CN114268961A

CN114268961A - Model training method, parameter adjusting method, device and equipment of shared base station

Info

Publication number: CN114268961A
Application number: CN202111493013.3A
Authority: CN
Inventors: 张国鹏; 谷俊江; 张进; 赵煜; 周奕昕; 祝海亮; 孙宏; 盛莉莉; 薛金明; 赵欢; 金立标; 支亚光
Original assignee: China United Network Communications Group Co Ltd
Current assignee: China United Network Communications Group Co Ltd
Priority date: 2021-12-08
Filing date: 2021-12-08
Publication date: 2022-04-01

Abstract

The application provides a model training method, a parameter adjusting device and a parameter adjusting device of a shared base station, wherein the model training method comprises the following steps: obtaining a plurality of training samples, wherein the training samples comprise sample service indexes of a sample base station; obtaining label data corresponding to the sample service index, wherein the label data is used for representing expected configuration parameters corresponding to the sample base station; inputting the training sample into a target network model to obtain a training output configuration parameter; and obtaining a trained target network model according to the expected configuration parameters and the training output configuration parameters. In the embodiment of the application, the sample service index of the sample base station is used as the training sample, the expected configuration parameter of the sample base station is used as the label data, and the trained target network model can be used for adjusting the configuration parameter of the shared base station, so that the signal levels of the shared base station with the adjusted configuration parameter for users of different communication operators are consistent, and the configuration efficiency of the configuration parameter of the shared base station is further improved.

Description

Model training method, parameter adjusting method, device and equipment of shared base station

Technical Field

The present application relates to the field of computers, and in particular, to a model training method, a parameter adjusting method for a shared base station, an apparatus and a device thereof.

Background

In order to achieve fast full coverage of network services, base stations are shared among operators. Under the mode of sharing the base station, the increase of the network speed, the network coverage and the network bandwidth can be quickly realized. In addition, the network efficiency can be greatly improved.

The shared base station has different service effects for users of different operators. For example, the shared base station provided by operator a has better service effect for the user of operator a and worse service effect for the user of operator B. For the above problems, the currently adopted method manually adjusts the configuration parameters of the shared base station for multiple times until the service effects of the shared base station on users of different operators are consistent, and the method has the problem of low efficiency, thereby affecting the sharing progress and the sharing scale of the shared base station.

Disclosure of Invention

The application provides a model training method, a parameter adjusting device and parameter adjusting equipment of a shared base station, and aims to solve the problem that in the prior art, the efficiency is low by manually adjusting configuration parameters of the shared base station for multiple times.

In a first aspect, the present application provides a model training method, including:

obtaining a plurality of training samples, wherein the training samples comprise sample service indexes of a sample base station;

obtaining label data corresponding to the sample service index, wherein the label data is used for representing expected configuration parameters corresponding to the sample base station;

inputting the training sample into a target network model to obtain a training output configuration parameter;

and adjusting the target network model according to the expected configuration parameters and the training output configuration parameters until the difference value between the expected configuration parameters and the training output configuration parameters is smaller than a threshold value, so as to obtain the trained target network model.

In one embodiment of the present application, a sample base station includes: a sample sharing base station for serving a first device of a first communications carrier and a second device of a second communications carrier; the desired configuration parameter is the configuration parameter of the sample shared base station at which the signal levels of the first and second devices are made consistent.

In one embodiment of the present application, a sample base station includes: the first sample base station, where the first sample base station is configured to serve a first device of a first communications carrier, is configured to expect the configuration parameter to be the configuration parameter of the first sample base station when a signal level of the first device is greater than a first signal level threshold and a signal level of a second device of a second communications carrier is less than a second signal level threshold, where the first signal level threshold is greater than a second signal level threshold.

In one embodiment of the present application, a sample base station includes: a second sample base station, configured to serve a second device of a second communications operator, where the desired configuration parameter is a configuration parameter of the second sample base station that causes a signal level of the second device to be greater than a third signal level threshold, and causes the signal level of the first device of the first communications operator to be less than a fourth signal level threshold, where the third signal level threshold is greater than a fourth signal level threshold.

In a second aspect, the present application provides a parameter adjusting method for a shared base station, where the shared base station is used to serve a first device of a first communications operator and a second device of a second communications operator, and the parameter adjusting method for the shared base station includes: acquiring a first service index of a shared base station; inputting the first service index into a target network model obtained by training in any one of the first aspect to obtain a target configuration parameter, wherein the target network model is used for determining the configuration parameter of the shared base station as the target configuration parameter, and when the shared base station configures the target configuration parameter, the signal level of the first device is consistent with the signal level of the second device; and adjusting the configuration parameters of the shared base station according to the target configuration parameters.

In an embodiment of the present application, before obtaining the first service index of the shared base station, the method further includes: acquiring a second service index of the current base station; determining the target score of the current base station by adopting an entropy weight algorithm according to the second service index; and determining the current base station as a shared base station according to the target score.

In an embodiment of the present application, determining, according to the target score, that the current base station is a shared base station includes: determining the sharable priority level of the current base station according to the target score and the score threshold; and if the priority level is greater than the level threshold, determining the current base station as the shared base station.

In a third aspect, the present application provides a model training apparatus, comprising:

the device comprises a first acquisition module, a second acquisition module and a third acquisition module, wherein the first acquisition module is used for acquiring a plurality of training samples, and the training samples comprise sample service indexes of a sample base station;

the second obtaining module is used for obtaining label data corresponding to the sample service index, and the label data is used for representing an expected configuration parameter corresponding to the sample base station;

the input module is used for inputting the training samples into the target network model to obtain training output configuration parameters;

and the adjusting module is used for adjusting the target network model according to the expected configuration parameters and the training output configuration parameters until the difference value between the expected configuration parameters and the training output configuration parameters is smaller than a threshold value, so as to obtain the trained target network model.

In a fourth aspect, the present application provides a parameter adjusting apparatus for a shared base station, where the shared base station is configured to serve a first device of a first communications carrier and a second device of a second communications carrier, and the parameter adjusting apparatus for the shared base station includes:

the acquisition module is used for acquiring a first service index of the shared base station;

an input module, configured to input the first service indicator into the target network model obtained through training in claim 8, to obtain a target configuration parameter, where the target network model is configured to determine that the configuration parameter of the shared base station is the target configuration parameter, and when the shared base station configures the target configuration parameter, a signal level of the first device is consistent with a signal level of the second device;

and the adjusting module is used for adjusting the configuration parameters of the shared base station according to the target configuration parameters.

In an embodiment of the present application, the parameter adjusting apparatus for sharing a base station further includes:

the service index acquisition module is used for acquiring a second service index of the current base station;

the first determining module is used for determining the target score of the current base station by adopting an entropy weight algorithm according to the second service index;

and the second determining module is used for determining the current base station as the shared base station according to the target score.

In an embodiment of the application, the second determining module is specifically configured to: determining the sharable priority level of the current base station according to the target score and the score threshold; and if the priority level is greater than the level threshold, determining the current base station as the shared base station.

In a fifth aspect, the present application provides an electronic device, comprising:

at least one processor; and

a memory communicatively coupled to the at least one processor; wherein the content of the first and second substances,

the memory stores instructions executable by the at least one processor to enable the electronic device to perform the model training method of any one of the first aspects or the parameter tuning method of the shared base station of any one of the second aspects.

In a sixth aspect, the present application provides a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements the model training method of any one of the first aspect or the parameter tuning method of the shared base station of any one of the second aspect of the present application.

In a seventh aspect, the present application provides a computer program product comprising a computer program, which when executed by a processor, implements the model training method of any one of the first aspect or the parameter tuning method of any one of the shared base stations of the second aspect of the present application.

The application provides a model training method, a parameter adjusting device and a parameter adjusting device of a shared base station, wherein the model training method comprises the following steps: obtaining a plurality of training samples, wherein the training samples comprise sample service indexes of a sample base station; obtaining label data corresponding to the sample service index, wherein the label data is used for representing expected configuration parameters corresponding to the sample base station; inputting the training sample into a target network model to obtain a training output configuration parameter; and adjusting the target network model according to the expected configuration parameters and the training output configuration parameters until the difference value between the expected configuration parameters and the training output configuration parameters is smaller than a threshold value, so as to obtain the trained target network model. In the embodiment of the application, the sample service index of the sample base station is used as the training sample, the expected configuration parameter of the sample base station is used as the label data, and the trained target network model can be used for adjusting the configuration parameter of the shared base station, so that the signal levels of the shared base station with the adjusted configuration parameter for users of different communication operators are consistent, and the configuration efficiency of the configuration parameter of the shared base station is further improved.

Drawings

In order to more clearly illustrate the technical solutions of the present application or the prior art, the drawings that are needed in the description of the prior art or the prior art are briefly introduced below, and it is obvious that the drawings in the following description are some of the present application, and other drawings can be obtained by those skilled in the art without inventive exercise.

FIG. 1 is a schematic view of a model training method provided in the present application;

FIG. 2 is a flow chart illustrating steps of a model training method provided herein;

FIG. 3 is a schematic diagram of a target network model provided herein;

fig. 4 is a flowchart illustrating steps of a method for tuning parameters of a shared base station according to the present application;

FIG. 5 is a schematic structural diagram of a model training device provided in the present application;

fig. 6 is a schematic structural diagram of a parameter adjusting apparatus for sharing a base station according to the present application;

fig. 7 is a schematic diagram of a hardware structure of an electronic device according to a first embodiment of the present disclosure.

The present application is made clear by the foregoing drawings and will be described in more detail hereinafter. These drawings and written description are not intended to limit the scope of the inventive concepts in any manner, but rather to illustrate the inventive concepts by reference to specific details that are particularly well within the skill of those in the art.

Detailed Description

To make the objects, aspects and advantages of the present application clearer, the technical solutions in the present application will be clearly and completely described below with reference to the accompanying drawings in the present application, and it is obvious that the description is a part of the present application, but not all of it. All others that would be obvious to one of ordinary skill in the art based on this disclosure, without any creative effort, are within the scope of this disclosure.

The terms "first," "second," "third," and the like in the description and in the claims of the present application and in the above-described drawings are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used is interchangeable under appropriate circumstances such that, for example, the applications described herein are capable of operation in sequences other than those illustrated or otherwise described herein.

Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.

The inventor finds that, at present, different communication operators vigorously develop 5G co-construction sharing work to co-construct a 5G network, and the 5G network deployment is promoted by the resultant force, so as to realize 5G coverage of each area. However, there are several challenges faced in 5G co-construction sharing. For example, due to different QoS (Quality of Service) policies of different communication carriers, coverage of the shared base station may differ, and the like, signal Quality of mobile devices corresponding to different communication carriers that share the Service of the base station may have a large difference.

In the related technology, when determining whether the configuration parameters of the shared base station cause large differences of the signal quality of the mobile devices corresponding to different communication operators, the method is adopted, after a user complains the network quality, according to a complaint work order of the user, the complaint work order is sent to roadside staff and dial-up staff in a corresponding area, the roadside staff and the dial-up staff measure the network quality indexes of the area to which the complaint user belongs, then analysis staff analyze the measured data, and the problem of inconsistent signal levels of the mobile devices of the different communication operators is comprehensively judged by manually checking a large number of QoS strategies, DT (road test) data, MR (measurement report) data, network cfg (configuration) files, PM (network performance indexes) and other data, so as to adjust the configuration parameters of the shared base station.

In the related art, the method for judging the problem of inconsistent signal levels of the mobile devices of different communication operators needs the participation of workers of multiple work types, and has the disadvantages of low efficiency, poor accuracy and high error rate. In addition, a communication operator is required to invest a large amount of human resources to perform data docking, process communication, data extraction and analysis, and the like. Due to the fact that a plurality of workers are involved, a plurality of management links are provided, the difference of levels of the workers is analyzed, the interaction time is delayed, communication among the workers is limited, human resources are insufficient, acquired data are lost, evaluation tendencies of different communication operators are different, and the like, the problem that evaluation results of signal levels of mobile devices of different communication operators are inconsistent is not practical, and further the network quality of the shared base station cannot meet user requirements of different communication operators.

Moreover, sharing of the base station is also determined by human judgment, such as "a certain site can be shared", "a part of sites in a certain area can be shared", "this is a special scene and does not have shared resources", and the like, and a unified evaluation system is lacked, so that the sharing progress and the sharing scale of the base station are also seriously hindered.

The model training method provided by the embodiment of the application can be used for adjusting the configuration parameters of the shared base station by using the sample service index of the sample base station as a training sample and using the expected configuration parameters of the sample base station as label data, so that the signal levels of the shared base station with the adjusted configuration parameters for users of different communication operators are consistent, and the configuration efficiency of the configuration parameters of the shared base station is further improved.

Before introducing the model training method provided by the present application, an application scenario of the model training method is briefly introduced.

Fig. 1 is a schematic view of a scenario of a model training method provided in this application, and as shown in fig. 1, the scenario includes a sample base station 11, a first device a and a second device B in an area 12 served by the sample base station 11, and a management device 13. The first device a is a device corresponding to a first communications carrier, the second device B is a device of a second communications carrier, and the sample base station 11 may be a device provided by the first communications carrier or the second communications carrier. If the sample base station 11 is a shared base station, the service index of the sample base station 11 currently serving the first device a and the second device B is sent to the management device 13 as a sample service index, and then the parameters of the sample base station are adjusted, so that the configuration parameters of the sample base station when the signal levels of the first device a and the second device B are consistent are sent to the management device 13 as expected configuration parameters. The management device 13 takes the sample service indexes uploaded by the plurality of sample base stations 11 as training samples, takes the expected configuration parameters as label data, trains a target network model, and obtains the trained target network model. When the management device 13 receives the target service index sent by the other shared base station, the target service index may be input into the target network model to obtain the target configuration parameter of the other shared base station, and the target configuration parameter of the other shared base station may enable the signal levels of the first device a and the second device B to be consistent when the other shared base station serves the first device a and the second device B.

In connection with the application scenario of fig. 1, a method for model training according to an exemplary embodiment of the present application is described with reference to fig. 2. It should be noted that the above application scenarios are only presented to facilitate understanding of the spirit and principles of the present application, and the embodiments of the present application are not limited in this respect. Rather, embodiments of the present application may be applied to any scenario where applicable.

Fig. 2 is a flowchart of a model training method provided in the present application. The model training method specifically comprises the following steps:

s201, a plurality of training samples are obtained.

Wherein the training samples comprise sample service indexes of the sample base station. Specifically, the sample service index includes five types of indexes. The method specifically comprises the following steps: coverage class, load class, quality class, interference class, and user perception class. Wherein the overlay class includes: the downlink RSRP is less than the proportion (%) of-110, the uplink physical shared channel (PUSCH) RSRP average level (dBm) and the TA proportion (%) of the ultra-far (> 1 KM). The load classes include: the average occupancy rate (%) of the single-board CPU, the average number of users, the average utilization rate (%) of uplink PRB, and the average utilization rate (%) of downlink PRB. The quality classes include: the rate (%) of the CQI (channel quality indicator), the rate (%) of the cell service, the rate (%) of the call completing, the rate (%) of the disconnection, and the rate (%) of the 5G backflow 4G. The interference classes include: the interference noise mean (dBm) is received per PRB on the uplink. The user perception classes include: the average uplink rate (Mbps) of user equipment in a cell, the average downlink rate (Mbps) of the user equipment in the cell, the average uplink MCS (modulation and coding strategy) and the average downlink MCS.

In addition, the sample service index is data generated when the sample base station adopts the current configuration parameter for service.

In the present application, a training sample corresponds to a sample service indicator of a sample base station. In the embodiment of the application, a plurality of training samples are needed to train the target network model.

S202, label data corresponding to the sample service index is obtained.

The label data is used for representing expected configuration parameters corresponding to the sample base station.

In an alternative embodiment, the sample base station comprises: a sample sharing base station for serving a first device of a first communications carrier and a second device of a second communications carrier; the desired configuration parameter is the configuration parameter of the sample shared base station at which the signal levels of the first and second devices are made consistent.

Specifically, the expected configuration parameter refers to a signal level of the first device and a signal level of the second device are consistent when the configuration parameter of the sample base station is adjusted from the current configuration parameter to the expected configuration parameter.

In an optional embodiment, the sample base station further comprises: the first sample base station, where the first sample base station is configured to serve a first device of a first communications carrier, is configured to expect the configuration parameter to be the configuration parameter of the first sample base station when a signal level of the first device is greater than a first signal level threshold and a signal level of a second device of a second communications carrier is less than a second signal level threshold, where the first signal level threshold is greater than a second signal level threshold.

Specifically, the desired configuration parameter refers to the configuration parameter of the first sample base station when the signal level of the first device is greater than the first signal level threshold and the signal level of the second device of the second communications operator is less than the second signal level threshold when the configuration parameter of the sample base station is adjusted from the current configuration parameter to the desired configuration parameter. If the signal levels are 0, 0.2, 0.4, 0.6, 0.8 and 1, the signal level 0 indicates no signal, the signal level 0.2 indicates very weak signal, the signal level 0.4 indicates weak signal, the signal level 0.6 indicates strong signal, the signal level 0.8 indicates strong signal, and the signal level 1 indicates very strong signal. The signal level in the embodiment of the present application may be determined according to the call effect, the network effect, and the like of the first device and the second device.

Illustratively, the first signal level threshold may be 0.6 and the second signal level threshold may be 0.2.

Wherein, the sample base station includes: a second sample base station, configured to serve a second device of a second communications operator, where the desired configuration parameter is a configuration parameter of the second sample base station that causes a signal level of the second device to be greater than a third signal level threshold, and causes the signal level of the first device of the first communications operator to be less than a fourth signal level threshold, where the third signal level threshold is greater than a fourth signal level threshold.

Illustratively, the third signal level threshold may be 0.6 and the fourth signal level threshold may be 0.2.

In addition, in the embodiment of the present application, the training samples and the expected configuration parameters corresponding to the sample sharing base station may be used as data for training the target network model. The training sample and the expected configuration parameter corresponding to the first sample base station or the second sample base station may be used as data for testing a trained target network model.

S203, inputting the training sample into the target network model to obtain a training output configuration parameter.

The target neural network is structured as a Back Propagation (BP) neural network, which is a multi-layer feedforward neural network trained according to an error back propagation algorithm. In addition, the BP neural network is divided into a forward propagation stage and an error backward propagation stage, a Sigmoid activating function is adopted to calculate training output configuration parameters during the forward propagation, and then a gradient descent method is adopted to perform backward propagation according to the difference value between the training output configuration parameters and expected configuration parameters, so that the model parameters of the target network model are updated.

Specifically, referring to fig. 3, the target network model includes an input layer 31, a hidden layer 32, and an output layer 33. Wherein the input layer 31 includes a plurality of input nodes (a1 to An), n being An integer greater than 1, each input node inputting different sample data in the training samples. The hidden layer 32 includes a plurality of hidden nodes (B1 to Bm), m is an integer greater than 1, and data output from the input layer 31 enters the hidden layer 32. The output layer 33 includes a plurality of output nodes (C1 to Cp), p is an integer greater than or equal to 1, and the data output by the hidden layer 32 enters the output layer and is processed by the output layer to output the training output configuration parameters.

In fig. 3, the training samples are preprocessed to obtain input data X ═ (X)₁，x₂，…，x_n)^TWherein x is_iInput corresponding input node A_iAnd i is 1 to n. Wk ═ w_k1，w_k2，…，w_km) Is the connection weight of the kth hidden node in the hidden layer 32 to the input layer 31. Wo ═ w₁，w₂，…，w_m) Weights are connected to the output layer 33 and the hidden nodes. Ho ═ h₁，h₂，…，h_m) The output value of each hidden node of the hidden layer. B ═ B₁，b₂，…，b_m) A threshold corresponding to each hidden node. bo is the threshold of each output node of the output layer. The hidden layer output is Ho ═ f₁(w_kx-b_k) Wherein f is₁Is the activation function (Sigmoid) of the hidden layer 32. The training output configuration parameter output by the output layer is O ═ f₂(w_oH_o-b_o) Wherein f is₂Is the activation function (Sigmoid) of the output layer 33.

And S204, adjusting the target network model according to the expected configuration parameters and the training output configuration parameters until the difference value between the expected configuration parameters and the training output configuration parameters is smaller than a threshold value, and obtaining the trained target network model.

The adjustment target network model is substantially a model parameter of the adjustment target network model, and specifically, when a difference value between a desired configuration parameter and a training output configuration parameter is greater than a threshold value, the connection weight between the hidden node and the input layer, the connection weight between the output layer and the hidden node, the threshold value of each hidden node, and the threshold value of each output node are adjusted through back propagation. And obtaining the trained target network model until the difference value between the expected configuration parameters and the training output configuration parameters is smaller than the threshold value.

The model training method provided by the application comprises the following steps: obtaining a plurality of training samples, wherein the training samples comprise sample service indexes of a sample base station; obtaining label data corresponding to the sample service index, wherein the label data is used for representing expected configuration parameters corresponding to the sample base station; inputting the training sample into a target network model to obtain a training output configuration parameter; and adjusting the target network model according to the expected configuration parameters and the training output configuration parameters until the difference value between the expected configuration parameters and the training output configuration parameters is smaller than a threshold value, so as to obtain the trained target network model. In the embodiment of the application, the sample service index of the sample base station is used as the training sample, the expected configuration parameter of the sample base station is used as the label data, and the trained target network model can be used for adjusting the configuration parameter of the shared base station, so that the signal levels of the shared base station with the adjusted configuration parameter for users of different communication operators are consistent, and the configuration efficiency of the configuration parameter of the shared base station is further improved.

Fig. 4 is a flowchart illustrating steps of a method for tuning a shared base station according to an embodiment of the present invention. The shared base station is used for serving a first device of a first communication operator and a second device of a second communication operator, and the parameter adjusting method of the shared base station specifically comprises the following steps:

s401, a first service index of the shared base station is obtained.

Before obtaining the first service index of the shared base station, the method further includes: acquiring a second service index of the current base station; determining the target score of the current base station by adopting an entropy weight algorithm according to the second service index; and determining the current base station as a shared base station according to the target score.

Specifically, the current base station is a currently unshared base station. The base station serves only the first communication carrier or the second communication carrier. The second service index of the current base station comprises: three major indexes are load capacity, throughput and perceptibility. Wherein the load capacity comprises: the method comprises the steps of single-user perception rate, online activated user number and wireless resource utilization rate. The throughput includes: three high user traffic (GB), single cell downlink traffic (GB), "720 p + video traffic (GB)" and "71080 p + video traffic (GB)". The perceptibility comprises: game stuck rate (%) and video stuck rate (%).

Illustratively, the single user perception rate refers to { total throughput of downlink data transmitted by a cell PDCP (radio network layer and transport network layer) layer minus downlink PDCP throughput transmitted by the last TTI that has the buffer left empty }/(data transmission duration after the last TTI (transmission time interval) that has the buffer left empty)/{ 1000 }. The number of online active users refers to the average number of active users. The radio resource utilization ratio is the average number of used downlink PRBs (physical resource blocks)/the number of PRBs available for downlink/{ 0.9}/{0.6}/{0.5} {100 }. The three-high user traffic refers to high ARPU (income) user traffic, high DOU (monthly internet traffic per household) and high end user traffic. The single-cell downlink traffic (GB) refers to the total throughput/({ 1000 }. multidot {1000 }) of downlink data transmitted by the cell PDCP layer.

In the embodiment of the application, the entropy is a measure of uncertainty of data, and the larger the data amount is, the higher the intelligibility is, the smaller the entropy is, and conversely, the smaller the data amount is, the larger the entropy is. The entropy weight algorithm is introduced, the second service indexes of all the base stations are scored uniformly, and the base stations which can serve as the shared base stations are determined according to the final target scores of the base stations. And the base station recommended by the communication operator and the key base station can be subjected to scoring calculation.

In addition, the entropy weight algorithm is specifically as follows: the data X1, X2, … and Xk of the second service index are normalized to obtain normalized data Y1, Y2, … and Yk. Wherein the content of the first and second substances,

wherein j is 1 to k, min (X) in sequence_i) Refers to X1, X2, …, X_kMinimum data of (1), max (X)_m) Means X₁，X₂，…，X_kThe largest data in the data. Information entropy corresponding to second service index

Wherein the content of the first and second substances,

wherein, if P_jWhen 0, then

Determining each data X according to the information entropy₁，X₂，…，X_kCorresponding information entropy is respectively E₁，E₂，…，E_k. Then, the weight corresponding to each data is obtained:

and i takes 1 to k in sequence.

Illustratively, referring to table 1, the weight values corresponding to the three types of indicators and the weight corresponding to each data in each type of indicator are obtained in the above manner. Wherein the sum of the weights of the data under each index is 100%. The sum of the weights of the indexes is 100 percent.

Specifically, the weight corresponding to each data under each type of index may be calculated based on each type of index. The weight of each type of index is then calculated.

TABLE 1

Referring to table 2, for the entropy corresponding to each data, the score corresponding to each index may be calculated according to the sum of products of the weight corresponding to each data and the corresponding entropy, and the target score of the current base station is obtained according to the sum of products of the weight corresponding to each index and the score corresponding to each index.

For example, if the entropy value corresponding to the single user perception rate is 100, the entropy value corresponding to the number of online active users is 60, and the entropy value corresponding to the radio resource utilization rate is 30, the score corresponding to the load amount is 28.26% × 100+ 35.25% × 60+ 36.49% × 30, and is 60.357. If the entropy value corresponding to the three-highest user traffic is 60, the entropy value corresponding to the single-cell downlink traffic is 30, the entropy value corresponding to the 720p + video traffic is 100, and the entropy value corresponding to the 1080p + video traffic is 60, the score corresponding to the throughput is 23.59% × 60+ 22% + 30.65% × 100+ 24.76% × 60, which is 65.26. If the entropy value for the game kation rate is 60 and the entropy value for the video kation rate is 30, the perceptibility score is S3 ═ 50% × 60+ 50% × 30, and is 45. The goal score S of the previous base station is 50% S1+ 40% S2+ 10% S3, which is 60.78.

In the embodiment of the application, entropy values corresponding to reference value intervals of a plurality of data can be preset, and then a target score of the current base station is determined, so as to accurately determine whether the current base station can be used as a shared base station.

TABLE 2

Further, determining the current base station as a shared base station according to the target score includes: determining the sharable priority level of the current base station according to the target score and the score threshold; and if the priority level is greater than the level threshold, determining the current base station as the shared base station.

For example, if the target score is greater than or equal to 60 points, the current base station is determined to be a non-shared base station. And if the target score is greater than or equal to 50 points and less than 60 points, determining the current base station as the negotiable shared base station. And if the target score is greater than or equal to 40 points and less than 50 points, determining the current base station as the sharable base station. And if the target score is greater than or equal to 30 points and less than 40 points, determining the current base station as a priority sharing base station.

S402, inputting the first service index into any one of the trained target network models to obtain a target configuration parameter.

The target network model is used for determining that the configuration parameters of the shared base station are target configuration parameters, and when the shared base station configures the target configuration parameters, the signal level of the first device is consistent with the signal level of the second device.

S403, adjusting the configuration parameters of the shared base station according to the target configuration parameters.

According to the method and the device, the target network model can be used for adjusting the configuration parameters of the shared base station, so that the signal levels of the shared base station with the adjusted configuration parameters to users of different communication operators are consistent, and the configuration efficiency of the configuration parameters of the shared base station is improved. In addition, whether the current base station can be shared can be determined through an entropy weight algorithm, and the sharing progress and the sharing scale of the shared base station are further promoted.

Fig. 5 is a block diagram of a model training apparatus 50 according to a first embodiment of the present invention. As shown in fig. 5, the present application provides a model training apparatus 50 including: a first obtaining module 51, a second obtaining module 52, an input module 53 and an adjusting module 54. Wherein:

a first obtaining module 51, configured to obtain a plurality of training samples, where a training sample includes a sample service index of a sample base station;

a second obtaining module 52, configured to obtain tag data corresponding to the sample service indicator, where the tag data is used to indicate an expected configuration parameter corresponding to the sample base station;

an input module 53, configured to input the training samples into the target network model to obtain training output configuration parameters;

and an adjusting module 54, configured to adjust the target network model according to the expected configuration parameter and the training output configuration parameter until a difference between the expected configuration parameter and the training output configuration parameter is smaller than a threshold value, so as to obtain a trained target network model.

The model training device provided in the present application is used to execute the technical solution in the method corresponding to fig. 2, and the implementation principle and the technical effect are similar, which are not described herein again.

Fig. 6 is a block diagram of a parameter adjusting apparatus 60 for sharing a base station according to a first embodiment of the present invention. As shown in fig. 6, the parameter adjusting apparatus 60 of the shared base station provided in the present application includes: an acquisition module 61, an input module 62 and an adjustment module 63. Wherein:

an obtaining module 61, configured to obtain a first service index of a shared base station;

an input module 62, configured to input the first service indicator into the target network model obtained by the training of claim 8, to obtain a target configuration parameter, where the target network model is configured to determine that the configuration parameter of the shared base station is the target configuration parameter, and when the shared base station configures the target configuration parameter, the signal level of the first device is consistent with the signal level of the second device;

and an adjusting module 63, configured to adjust the configuration parameters of the shared base station according to the target configuration parameters.

a service index obtaining module (not shown) for obtaining a second service index of the current base station;

a first determining module (not shown) for determining a target score of the current base station by using an entropy weight algorithm according to the second service index;

and a second determining module (not shown) for determining the current base station as the shared base station according to the target score.

The parameter adjusting apparatus for a shared base station provided in the present application is configured to execute the technical scheme in the method corresponding to fig. 4, and the implementation principle and the technical effect are similar, which are not described herein again.

Fig. 7 is a schematic diagram of a hardware structure of an electronic device provided in the present application. As shown in fig. 7, the electronic device 70 of the present application may include: at least one processor 71 (only one processor is shown in FIG. 7); and a memory 72 communicatively coupled to the at least one processor. The memory 72 stores instructions executable by the at least one processor 71, and the instructions are executed by the at least one processor 71, so that the electronic device 70 can execute any one of the foregoing methods.

Alternatively, the memory 72 may be separate or integrated with the processor 71.

When the memory 72 is a device separate from the processor 71, the electronic device 70 further includes: a bus 73 for connecting the memory 72 and the processor 71.

The electronic device provided by the application can execute the technical scheme of any one of the methods, and the implementation principle and the technical effect are similar, and are not described herein again.

The present application further provides a computer-readable storage medium, in which a computer program is stored, and when the computer program is executed by a processor, the computer program is used to implement the technical solution in any one of the foregoing methods.

The present application provides a computer program product comprising a computer program which, when executed by a processor, implements an aspect of any of the methods described above.

The present application also provides a chip, including: and the processing module and the communication interface can execute the technical scheme in the method.

Further, the chip further includes a storage module (e.g., a memory), where the storage module is configured to store instructions, and the processing module is configured to execute the instructions stored in the storage module, and the execution of the instructions stored in the storage module causes the processing module to execute the technical solution in the foregoing method.

It should be understood that the Processor may be a Central Processing Unit (CPU), other general purpose Processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), etc. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. The steps of a method disclosed in connection with the present invention may be embodied directly in a hardware processor, or in a combination of the hardware and software modules within the processor.

The memory may comprise a high-speed RAM memory, and may further comprise a non-volatile storage NVM, such as at least one disk memory, and may also be a usb disk, a removable hard disk, a read-only memory, a magnetic or optical disk, etc.

The bus may be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, an Extended ISA (EISA) bus, or the like. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, the buses in the figures of the present application are not limited to only one bus or one type of bus.

The storage medium may be implemented by any type or combination of volatile or non-volatile memory devices, such as Static Random Access Memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic or optical disks. A storage media may be any available media that can be accessed by a general purpose or special purpose computer.

An exemplary storage medium is coupled to the processor such the processor can read information from, and write information to, the storage medium. Of course, the storage medium may also be integral to the processor. The processor and the storage medium may reside in an Application Specific Integrated Circuits (ASIC). Of course, the processor and the storage medium may reside as discrete components in an electronic device.

Finally, it should be noted that: the above embodiments are merely illustrative of the technical solutions of the present application, and not restrictive; although the present application has been described in detail with reference to the foregoing aspects, those of ordinary skill in the art will understand that: the technical solutions described above may still be modified, or some or all of the technical features may be equivalently replaced; such modifications or substitutions do not depart from the scope of the present disclosure.

Claims

1. A method of model training, comprising:

obtaining label data corresponding to the sample service index, wherein the label data is used for representing an expected configuration parameter corresponding to the sample base station;

2. The model training method of claim 1, wherein the sample base station comprises: a sample sharing base station to serve a first device of a first communications carrier and a second device of a second communications carrier; the desired configuration parameter is a configuration parameter of the sample shared base station at which the signal levels of the first device and the second device are made consistent.

3. The model training method of claim 2, wherein the sample base station comprises: a first sample base station, the first sample base station configured to serve a first device of a first communications operator, the desired configuration parameter being the configuration parameter of the first sample base station when the signal level of the first device is greater than a first signal level threshold and the signal level of a second device of a second communications operator is less than a second signal level threshold, the first signal level threshold being greater than the second signal level threshold.

4. The model training method of claim 2 or 3, wherein the sample base station comprises: a second sample base station, configured to serve a second device of a second communications operator, where the desired configuration parameter is a configuration parameter of the second sample base station when a signal level of the second device is greater than a third signal level threshold and a signal level of a first device of the first communications operator is less than a fourth signal level threshold, and the third signal level threshold is greater than the fourth signal level threshold.

5. A parameter adjusting method of a shared base station, wherein the shared base station is configured to serve a first device of a first communication operator and a second device of a second communication operator, and the parameter adjusting method of the shared base station comprises:

acquiring a first service index of a shared base station;

inputting the first service index into the target network model obtained by training according to any one of claims 1 to 4 to obtain a target configuration parameter, where the target network model is used to determine that the configuration parameter of the shared base station is the target configuration parameter, and when the shared base station configures the target configuration parameter, the signal level of the first device is consistent with the signal level of the second device;

and adjusting the configuration parameters of the shared base station according to the target configuration parameters.

6. The method of claim 5, wherein before obtaining the first service indicator of the shared base station, the method further comprises:

acquiring a second service index of the current base station;

determining a target score of the current base station by adopting an entropy weight algorithm according to the second service index;

and determining the current base station as the shared base station according to the target score.

7. The method of claim 6, wherein the determining the current base station as the shared base station according to the target score comprises:

determining the shared priority level of the current base station according to the target score and the score threshold;

and if the priority level is greater than a level threshold, determining the current base station as the shared base station.

8. A model training apparatus, comprising:

a second obtaining module, configured to obtain tag data corresponding to the sample service indicator, where the tag data is used to indicate an expected configuration parameter corresponding to the sample base station;

the input module is used for inputting the training samples into a target network model to obtain training output configuration parameters;

9. A parameter adjusting apparatus of a shared base station, wherein the shared base station is configured to serve a first device of a first communications operator and a second device of a second communications operator, and the parameter adjusting apparatus of the shared base station comprises:

10. An electronic device, comprising:

at least one processor; and

the memory stores instructions executable by the at least one processor to enable the electronic device to perform the model training method of any one of claims 1 to 4 or the tuning method of a shared base station of any one of claims 5 to 7.

11. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, implements the model training method of any one of claims 1 to 4 or the method for tuning a shared base station of any one of claims 5 to 7.