CN118803929A - Base station parameter optimization method, device, equipment, medium and program product - Google Patents

Abstract

The present application relates to the field of wireless network technology, and provides a base station parameter optimization method, device, equipment, medium and program product, which obtains the wireless key data indicator KPI of the base station to be optimized; adjusts the log data distribution of the preset training base station according to the KPI to obtain the corrected log data; inputs the corrected log data into a pre-trained test model, predicts the predicted base station performance of the base station to be optimized, and optimizes the base station operating parameters of the base station to be optimized based on the prediction results, wherein the test model is trained based on the log data of the preset training base station. When optimizing parameters, the present application selects wireless key data indicators KPIs that are easy to obtain, and the actual parameters of the base station need to be modified. There is no need to traverse different parameters for each base station. The base station parameter optimization is achieved by model estimation, which improves the efficiency of base station parameter optimization.

Description

Base station parameter optimization method, device, equipment, medium and program product

Technical Field

The present application relates to the field of wireless network technology, and in particular to a base station parameter optimization method, device, equipment, medium and program product.

Background Art

The specific functions or parameters of a base station have different performance in different business scenarios. In order to obtain better network performance, it is necessary to configure different parameters for different base stations.

At present, when optimizing base station networks, optimizers usually collect base station parameters and then configure and adjust the base station parameters based on experience. However, the various influencing factors in wireless communication networks are extremely complex, and it is impossible to directly determine the performance of different parameters in different business scenarios through manual experience. If different parameters are directly traversed for each base station through trial and error, and the optimal parameters are finally determined based on the actual performance of each base station, the workload is too huge and the efficiency of base station parameter optimization is low.

Summary of the invention

The present application provides a base station parameter optimization method, device, equipment, medium and program product to solve the defect of low efficiency of base station parameter optimization.

In a first aspect, the present application provides a base station parameter optimization method, comprising:

Obtain the wireless key data indicators KPI of the base station to be optimized;

Adjust the log data distribution of the preset training base station according to the KPI to obtain the corrected log data;

The corrected log data is input into a pre-trained test model to predict the predicted base station performance of the base station to be optimized, and the base station operating parameters of the base station to be optimized are optimized based on the prediction result, wherein the test model is trained based on the log data of a preset training base station.

In one embodiment, adjusting the log data distribution of the preset training base station according to the KPI to obtain the corrected log data includes:

Based on the KPI, determine the sampling quantity of the log data of the preset training base station;

Sampling is performed based on the sampling quantity to obtain the corrected log data.

In one embodiment, determining the sampling quantity of log data of a preset training base station based on the KPI includes:

Based on the KPI, obtaining a first distribution corresponding to a communication network performance indicator;

Based on the log data of the preset training base station, obtaining a second distribution corresponding to the first distribution;

According to the first distribution and the second distribution, a preset number of samples of log data of the training base station is determined.

In one embodiment, the communication network performance indicator is path loss, and determining the number of samples of log data of a preset training base station based on the KPI includes:

Based on the KPI, a first path loss distribution with an interval of first loss is obtained;

Based on the log data of the preset training base station, obtaining a second path loss distribution corresponding to the first path loss distribution;

Based on the first path loss distribution and the second path loss distribution, a sampling quantity of log data of a preset training base station is determined.

In one embodiment, determining the number of samples of log data of a preset training base station based on the first path loss distribution and the second path loss distribution includes:

Determining a target road loss interval according to a ratio between the first road loss distribution and the second road loss distribution in the same road loss interval;

Determine the total number of samples of the log data of the preset training base station according to the distribution of the KPI in the target path loss interval and the number of samples of the log data of the preset training base station in the target path loss interval;

According to the total sampling number and the first path loss distribution, the sampling number of the log data of the preset training base station in each path loss interval is determined.

In one embodiment, the test model trains the boosted tree model based on log data of a preset training base station.

In one embodiment, the step of obtaining a wireless key data indicator KPI of the base station to be optimized includes:

Acquire the wireless key data indicator KPI corresponding to the to-be-optimized parameter of the to-be-optimized base station based on the gateway;

Correspondingly, the inputting the corrected log data into the pre-trained test model, predicting the predicted base station performance of the base station to be optimized, and optimizing the base station operating parameters of the base station to be optimized based on the prediction result, includes:

Inputting the modified log data into a pre-trained test model to obtain a predicted base station performance corresponding to any parameter value of the parameter to be optimized;

The base station operating parameters are configured to be parameter values corresponding to the optimal predicted base station performance to complete base station parameter optimization.

In a second aspect, the present application also provides a base station parameter optimization device, comprising:

An acquisition module is used to acquire the wireless key data indicators KPI of the base station to be optimized;

An adjustment module, used for adjusting the log data distribution of a preset training base station according to the KPI to obtain corrected log data;

An optimization module is used to input the corrected log data into a pre-trained test model, predict the predicted base station performance of the base station to be optimized, and optimize the base station operating parameters of the base station to be optimized based on the prediction result, wherein the test model is trained based on the log data of a preset training base station.

In a third aspect, the present application also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, it implements any of the base station parameter optimization methods described above.

In a fourth aspect, the present application provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the base station parameter optimization methods described above.

In a fifth aspect, the present application also provides a computer program product, including a computer program, which, when executed by a processor, implements any of the base station parameter optimization methods described above.

The base station parameter optimization method, device, equipment, medium and program product provided in the present application obtain the wireless key data indicator KPI of the base station to be optimized; adjust the log data distribution of the preset training base station according to the KPI to obtain the corrected log data; input the corrected log data into the pre-trained test model, predict the predicted base station performance of the base station to be optimized, and optimize the base station operating parameters of the base station to be optimized based on the prediction result, wherein the test model is trained based on the log data of the preset training base station. When optimizing parameters, the present application selects a wireless key data indicator KPI that is easy to obtain, adjusts the log data distribution of the preset training base station to obtain the corrected log data, inputs the corrected log data into the pre-trained test model, predicts the predicted base station performance of the base station to be optimized, and optimizes the base station operating parameters of the base station to be optimized based on the prediction result, without modifying the actual parameters of the base station, without traversing different parameters for each base station, and realizes base station parameter optimization by model estimation, thereby improving the efficiency of base station parameter optimization.

BRIEF DESCRIPTION OF THE DRAWINGS

In order to more clearly illustrate the technical solutions in the present application or the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

FIG1 is a flow chart of a method for optimizing base station parameters provided by the present application;

FIG2 is a schematic diagram of the weight of log data of a preset training base station provided in the present application;

FIG3 is a schematic diagram of a process flow of predicted performance and actual performance provided by the present application;

FIG4 is a second flow chart of the base station parameter optimization method provided by the present application;

FIG5 is a schematic diagram of the structure of a base station parameter optimization device provided by the present application;

FIG. 6 is a schematic diagram of the structure of an electronic device provided in the present application.

DETAILED DESCRIPTION

In order to make the purpose, technical solutions and advantages of this application clearer, the technical solutions in this application will be clearly and completely described below in conjunction with the drawings in this application. Obviously, the described embodiments are part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

FIG. 1 is a flow chart of a base station parameter optimization method provided by the present application. As shown in FIG. 1 , the present application provides a base station parameter optimization method, including:

Step S100, obtaining the wireless key performance indicator (KPI) of the base station to be optimized; the wireless key performance indicator KPI includes indicator data of uplink rate related parameters of the base station to be optimized, which is used to characterize the network performance characteristics of the base station to be optimized.

KPI can be obtained through the gateway without going to the base station site. It is easier to obtain than large amounts of log data, and is related to the base station, which can reflect the characteristics of each base station to be optimized.

Step S200, adjusting the log data distribution of the preset training base station according to the KPI to obtain corrected log data;

The log data of the preset training base station (hereinafter referred to as log data or log data) is the log data of a small batch (at least dozens) of base stations for the preset business scenario. The log data distribution of the preset training base station is adjusted according to the KPI, that is, sampling is performed from the log data of the preset training base station according to the data distribution of the KPI to obtain corrected log data, so that the corrected log data is more strongly correlated with the base station to be optimized, so as to improve the accuracy of network performance prediction.

Step S300, input the corrected log data into a pre-trained test model, predict the predicted base station performance of the base station to be optimized, and adjust the base station operating parameters of the base station to be optimized based on the prediction result, wherein the test model is trained based on the log data of a preset training base station.

The preset boosting tree model is trained through the log data of the preset training base station and the actual network performance of a small batch of base stations, and the relationship between different business scenarios, different parameters and network performance is automatically learned to build a test model.

Specifically, by inputting the corrected log data into the test model, the test model can output the predicted network performance of the base station to be optimized under various parameters. By selecting the predicted network performance, the optimal parameters of the base station to be optimized can be determined, and the base station parameter optimization can be completed.

It can be understood that when optimizing parameters, the present application selects wireless key data indicators KPIs that are easy to obtain, adjusts the log data distribution of the preset training base stations to obtain corrected log data, inputs the corrected log data into the pre-trained test model, predicts the performance of the base station to be optimized, and optimizes the base station operating parameters of the base station to be optimized based on the prediction results. There is no need to modify the actual parameters of the base station, and there is no need to traverse different parameters for each base station. The base station parameter optimization is achieved through model estimation, which improves the efficiency of base station parameter optimization.

Based on the above embodiment, as an optional embodiment, the test model trains the boosted tree model based on log data of a preset training base station.

This application mainly considers the impact of different parameter modifications on the user's uplink rate. Therefore, when selecting log data, consider using log data related to the uplink rate for modeling.

Optionally, the log data includes user behavior data, channel condition data, network load data, base station parameters and terminal capability data. Specifically, the user behavior data includes an uplink user buffer status report, the channel condition data includes uplink path loss and uplink interference, the network load data includes the number of remaining resource blocks and the number of scheduled users, the base station parameters include uplink waveform parameters, and the terminal capability data includes maximum transmit power.

In one embodiment, the current uplink rate of the user is determined by the maximum uplink rate that can be supported and the uplink rate currently required by the user, whichever is smaller. The uplink rate currently required by the user is indicated by the user buffer status report (Buffer Status Report, BSR), and the maximum uplink rate is determined by the schedulable wireless resources and the carrying efficiency. Among them, the schedulable wireless resources are related to the number of remaining resource blocks (Resource Block, RB) and the number of scheduled users. The carrying efficiency is related to the channel conditions (path loss, uplink interference), terminal capabilities (maximum transmit power), and base station parameters (uplink waveform parameters). Therefore, the preset log data related to the uplink rate is mainly reflected in the following aspects:

User behavior: Uplink BSR (lsgbsr);

Channel conditions: uplink path loss (ulPl), uplink interference (puschNi);

Network load: number of remaining RBs (restRBNum), number of scheduled users (schdUeNum);

Base station parameters: uplink waveform parameters (para);

Terminal capability: maximum transmit power (pcMax).

Optionally, the boosted tree model uses the XGBoost model, which is also called the extreme gradient boosted tree. It is an implementation of the boosting algorithm and can be used to solve classification or regression problems with good results. The algorithm idea of the XGBoost model is to continuously add trees and continuously perform feature splitting to grow a tree. Each time a tree is added, it actually learns a new function to fit the residual of the last prediction. After training, k trees are obtained. If you want to predict the score of a sample, that is, according to the characteristics of the sample, it will fall into a corresponding leaf node in each tree. Each leaf node corresponds to a score. Finally, you only need to add up the scores corresponding to each tree to get the predicted value of the sample.

The cost function of the XGBoost model is:

Where n is the number of training samples, l is the loss of a single sample, y′ _i is the model’s predicted value for the training sample, y _i is the true label value of the training sample, and the ∑ _k Ω(f _k ) regularization term defines the complexity of the model.

_yi ′＝f(x), where x is the input value, specifically a 7-dimensional vector (i.e., x＝[x_ulPl,x_lcgBsr,x_schdUeNum,x_restRbNum,x_puschNi,x_pcMax,x_para]), which is the seven log data related to the uplink rate mentioned above. f(x) is the output value of the model (i.e., the predicted value of the uplink rate), and _yi is the actual uplink rate.

Figure 2 is a weight diagram of the log data of the preset training base station provided in the present application. As shown in Figure 2, the log data is used as a training sample, and the actual performance of the base station is used as the label of the training sample. The XGBoost model is trained to obtain a test model. After the model is trained, the weight value can be automatically output through feature_importance: the feature_importance attribute is an attribute in the XGBoost model used to indicate the importance of each feature in the model.

Figure 3 is a flow chart of the predicted performance and actual performance provided by the present application. As shown in Figure 3, after the test model training is completed, test samples are extracted from the log data and input into the test model. The numerical comparison between the predicted performance output by the test model and the actual performance of the test samples is shown in Figure 3. It can be seen that the test model obtained based on XGBoost model training can predict the performance of the base station more accurately and improve the accuracy of parameter optimization.

It can be understood that the present application creatively introduces the XGBoost model used in the prior art for image classification, medical simulation prediction, and industry into the field of base station parameter optimization. There is no need to modify the actual parameters of the base station, and there is no need to traverse different parameters for each base station. The base station parameter optimization is achieved through model estimation, thereby improving the efficiency of base station parameter optimization.

Based on the above embodiment, as an optional embodiment, the obtaining of the wireless key data indicator KPI of the base station to be optimized includes:

Step S100, obtaining a wireless key data indicator KPI corresponding to the to-be-optimized parameter of the to-be-optimized base station based on the gateway;

Specifically, collecting base station log data requires on-site testing. If log data of each base station is collected during parameter optimization, the workload is huge and the efficiency is extremely low. It is easier to obtain KPIs through the gateway.

KPIs include cell uplink UE Throughput rate distribution, PUSCH PRB utilization rate distribution, QPSK uplink transmission times distribution, 16QAM uplink transmission times distribution, 64QAM uplink transmission times distribution, 256QAM uplink transmission times distribution, PUSCH SINR distribution, path loss distribution of uplink service users between partitions, SRS NI distribution, cell uplink UE Throughput data transmission time, cell uplink UE Throughput data volume, number of PRBs used for uplink DTCH, traffic volume of successful uplink DTCH transmission, number of slots scheduled for uplink DTCH, number of slots schedulable for uplink in the cell, total number of uplink scheduled users in the cell, average uplink interference level in the cell, number of uplink DTX times for different modulation modes, number of uplink residual packet errors for different modulation modes, number of uplink successful reception TBs for different modulation modes, number of uplink failed reception TBs for different modulation modes, number of PRBs occupied by uplink PRACH, number of PRBs used for uplink DCCH, number of PRBs used for uplink DTCH, number of available PRBs for PUSCH, number of PRBs used for PUSCH, number of PRBs used for PUCCH, distribution of QPSK uplink transmission times, distribution of 16QAM uplink transmission times, distribution of 64QAM uplink transmission times, distribution of 256QAM uplink transmission times, cumulative path loss value of uplink service users, cumulative number of path losses for uplink service users.

Depending on the parameters to be optimized, the extracted KPI is also different. For example, if the waveform adaptive function parameters of a certain base station need to be modified, when only considering the modification of the waveform adaptive function parameters, the effective signal power received on the base station side is obtained by subtracting the path loss from the maximum transmission power of the terminal. The effective signal power is divided by the uplink interference to obtain the SINR. When the maximum transmission power and interference of the terminal remain unchanged, the SINR is only related to the path loss. The higher the SINR, the higher the carrying efficiency, which affects the uplink rate. Therefore, if the waveform parameters of a certain base station need to be modified, the KPI is the path loss distribution of uplink service users between partitions.

Correspondingly, the inputting the corrected log data into the pre-trained test model, predicting the predicted base station performance of the base station to be optimized, and optimizing the base station operating parameters of the base station to be optimized based on the prediction result, includes:

Step S310, inputting the modified log data into a pre-trained test model to obtain predicted base station performance corresponding to any parameter value of the parameter to be optimized;

For example, if the parameter to be optimized is a waveform adaptation function parameter, which has three possible values: auto, CP-OFDM, and DFT-S-OFDM, the log data related to the path loss distribution can be input into the network performance prediction model to obtain the gains corresponding to the three possible values of auto, CP-OFDM, and DFT-S-OFDM output by the network performance prediction model.

Step S320: configuring the base station operating parameters to the parameter values corresponding to the optimal predicted base station performance to complete base station parameter optimization.

It can be understood that the present application uses KPI to adjust the log data, that is, maps the log data through KPI. Compared with the parameter optimization method directly based on whether the KPI is stable or whether the log data is reasonable, the present application creatively optimizes the parameters through KPI mapping, which can reduce the difficulty of data acquisition and eliminate the need to traverse all the data for judgment, thereby improving the efficiency of parameter optimization.

Based on the above embodiment, as an optional embodiment, the step of adjusting the log data distribution of the preset training base station according to the KPI to obtain the corrected log data includes:

Step S210, determining the number of preset sampling of log data of the training base station based on the KPI;

There are a huge number of base stations in the existing network, and the acquisition of base station log data requires on-site network pulling. Therefore, it is very difficult and time-consuming to obtain a large number of base station log data for model correction. Considering the operability of large-scale implementation across the entire network, this application proposes a method based on KPI mapping to replace base station log data.

Step S220: performing sampling based on the sampling quantity to obtain the corrected log data.

Optionally, the determining, based on the KPI, a sampling quantity of log data of a preset training base station includes:

Step S211, based on the KPI, obtaining a first distribution corresponding to the communication network performance indicator;

Step S212, obtaining a second distribution corresponding to the first distribution based on the log data of the preset training base station;

Step S213: Determine the sampling quantity of the log data of the preset training base station according to the first distribution and the second distribution.

In one embodiment, the communication network performance indicator is path loss, and determining the number of samples of log data of a preset training base station based on the KPI includes:

Step S211, based on the KPI, obtaining a first path loss distribution with an interval of first loss;

Taking KPI as path loss as an example, determining the first path loss distribution of the KPI in multiple preset intervals may specifically be a path loss distribution with an interval of 5 dB, such as: dict_pl_kpi={[-90--95]: 20%, [-95--100]: 50%, [-100--105]: 30%}.

Step S212: obtaining a second path loss distribution corresponding to the first path loss distribution based on the log data of the preset training base station;

Taking KPI as path loss as an example, the second indicator distribution of log data in multiple path loss intervals can be the path loss distribution of base station log data at an interval of 5dB, such as dict_pl_log={[-90～95]:50% (5000 items), [-95～100]:30% (3000 items), [-100～105]:20% (2000 items)}.

Step S213: Determine the number of preset samples of the log data of the training base station based on the first path loss distribution and the second path loss distribution.

In one embodiment, determining the number of samples of log data of a preset training base station based on the first path loss distribution and the second path loss distribution includes:

Step S2131, determining a target path loss interval according to a ratio between the first path loss distribution and the second path loss distribution in the same path loss interval;

Specifically, the two distributions obtained by dict_pl_log and dict_pl_kpi are divided to obtain the following distributions: [-90～95]: 20%/50%=0.4, [-95～100]: 50%/30%=1.67, [-100～105]: 30%/20%=1.5, and the preset performance interval with the largest ratio is [-95～100].

Step S2132, determining the total number of samples of the log data of the preset training base station according to the distribution of the KPI in the target path loss interval and the number of samples of the log data of the preset training base station in the target path loss interval;

That is, the number of log data samples of the base station in the preset path loss interval with the largest ratio is divided by the ratio of the KPI to obtain the total number of log data samples, and the total number of samples = 3000/50% = 6000.

Step S2133: Determine the sampling number of the log data of the preset training base station in each path loss interval according to the total sampling number and the first path loss distribution.

Calculate the number of samples in each path loss interval and sample the log data. The sampled log samples are consistent with the KPI distribution. For example: [-90～95]: 6000*20%=1200; [-95～100]: 6000*50%=3000; [-100～105]: 6000*30%=1800.

Specifically, 1200 are randomly sampled from 5000 in the [-90～95] gear, 3000 are sampled from 3000 in the [-95-100] gear (no sampling), and 1800 are randomly sampled from 2000 in the [-100-105] gear. The final corrected base station log data is the test log data: dict_pl_log={[-90～95]: 20% (1200 items), [-5～100]: 50% (3000 items), [-100～105]: 30% (1800 items)}.

It can be understood that when modifying the parameters of the base station, the present application takes into account the accuracy and timeliness of model establishment, as well as the operability of large-scale implementation across the entire network, and maps the base station log data through KPI, thereby achieving the purpose of efficient and accurate parameter optimization and improving the accuracy and efficiency of parameter optimization.

The experimental results of this application are illustrated below.

Experiment 1

If you need to modify the waveform adaptation function parameters of a certain base station, the KPI is the path loss distribution. The waveform adaptation function parameters are divided into auto, CP-OFDM and DFT-S-OFDM. CP uplink supports dual streams, so it is suitable for areas where users in the cell center are concentrated. For edge users, DFT-S-OFDM is used. There is no power loss, but it only supports single stream, which is suitable for areas with weak coverage.

The waveform adaptation function parameters were modified at three base stations in the existing network. The gain change data before and after the modification and the gain change prediction value output by the network performance prediction model are shown in the following table.

Experiment 2

The PRB randomization function parameters are mainly used to solve the interference problem in low-load scenarios. The optimal parameters under different interference distributions and load distributions will be different, so the main correction is the distribution of interference and load. The parameter field name is: reserved250, which has three dimensions. The first dimension parameter meaning: This parameter is used to control whether to enter the optimization function of uplink enhanced PRB randomization based on the uplink cell coordinated allocation strategy. 0 means off and 1 means on. The second dimension parameter meaning: This parameter indicates the interval index of the randomized RB starting position when the uplink enhanced PRB randomization is turned on. That is, the number of RBs between the randomized RB starting positions is 2 to the power of this parameter. The current default value is set to 3, which means that there are 8 random RB starting positions in the three-segment cell. The third dimension parameter meaning: This parameter is the uplink PRB utilization threshold. When the utilization of this slot is lower than the parameter threshold, the uplink enhanced PRB randomization optimization function is performed.

The PRB randomization function parameters were modified in three base stations in the existing network. The gain change data before and after the modification and the gain change prediction value output by the network performance prediction model are shown in the following table.

Due to the complexity of the existing network environment, it is very likely that the parameter modification based on conventional principles will not be consistent with the actual situation. For example, in Industrial Park 11580000, after the optimization function of uplink enhanced PRB randomization was turned off, the uplink rate increased instead. This may be because there are many edge users and larger continuous RBs are required for output transmission. PRB randomization will affect performance. Therefore, according to the verification results, this algorithm can evaluate the effect of parameter modification without modifying the parameters, and the gain judgment accuracy is high.

Figure 4 is the second flow chart of the base station parameter optimization method provided by the present application. As shown in Figure 4, the base station parameter optimization method provided by the present application includes a training process and a testing process. The base station log data is selected as the data source for model establishment, and the KPI is selected as the data source used for parameter optimization. That is, a model is first trained through a small batch of base station log data. For base stations that need to be optimized, the KPI of the base station is first obtained, and then the distribution of the KPI in each dimension of the log data in the model is adjusted through an algorithm to obtain a new set of log data, and then the new log data is used to obtain the optimal parameters.

In the network, certain functions or parameters of the base station may have different performance in different business scenarios. Therefore, it is necessary to configure different parameters for different base stations to obtain better network performance. This application can automatically learn the relationship between different business scenarios, different parameters and network performance, and then use this model to estimate the performance of other base stations under different parameters, so as to achieve the purpose of parameter optimization. This application takes into account the operability of large-scale implementation of the entire network. The data source used for parameter optimization selects easy-to-access data indicators and combines the advantages of log data. Therefore, it is very likely to be used by the company, other operators and suppliers, and has good promotion prospects in the ZTE area. At the same time, since this application does not have too many restrictions on the underlying structure and logic, it also has good application prospects in other manufacturers' areas.

The base station parameter optimization device provided in the present application is described below. The base station parameter optimization device described below and the base station parameter optimization method described above can be referenced to each other.

FIG5 is a principle block diagram of a base station parameter optimization device provided by the present application. As shown in FIG5 , the present application further provides a base station parameter optimization device, including:

An acquisition module 510 is used to acquire a wireless key data indicator KPI of the base station to be optimized;

An adjustment module 520, configured to adjust the log data distribution of a preset training base station according to the KPI to obtain corrected log data;

The optimization module 530 is used to input the corrected log data into a pre-trained test model, predict the predicted base station performance of the base station to be optimized, and optimize the base station operating parameters of the base station to be optimized based on the prediction result, wherein the test model is trained based on the log data of a preset training base station.

It should be noted that the base station parameter optimization device provided by the present invention can execute the base station parameter optimization method described in any of the above embodiments during specific operation, and has the technical effect corresponding to the method, which will not be elaborated in this embodiment.

FIG6 illustrates a schematic diagram of a physical structure of an electronic device. As shown in FIG6 , the electronic device may include: a processor 610, a communications interface 620, a memory 630, and a communication bus 640, wherein the processor 610, the communications interface 620, and the memory 630 communicate with each other through the communication bus 640. The processor 610 may call the logic instructions in the memory 630 to execute the base station parameter optimization method, which includes:

Obtain the wireless key data indicators KPI of the base station to be optimized;

Adjust the log data distribution of the preset training base station according to the KPI to obtain the corrected log data;

The corrected log data is input into a pre-trained test model to predict the predicted base station performance of the base station to be optimized, and the base station operating parameters of the base station to be optimized are optimized based on the prediction result, wherein the test model is trained based on the log data of a preset training base station.

In addition, the logic instructions in the above-mentioned memory 630 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art, and the computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a disk or an optical disk.

On the other hand, the present application also provides a computer program product, the computer program product includes a computer program, the computer program can be stored on a non-transitory computer-readable storage medium, and when the computer program is executed by a processor, the computer can execute the base station parameter optimization method provided by the above methods, the method comprising:

Obtain the wireless key data indicators KPI of the base station to be optimized;

Adjust the log data distribution of the preset training base station according to the KPI to obtain the corrected log data;

The corrected log data is input into a pre-trained test model to predict the predicted base station performance of the base station to be optimized, and the base station operating parameters of the base station to be optimized are optimized based on the prediction result, wherein the test model is trained based on the log data of a preset training base station.

On the other hand, the present application also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which is implemented when the computer program is executed by a processor to execute the base station parameter optimization method provided by the above methods, the method comprising:

Obtain the wireless key data indicators KPI of the base station to be optimized;

Adjust the log data distribution of the preset training base station according to the KPI to obtain the corrected log data;

The corrected log data is input into a pre-trained test model to predict the predicted base station performance of the base station to be optimized, and the base station operating parameters of the base station to be optimized are optimized based on the prediction results, wherein the test model is trained based on the log data of a preset training base station.

The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM/RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims (11)

1. A method for optimizing base station parameters, comprising: Obtain the wireless key data indicators KPI of the base station to be optimized; Adjust the log data distribution of the preset training base station according to the KPI to obtain the corrected log data; The corrected log data is input into a pre-trained test model to predict the predicted base station performance of the base station to be optimized, and the base station operating parameters of the base station to be optimized are optimized based on the prediction result, wherein the test model is trained based on the log data of a preset training base station. 2. The base station parameter optimization method according to claim 1, wherein the step of adjusting the log data distribution of the preset training base station according to the KPI to obtain the corrected log data comprises: Based on the KPI, determine the sampling quantity of the log data of the preset training base station; Sampling is performed based on the sampling quantity to obtain the corrected log data. 3. The base station parameter optimization method according to claim 2, wherein the step of determining the number of samples of log data of a preset training base station based on the KPI comprises: Based on the KPI, obtaining a first distribution corresponding to a communication network performance indicator; Based on the log data of the preset training base station, obtaining a second distribution corresponding to the first distribution; According to the first distribution and the second distribution, a preset number of samples of log data of the training base station is determined. 4. The base station parameter optimization method according to claim 3, wherein the communication network performance indicator is path loss, and the step of determining the number of samples of log data of a preset training base station based on the KPI comprises: Based on the KPI, a first path loss distribution with an interval of first loss is obtained; Based on the log data of the preset training base station, obtaining a second path loss distribution corresponding to the first path loss distribution; Based on the first path loss distribution and the second path loss distribution, a sampling quantity of log data of a preset training base station is determined. 5. The base station parameter optimization method according to claim 4, characterized in that the step of determining the sampling quantity of the log data of the preset training base station based on the first path loss distribution and the second path loss distribution comprises: Determining a target road loss interval according to a ratio between the first road loss distribution and the second road loss distribution in the same road loss interval; Determine the total number of samples of the log data of the preset training base station according to the distribution of the KPI in the target path loss interval and the number of samples of the log data of the preset training base station in the target path loss interval; According to the total sampling number and the first path loss distribution, the sampling number of the log data of the preset training base station in each path loss interval is determined. 6. The base station parameter optimization method according to claim 1 is characterized in that the test model trains a preset boosting tree model based on log data of a preset training base station. 7. The base station parameter optimization method according to claim 1, wherein the step of obtaining the wireless key data indicator KPI of the base station to be optimized comprises: Acquire the wireless key data indicator KPI corresponding to the to-be-optimized parameter of the to-be-optimized base station based on the gateway; Correspondingly, the inputting the corrected log data into the pre-trained test model, predicting the predicted base station performance of the base station to be optimized, and optimizing the base station operating parameters of the base station to be optimized based on the prediction result, includes: Inputting the modified log data into a pre-trained test model to obtain a predicted base station performance corresponding to any parameter value of the parameter to be optimized; The base station operating parameters are configured to be parameter values corresponding to the optimal predicted base station performance to complete base station parameter optimization. 8. A base station parameter optimization device, characterized by comprising: An acquisition module is used to acquire the wireless key data indicators KPI of the base station to be optimized; An adjustment module, used for adjusting the log data distribution of a preset training base station according to the KPI to obtain corrected log data; An optimization module is used to input the corrected log data into a pre-trained test model, predict the predicted base station performance of the base station to be optimized, and optimize the base station operating parameters of the base station to be optimized based on the prediction result, wherein the test model is trained based on the log data of a preset training base station. 9. An electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the base station parameter optimization method as described in any one of claims 1 to 7 is implemented. 10. A non-transitory computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the base station parameter optimization method according to any one of claims 1 to 7 is implemented. 11. A computer program product, comprising a computer program, characterized in that when the computer program is executed by a processor, the base station parameter optimization method according to any one of claims 1 to 7 is implemented.