Adjustment module 6 is used to read the current local temperature in real time, obtain the matching phase compensation value by interpolation or table lookup according to the temperature phase compensation correspondence table, apply the phase compensation value to the calculation of downlink beamforming weight, and output the optimized beamforming result.

[0134] Alarm module 7 is used to skip phase compensation and output a targeted maintenance alarm if the root cause of the fault is determined to be a hardware fault that is not caused by temperature factors.

[0135] The solution proposed in this application achieves dynamic optimization of 5G base station signals through a modular system design.

[0136] Through the above technical solution, this system provides a concrete approach to dynamically optimize 5G base station signals, overcoming the limitations of relying solely on methods and steps in actual deployment and operation. Through modular design, the system achieves real-time monitoring of base station operating status, anomaly identification, accurate differentiation of fault root causes, and targeted compensation or alarms. Specifically, the system can effectively identify and compensate for asymmetric, dynamic uplink and downlink channel reciprocity failures caused by temperature changes, thereby maintaining beamforming performance and avoiding communication quality degradation due to channel reciprocity disruption. Simultaneously, for hardware failures not caused by temperature factors, the system can promptly issue maintenance alarms, avoiding unnecessary compensation operations and improving the efficiency and accuracy of fault handling. This enables the base station to continuously provide communication service quality under complex environments and high load conditions.

[0137] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for dynamic optimization of 5G base station signals, characterized in that, include: Acquire multi-dimensional operational status data for each radio frequency channel of the base station. The multi-dimensional operational status data includes at least service load intensity, power amplifier output power, local temperature change rate, and user service quality indicators. Based on preset combination judgment rules, the multi-dimensional operation status data is subjected to correlation analysis; When the correlation analysis simultaneously meets the combined conditions of high service load, high power output, continuous rise in local temperature and decline in user service quality, the radio frequency channel is determined to have a suspected reciprocity failure. For radio frequency channels suspected of reciprocity failure, a controlled power downsampling test is performed. Local temperature change trends, uplink and downlink phase offset change trends, and user service quality index change trends are monitored in real time. The root cause of the failure is identified according to the preset response relationship. If the root cause of the fault is determined to be a reciprocity failure due to temperature, two downlink probe signals are injected, and the corresponding composite phase offset is obtained using loopback measurement. Based on the two composite phase offsets, the power amplifier memory effect component is eliminated. Obtain the true phase offset value, and record the current local temperature and the true phase offset value in the temperature phase compensation correspondence table; The system reads the current local temperature in real time, obtains the matching phase compensation value by interpolation or table lookup according to the temperature phase compensation correspondence table, applies the phase compensation value to the calculation of downlink beamforming weight, and outputs the optimized beamforming result. If the root cause of the fault is determined to be a hardware failure not caused by temperature factors, skip phase compensation and output a targeted maintenance alarm.

2. The 5G base station signal dynamic optimization method according to claim 1, characterized in that, The method of distinguishing the root cause of a fault according to a preset response relationship includes: The local temperature change trend, uplink and downlink phase offset change trend and user service quality index change trend collected during the controlled power downgrade test are normalized to generate corresponding trend feature data. Based on the trend feature data, the trend direction, change magnitude and response time sequence information of each indicator are extracted, and a response feature vector is constructed. The response feature vector is compared with a preset temperature-dominant response relationship template and a preset non-temperature-dominant fault response template. If the response feature vector is consistent with the temperature-dominant response relationship, the root cause of the fault is determined to be a reciprocity failure caused by temperature. If the response feature vector is consistent with the non-temperature-dominant fault response relationship, the root cause of the fault is determined to be a hardware fault caused by non-temperature factors.

3. The 5G base station signal dynamic optimization method according to claim 2, characterized in that, The process of eliminating the power amplifier memory effect component based on the two composite phase offsets to obtain the true phase offset value includes: Two downlink probe signals are injected into the target radio frequency channel within a preset interval time to obtain the corresponding two sets of composite phase offset measurement data; Calculate the phase response difference caused by the change in power amplifier output power in the two sets of measurement data, and extract the corresponding memory effect component; The memory effect component is removed from the composite phase offset to obtain the true phase offset value related to temperature change.

4. The 5G base station signal dynamic optimization method according to claim 1, characterized in that, The step of obtaining the matching phase compensation value based on the temperature phase compensation correspondence table through interpolation or table lookup includes: If the current local temperature completely matches the preset temperature point in the temperature phase compensation correspondence table, the corresponding phase compensation value is obtained by looking up the table. Otherwise, select the upper limit temperature point and lower limit temperature point adjacent to the current local temperature and their corresponding adjacent temperature points from the temperature phase compensation correspondence table, and calculate the phase compensation value corresponding to the current local temperature using a linear interpolation algorithm based on the phase compensation values ​​of the adjacent temperature points.

5. The 5G base station signal dynamic optimization method according to claim 1, characterized in that, The step of recording the current local temperature and the actual phase offset value into a temperature phase compensation correspondence table includes: Multiple consecutive data pairs of the current local temperature and the corresponding true phase offset value are obtained to form a new set of measurement samples; Calculate the dispersion of the true phase shift value at each temperature point in the measurement sample to determine whether the measurement sample is stable. When the degree of dispersion is lower than a preset stability threshold, the weighted average method is used to fuse the true phase offset value corresponding to the temperature point, and the phase offset value associated with the corresponding temperature point is updated.

6. The 5G base station signal dynamic optimization method according to claim 5, characterized in that, Before updating the phase offset value associated with the corresponding temperature point, the method further includes: The test context information for obtaining the weighted average fused true phase offset value includes the service load level, output power level and power stability at the time of measurement. Based on the test context information and the preset credibility evaluation rules, the credibility level of the true phase offset value is evaluated. When the confidence level is higher than the preset confidence threshold, the actual phase offset value is used to update the temperature phase compensation correspondence table.

7. The 5G base station signal dynamic optimization method according to claim 1, characterized in that, The controlled power downsampling test includes: According to the preset power reduction curve, the power amplifier output power is reduced in stages for radio frequency channels that are suspected of reciprocity failure. During each power reduction phase, corresponding local temperature, uplink and downlink phase offset, and user service quality indicators are collected synchronously to form multiple sets of power response data. Based on these multiple sets of power response data, the response sensitivity of the multi-dimensional operating status data under different power conditions is evaluated to help distinguish between temperature-dominated reciprocity failures and hardware failures caused by non-temperature factors.

8. The 5G base station signal dynamic optimization method according to claim 1, characterized in that, The correlation analysis of the multi-dimensional operational status data includes: Within a preset monitoring period, if the service load intensity exceeds a preset service threshold, the power amplifier output power exceeds a preset power threshold, the local temperature change rate continues to rise, and the user service quality index continues to decline, it is determined that the user service quality has deteriorated, triggering the correlation analysis of the multi-dimensional operating status data.

9. The 5G base station signal dynamic optimization method according to claim 1, characterized in that, The step of applying the phase compensation value to the calculation of downlink beamforming weights includes: Based on a preset beam direction configuration strategy, construct the initial weights for downlink beamforming; The phase compensation value is superimposed on the phase factor of the corresponding RF channel in the downlink beamforming initial weight to generate the temperature-compensated beamforming weight. The temperature-compensated beamforming weights are used for beamforming calculations of downlink signals.

10. A 5G base station signal dynamic optimization system, characterized in that, include: The data acquisition module is used to acquire multi-dimensional operating status data of each radio frequency channel of the base station. The multi-dimensional operating status data includes at least the service load intensity, power amplifier output power, local temperature change rate, and user service quality indicators. The correlation analysis module is used to perform correlation analysis on the multi-dimensional operating status data based on preset combination judgment rules; The determination module is used to determine that the radio frequency channel has a suspected reciprocity failure when the correlation analysis simultaneously meets the combined conditions of high service load, high power output, continuous rise in local temperature and decline in user service quality. The fault analysis module is used to perform controlled power downsampling tests on radio frequency channels suspected of reciprocity failure, monitor local temperature change trends, uplink and downlink phase offset change trends, and user service quality index change trends in real time, and distinguish the root cause of the fault according to the preset response relationship. The test and analysis module is used to inject two downlink detection signals and obtain the corresponding composite phase offset by using loopback measurement if the root cause of the fault is determined to be reciprocity failure caused by temperature. Based on the two composite phase offsets, the power amplifier memory effect component is removed to obtain the true phase offset value. The current local temperature and the true phase offset value are recorded in the temperature phase compensation correspondence table. The adjustment module is used to read the current local temperature in real time, obtain the matching phase compensation value by interpolation or table lookup according to the temperature phase compensation correspondence table, apply the phase compensation value to the calculation of downlink beamforming weight, and output the optimized beamforming result. The alarm module is used to skip phase compensation and output a targeted maintenance alarm if the root cause of the fault is determined to be a hardware failure that is not caused by temperature factors.

Images