CN113133069A

CN113133069A - Method and device for determining target cell, electronic equipment and storage medium

Info

Publication number: CN113133069A
Application number: CN202010028236.1A
Authority: CN
Inventors: 李铁钧; 王斐; 刘献玲
Original assignee: Shanghai Datang Mobile Communications Equipment Co ltd
Current assignee: Shanghai Datang Mobile Communications Equipment Co ltd
Priority date: 2020-01-10
Filing date: 2020-01-10
Publication date: 2021-07-16

Abstract

The embodiment of the invention discloses a method and a device for determining a target cell, electronic equipment and a storage medium, wherein the method for determining the target cell comprises the following steps: acquiring measurement data of each cell reported by UE, and determining a service quality value corresponding to each cell based on the measurement data of each cell through a preset quality evaluation model; and determining a target cell corresponding to the UE based on the service quality value corresponding to each cell. The invention can effectively improve the network service quality and improve the user experience.

Description

Method and device for determining target cell, electronic equipment and storage medium

Technical Field

The present invention relates to the field of communications technologies, and in particular, to a method and an apparatus for determining a target cell, an electronic device, and a storage medium.

Background

In a mobile communication system, when a UE (User Equipment) moves, the UE may exceed a service range of a current cell, and in order to ensure communication continuity, the mobile communication system may need to handover the UE from the current cell to another cell (hereinafter, referred to as a target cell).

At present, the aforementioned cell switching of the UE is implemented based on the measurement of the reference signal. Specifically, first, the network side may issue measurement configuration, such as information of measurement frequency points, measurement events, reporting times, and the like, to the UE. Then, after receiving the Measurement configuration, the UE may Report the detected MR (Measurement Report) of each neighboring cell. After Receiving the measurement report reported by the UE, the network side may sort the level RSRP (Reference Signal Receiving Power) values corresponding to each neighboring cell reported by the UE, select the neighboring cell with the highest level value as a target cell according to a level optimization principle, and send a handover command to the UE, so that the UE may be handed over to the target cell based on the handover command.

In the prior art, with the densification of a network, the situations of good reference signal quality and poor service quality often occur, so that selecting a target cell according to a level value (i.e., a reference signal) of an adjacent cell may affect the network service quality to a certain extent, and reduce user experience.

Disclosure of Invention

Because the existing methods have the above problems, embodiments of the present invention provide a method and an apparatus for determining a target cell, an electronic device, and a storage medium.

In a first aspect, an embodiment of the present invention provides a method for determining a target cell, where the method includes:

acquiring measurement data of each cell reported by UE, and determining a service quality value corresponding to each cell based on the measurement data of each cell through a preset quality evaluation model;

and determining a target cell corresponding to the UE based on the service quality value corresponding to each cell.

Optionally, the determining, by using a preset quality evaluation model, the service quality value corresponding to each cell based on the measurement data of each cell includes:

and determining a preset service quality evaluation parameter corresponding to the current cell based on the measurement data of the current cell through the preset quality evaluation model, and determining a service quality value corresponding to the current cell based on the preset service quality evaluation parameter corresponding to the current cell.

Optionally, the preset service quality evaluation parameter includes a throughput rate, a time delay, and a packet loss rate corresponding to the current cell;

the determining the service quality value corresponding to the current cell based on the preset service quality evaluation parameter corresponding to the current cell includes:

determining a first preset evaluation weight value corresponding to the throughput rate, a second preset evaluation weight value corresponding to the time delay and a third preset evaluation weight value corresponding to the packet loss rate through the preset quality evaluation model;

and determining the service quality value corresponding to the current cell based on the first preset evaluation weight value, the second preset evaluation weight value, the third preset evaluation weight value, and the throughput rate, the time delay and the packet loss rate corresponding to the current cell through the preset quality evaluation model.

Optionally, the method further includes:

determining all cell pairs within the coverage of a base station, wherein the cell pairs consist of any two cells within the coverage of the base station;

and acquiring historical cell switching data information corresponding to each cell pair, and generating a preset quality evaluation model based on the historical cell switching data information through a classification prediction algorithm.

Optionally, before the generating a preset quality evaluation model based on the historical cell switching data information by using a classification prediction algorithm, the method further includes:

performing data preprocessing on the historical cell switching data information to obtain a preprocessed data set, wherein the data preprocessing comprises data deduplication processing and abnormal data processing;

performing feature tag processing on the preprocessed data set to obtain a switching data set, wherein the feature tag processing comprises feature data processing and tag data processing;

and constructing a training set based on the switching data set according to a preset training proportion, and constructing a test set based on historical cell switching data information except the training set in the switching data set.

Optionally, the performing data preprocessing on the historical cell handover data information to obtain a preprocessed data set includes:

performing duplicate removal processing on the historical cell switching data information, and deleting the historical cell switching data information which is completely the same as the current historical cell switching data information in the historical cell switching data information to obtain a duplicate removal data set;

and deleting abnormal historical cell switching data information in the duplicate removal data set through a three-sigma rule to obtain a preprocessed data set.

Optionally, the performing feature tag processing on the preprocessed data set to obtain a switching data set includes:

extracting feature data based on the preprocessed data set, wherein the feature data at least comprises a time feature, a source cell feature and a target cell feature;

and acquiring the switching rate of each historical cell switching data information in the preprocessed data set, and performing service quality grading on the historical cell switching data information of the preprocessed data set according to a preset service quality grading rule based on the switching rate of each historical cell switching data information.

Optionally, the generating a preset quality evaluation model based on the historical cell switching data information by using a classification prediction algorithm includes:

performing model training based on the training set to generate a first quality assessment model, and detecting whether the accuracy of the first quality assessment model is greater than or equal to a preset minimum accuracy based on the test set;

and if the first quality evaluation model is greater than or equal to the preset minimum accuracy, determining the first quality evaluation model as the preset quality evaluation model.

Optionally, the determining the target cell corresponding to the UE based on the service quality value corresponding to each cell includes:

and determining the maximum value in all the service quality values, and determining the cell corresponding to the maximum value as a target cell.

and carrying out grade division on all the service quality values according to a preset service quality grade division rule, and determining the cell corresponding to the service quality value with the highest service quality grade as a target cell.

In a second aspect, an embodiment of the present invention further provides a device for determining a target cell, including a quality value determining module and a cell determining module, where:

the quality value determining module is used for acquiring the measurement data of each cell reported by the UE and determining the service quality value corresponding to each cell based on the measurement data of each cell through a preset quality evaluation model;

and the cell determining module is used for determining the target cell corresponding to the UE based on the service quality value corresponding to each cell.

Optionally, the quality value determining module is configured to:

the quality value determination module is to:

Optionally, the system further comprises a model training module, configured to:

Optionally, the model training module is configured to:

and acquiring the switching rate of each historical cell switching data information in the preprocessed data set, and grading the historical cell switching data information of the preprocessed data set according to a preset grading rule based on the switching rate of each historical cell switching data information.

Optionally, the model training module is configured to:

Optionally, the cell determining module is configured to:

In a third aspect, an embodiment of the present invention further provides an electronic device, including:

at least one processor; and

at least one memory communicatively coupled to the processor, wherein:

the memory stores program instructions executable by the processor, which when called by the processor are capable of performing the above-described methods.

In a fourth aspect, an embodiment of the present invention further provides a non-transitory computer-readable storage medium storing a computer program, which causes the computer to execute the above method.

According to the technical scheme, the service quality value of each cell is determined through the preset quality evaluation model, and the target cell corresponding to the UE is determined according to the service quality value of each cell. Therefore, on one hand, the target cell corresponding to the UE is determined based on the service quality value corresponding to each cell, so that the condition that the reference signal is good but the actual service quality is poor when the target cell is determined based on the reference signal can be avoided, the service quality is effectively prevented from being influenced by cell switching, the network service quality can be effectively improved, and the user experience is improved. On the other hand, the service quality value corresponding to each cell is determined based on the preset quality evaluation model, and the accuracy of the determined service quality value can be improved, so that the accuracy of the target cell can be determined, the network service quality can be further improved, and the user experience is improved.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.

Fig. 1 is a flowchart illustrating a method for determining a target cell according to an embodiment of the present invention;

fig. 2 is a flowchart illustrating a method for determining a target cell according to an embodiment of the present invention;

fig. 3 is a schematic structural diagram of an apparatus for determining a target cell according to an embodiment of the present invention;

fig. 4 is a logic block diagram of an electronic device according to an embodiment of the present invention.

Detailed Description

The following further describes embodiments of the present invention with reference to the accompanying drawings. The following examples are only for illustrating the technical solutions of the present invention more clearly, and the protection scope of the present invention is not limited thereby.

An execution main body of the method for determining a target cell provided in the embodiment of the present invention is a network side device, and for a 2G (2-Generation wireless telephone technology, second-Generation mobile communication technology specification) and a 3G (3rd-Generation, third-Generation mobile communication technology) network, the network side device may be a controller; for 4G (the 4th Generation mobile communication technology, fourth Generation mobile communication technology) and 5G (5th-Generation, fifth Generation mobile communication technology) networks, the network side device may be a base station. The network side device may determine the service quality value of each cell reported by the user terminal through a preset quality evaluation model, and may determine the target cell of the user terminal based on the service quality value of each cell.

Fig. 1 shows a flowchart of a method for determining a target cell according to this embodiment, where the method includes:

s101, obtaining measurement data of each cell reported by UE, and determining a service quality value corresponding to each cell based on the measurement data of each cell through a preset quality evaluation model.

The preset quality evaluation model is a model obtained by training based on historical cell switching data information of each cell pair through a classification prediction algorithm in advance, and can be used for predicting the service quality value of each cell according to the measurement data of each cell reported by the UE. The cell pair is composed of a source cell and a target cell in any cell switching.

The service quality value refers to a value used for characterizing the condition of the service quality of the cell determined based on the preset quality evaluation model, and the service quality value can be determined based on one or more of service quality parameters such as throughput rate, time delay and packet loss rate.

In implementation, when the UE moves from one cell to another cell, the network side device may issue measurement configuration to the UE, and after receiving the measurement configuration, the UE may detect each neighboring cell, generate a measurement report corresponding to each cell, and report the measurement report corresponding to each cell to the network side device, and the network side device may determine, based on measurement data in each measurement report reported by the UE, a service quality value corresponding to each cell, and determine, according to the service quality value of each cell, a target cell of the UE, so that the UE may access the target cell. Specifically, first, measurement data of each cell reported by the UE may be obtained, where the measurement data may include a cell identifier, a measurement time, source cell characteristics, and target cell characteristics corresponding to the group of measurement data, where the source cell characteristics may include a source cell level value, a number of source cell RRC (Radio Resource Control) connected user terminals, and a number of idle source cell downlink PRBs (physical Resource blocks); the target cell characteristics may include a target cell level value, a target cell RRC connection user terminal number, and a target cell downlink PRB idle number. Then, the service quality value corresponding to each cell may be determined based on the measurement data of each cell reported by the UE through a preset quality evaluation model.

Before determining the service Quality value corresponding to each cell based on the measurement data of each cell by using the preset Quality evaluation model, historical cell handover data information that does not meet a preset minimum threshold may be deleted according to a preset minimum service threshold, where the threshold may be one or a combination of multiple of RSRP, RSRQ (Reference Signal Receiving Quality, LTE Reference Signal Receiving Quality), and SINR (Signal to Interference plus Noise Ratio). Taking the threshold as RSRP as an example, assuming that the preset minimum service threshold, that is, the preset minimum RSRP threshold is-105 dBm, RSRP values of 3 cells reported by a certain UE are shown in table 1, and since the RSRP value of the cell 3 is less than the preset minimum threshold of-105 dBm, the cell 3 may be deleted, see table 2.

TABLE 1

Cell identity	RSRP value
		Cell 1	-95dBm
Cell 2	-90dBm
		Cell 3	-106dBm

TABLE 2

Cell identity	RSRP value
		Cell 1	-95dBm
Cell 2	-90dBm

S102, determining a target cell corresponding to the UE based on the service quality value corresponding to each cell.

Wherein, the target cell refers to a cell to which the UE will access.

In implementation, after the service quality value corresponding to each cell is determined based on the measurement data of each cell reported by the UE through a preset quality evaluation model, a target cell corresponding to the UE may be determined in all cells reported by the UE based on the service quality value corresponding to each cell. It can be understood that the preset quality evaluation model may determine the service quality value of one cell at a time, so that the number of cells reported by the UE is equal to the number of times that the preset quality evaluation model determines the service quality value, that is, if the number of the cells is 4, the preset quality evaluation model needs to be executed 4 times to determine the service quality value corresponding to each cell.

Further, on the basis of the method embodiment, the service quality value corresponding to each cell may be determined based on the preset service quality evaluation parameter corresponding to each cell reported by the UE, and the corresponding partial processing in step S101 may be as follows: and determining a preset service quality evaluation parameter corresponding to the current cell based on the measurement data of the current cell through a preset quality evaluation model, and determining a service quality value corresponding to the current cell based on the preset service quality evaluation parameter corresponding to the current cell.

The current cell refers to any one of the cells reported by the UE.

The preset service quality evaluation parameter refers to a parameter for evaluating the service quality of the current cell, and may be one or more of service quality parameters such as throughput rate, delay, packet loss rate, and the like, and the preset service quality evaluation parameter may be determined based on measurement data of the current cell.

In implementation, after the measurement data of each cell reported by the UE is obtained, a preset service quality assessment parameter corresponding to each cell may be determined, so as to determine a service quality value corresponding to each cell. Specifically, the preset quality evaluation model may determine the preset service quality evaluation parameter of the current cell based on the measurement data of the current cell, for example, the preset service quality evaluation parameter may be a specific numerical value of one or more of throughput rate, delay, and packet loss rate. Then, the service quality value corresponding to the current cell can be determined based on the preset quality evaluation parameter corresponding to the current cell through the preset quality evaluation model. It can be understood that the service quality value corresponding to each cell reported by the UE can be determined in turn according to the above method. Therefore, the service quality value of each cell is determined based on the preset service quality evaluation parameters, so that the determined service quality value of each cell can better represent the real service quality of each cell, the accuracy of the determined target cell can be further improved, and the network service quality and the user experience are further improved.

Further, on the basis of the above method embodiment, the preset service quality evaluation parameter may be a throughput rate, a time delay, and a packet loss rate corresponding to the current cell, and the corresponding process of determining the service quality value of the current cell may be as follows: determining a first preset evaluation weight value corresponding to the throughput rate, a second preset evaluation weight value corresponding to the time delay and a third preset evaluation weight value corresponding to the packet loss rate through a preset quality evaluation model; and determining the service quality value corresponding to the current cell based on the first preset evaluation weight value, the second preset evaluation weight value, the third preset evaluation weight value, the throughput rate, the time delay and the packet loss rate corresponding to the current cell through a preset quality evaluation model.

The first preset evaluation weight value refers to a weight value corresponding to a preset throughput rate; the second preset evaluation weight value refers to a weight value corresponding to a preset time delay; the third preset evaluation weight value refers to a preset weight value corresponding to the packet loss rate.

In implementation, when the preset service quality evaluation parameter only includes throughput rate, time delay and packet loss rate corresponding to the current cell, the service quality value corresponding to the current cell may be determined based on weight values corresponding to different preset service quality evaluation parameters through the preset quality evaluation model. Specifically, first, the throughput rate, the time delay, and the packet loss rate corresponding to the current cell may be determined by a preset quality evaluation model. Then, a first preset evaluation weight value corresponding to the throughput rate, a second preset evaluation weight value corresponding to the time delay, and a third preset evaluation weight value corresponding to the packet loss rate can be determined through a preset quality evaluation model. And then, determining the service quality value of the current cell by a preset quality evaluation model based on the first preset evaluation weight value, the second preset evaluation weight value, the third preset evaluation weight value, and the throughput rate, the time delay and the packet loss rate corresponding to the current cell.

Taking the first preset evaluation weight value, the second preset evaluation weight value and the third preset evaluation weight value as 50%, 30% and 20% respectively as an example, assuming that the throughput rate corresponding to the current cell is a, the time delay is b, and the packet loss rate is c, the service quality value of the current cell can be determined to be a 50% + b 30% + c 20%. That is, the throughput rate, the time delay, and the packet loss rate corresponding to the current cell may be subjected to comprehensive weighted scoring according to the different weight values, and a comprehensive scoring result (i.e., a service quality value) is finally obtained, where a higher score of the service quality value indicates a better service quality, and the ranking is further forward. It can be understood that the preset service quality evaluation parameter may also be one or two of service quality indicators such as throughput rate, delay, packet loss rate, and the like, that is, the service quality value may also be determined based on one or more of the service quality indicators such as throughput rate, delay, packet loss rate, and the like. Therefore, the service quality value of the current cell is determined by combining the weighted values corresponding to different preset service quality evaluation parameters, so that the determined service quality value is more accurate, the accuracy of the determined target cell can be further improved, and the network service quality is further improved.

Further, on the basis of the above method embodiment, a preset quality evaluation model may also be generated based on historical cell handover data information of all cell pairs of the base station, and corresponding processing may be as follows: determining all cell pairs within the coverage of a base station; and acquiring historical cell switching data information corresponding to each cell pair, and generating a preset quality evaluation model based on the historical cell switching data information through a classification prediction algorithm.

The cell pair is composed of any two cells within the coverage area of the base station, and for any cell handover, the source cell and the target cell of the cell handover are a cell pair.

The classification prediction algorithm may be one of KNN (K-Nearest Neighbor) classification algorithm, SVM (Support Vector Machine), and the like.

In implementation, before determining the UE target cell, a preset quality assessment model may be trained and generated based on historical cell handover data information of all cell pairs within the coverage of the base station. First, all cell pairs within the coverage of the base station may be determined, and historical cell handover data information corresponding to a preset number (for example, 5000) of each cell pair may be obtained. Then, a preset quality evaluation model can be generated based on the historical cell switching data information by a classification prediction algorithm, that is, the historical cell switching data information corresponding to all the cells is trained by all the cells. It can be understood that the preset number of pieces of historical cell switching data information may be the preset number of pieces of historical cell switching data information with the shortest interval duration from the current time. Therefore, the preset quality evaluation model is trained and generated based on the historical cell switching data information of all cell pairs in the coverage area of the base station, so that the accuracy of the evaluation result of the preset quality evaluation model, namely the accuracy of the determined service quality value of each cell, is improved, the network service quality is further improved, and the user experience is improved; meanwhile, the complaints of users can be further reduced.

All the cell pairs corresponding to the base station are as follows: for example, the preset number of (source cell) cell S1- (target cell) cell 01, (source cell) cell S1- (target cell) cell 02 is 5000, and the obtained historical cell handover data information of all cell pairs within the coverage area of the base station may be shown in table 3.

TABLE 3

Wherein, the user identification: refers to a user ID that distinguishes different users.

Switching time: refers to the time when the UE performs cell handover, which may be recorded only in hourly minutes or more accurately (e.g. in seconds).

Source cell level value: the level value of the serving cell before the UE performs cell handover may be an average value of the reported level values of the last 1 or more MR source cells, and the unit is dBm (decibel relative to one milliwatt).

The number of source cell RRC connected user terminals: the number of the user terminals connected to the RRC in the serving cell before the UE performs cell handover may be counted as one user terminal connected to the RRC in 1 minute before the current time.

The idle number of the downlink PRBs of the source cell is as follows: the idle number of the downlink PRB of the serving cell before the UE performs cell switching can be counted within 1 minute or longer before the current time, and the unit is one.

Target cell level value: before cell handover, the level value of the target cell measured by the UE may be an average result of the level values of the last 1 or more reported MR target cells, and the unit is: dBm.

The number of target cell RRC connected user terminals is as follows: before the UE performs cell handover, the number of RRC connected user terminals in the target cell may be counted in units of one RRC connected user terminal in 1 minute before the current time.

The idle number of the target cell downlink PRB is as follows: before the UE performs cell switching, the idle number of the downlink PRB of the target cell can be counted in units of one idle number of the downlink PRB within 1 minute or more before the current time.

Switching rate: after the UE performs cell handover, the average rate of the user at a downlink PDCP (Packet Data Convergence Protocol) layer of the target cell, that is, the total Data amount of all transmission Data divided by all transmission time for transmitting the total Data amount, unit: and Mbps. I.e. the downlink rate (average) at which the user enters the target cell.

Further, on the basis of the above method embodiment, before training and generating a preset quality evaluation model based on the historical cell handover data information within the coverage area of the base station, data preprocessing may be performed on the data, and the corresponding processing may be as follows: performing data preprocessing on historical cell switching data information to obtain a preprocessed data set; performing characteristic label processing on the preprocessed data set to obtain a switching data set; and constructing a training set based on the switching data set according to a preset training proportion, and constructing a test set based on historical cell switching data information except the training set in the switching data set.

Wherein the data preprocessing at least comprises data deduplication processing and exception data processing.

The feature tag processing may include at least feature data processing and tag data processing.

The preprocessed data set refers to a data set obtained by preprocessing the historical cell switching data information, and the data set does not have completely repeated historical cell switching data information and abnormal data.

The switching data set refers to a data set obtained by performing feature tag processing on the preprocessed data set.

The preset proportion refers to the ratio of a training set to a test set in a preset switching data set, and can be 7:3, namely 70% of data in the switching data set is used as the training set, and the rest 30% of data is used as the test set.

In implementation, before the preset quality evaluation model is generated based on the historical cell switching data information through a classification prediction algorithm, data preprocessing and feature label processing can be further performed on the historical cell switching data information. Specifically, first, data preprocessing may be performed on all the historical cell handover data information, for example, deduplication processing and abnormal data processing may be performed, so as to obtain a preprocessed data set. Then, feature tag processing may be performed on the preprocessed data set, for example, feature data processing and tag data processing may be performed, so as to obtain a switching data set. And then, splitting the switching data set into training sets according to a preset proportion, taking historical cell switching data information except the training sets in the switching data set as test sets, wherein the training sets are used for model training, and the test sets can detect the accuracy of the trained models so that the accuracy of the trained preset quality assessment models meets preset requirements. Therefore, data preprocessing and feature tag processing are carried out on the historical cell switching data information, repeated data and abnormal data in the historical cell switching data information can be removed, feature tags are added to the historical cell switching data information, time consumed by model training can be reduced to a certain extent, accuracy of a preset quality evaluation model of training can be further improved, accuracy of a determined target cell is further improved, and user experience is improved.

Further, on the basis of the above method embodiment, the repeated data and abnormal data in the historical cell handover data information may be deleted, and the corresponding processing may be as follows: performing duplicate removal processing on all historical cell switching data information, and deleting historical cell switching data information which is completely the same as the current historical cell switching data information in the historical cell switching data information to obtain a duplicate removal data set; and deleting abnormal historical cell switching data information in the duplicate removal data set through a three-sigma rule to obtain a preprocessed data set.

The current historical cell switching data information refers to any one of the historical cell switching data information.

In implementation, when data preprocessing is performed, the repeated data and the abnormal data in the historical cell switching data information may be deleted. Specifically, first, it may be detected whether there is history cell switching data information (for example, one or more history cell switching data information) completely overlapping with the current history cell switching data information in the history cell switching data information. And if so, deleting the historical cell switching data information which is completely repeated with the current historical cell switching data information to obtain a duplication-removed data set. Then, abnormal data processing may be performed on the deduplication data set, for example, abnormal data processing may be performed on the deduplication data set through a three-sigma criterion, so as to delete abnormal historical cell switching data information in the deduplication data set, and obtain a preprocessed data set. Therefore, abnormal data in the duplicate removal data set can be effectively removed through abnormal data processing of the three sigma criteria process, so that historical cell switching data information in the preprocessed data set is all effective data (namely abnormal historical cell switching data information and repeated historical cell switching data information are avoided), time consumed for training the preset quality assessment model can be further reduced, accuracy of the trained preset quality assessment model can be further improved, accuracy of the determined target cell is improved, network service quality can be further improved, user experience is improved, and user complaints can be reduced.

The above-mentioned process of performing abnormal data processing on the duplicate data set by using the three-sigma criterion may be as follows: when all the above-mentioned historical cell switch data information obeys normal distribution, the value of most (e.g. 99%) of the historical cell switch data information should be within a distance of 3 standard deviations from the mean value, i.e. P (| x- μ | >3 σ) ≦ 0.003, where x represents a specific value of a certain parameter in the historical cell switch data information. When the value exceeds the distance, the value can be regarded as an abnormal value, and the historical cell switching data information corresponding to the abnormal value is deleted. The parameters of the historical cell handover data information may be: the level value of the source cell, the number of RRC connected user terminals of the source cell, the idle number of downlink PRBs of the source cell, the level value of the target cell, the number of RRC connected user terminals of the target cell, the idle number of downlink PRBs of the target cell and the downlink rate (average) of the user entering the target cell.

Further, on the basis of the above method embodiment, when performing the feature tag processing, feature extraction and ranking may be performed based on the preprocessed data set, and the corresponding processing may be as follows: extracting feature data based on the preprocessed data set; and acquiring the switching rate of each historical cell switching data information in the preprocessed data set, and performing service quality grade division on the historical cell switching data information of the preprocessed data set according to a preset service quality grade division rule based on the switching rate of each historical cell switching data information.

The characteristic data may include at least a time characteristic, a source cell characteristic, and a target cell characteristic.

The preset service quality grade division rule refers to a preset rule for performing service quality grade division on the cell according to the switching rate, and the higher the service quality grade is, the faster the switching rate is, for example, see table 4.

In implementation, after the above-described preprocessed data sets are obtained, feature extraction and ranking may be performed. Specifically, first, feature data may be extracted based on the preprocessed data set, such as time feature, source cell feature, and target cell feature extraction may be performed. Then, the switching rate of each historical cell switching data information in the preprocessing processing set can be obtained, the grade corresponding to each switching rate can be determined according to a preset service quality grade division rule, and the service quality grade of each historical cell switching data information can be determined according to the grade of each switching rate.

Specifically, the time feature extraction may be classified according to busy hours and idle hours, for example, a time period of the busy hours may be set to 7:00-20:00, and other time periods may be set to idle hours. Then, it can be determined according to the extraction time that each piece of historical cell handover data information in the preprocessed data set is idle data or busy data. The source cell characteristics at least include a source cell level value, the number of RRC connected user terminals in the source cell, and the idle number of downlink PRBs in the source cell. The target cell characteristics at least include a target cell level value, the number of target cell RRC connected user terminals, and the number of target cell downlink PRB idle.

The above-mentioned service quality ranking process may be as follows: taking the switching rate of each historical cell switching data information in the preprocessing processing set as a label, classifying (i.e. ranking) each historical cell switching data information in the preprocessing processing set, for example, ranking according to table 4, and it can be understood that in actual implementation, different ranking modes can be set according to different service quality parameters according to actual conditions.

TABLE 4

Tag data classification	Rate Range (units Mpbs)
		Throughput rate class 1	[0～0.5]
Throughput rate class 2	(0.5～1]
		Throughput class 3	(1～3]
Throughput class 4	(3～5]
		Throughput rate class 5	(5～10]
Throughput rate class 6	(10～15]
		Throughput class 7	(15～30]
Throughput class 8	(30～45]
		Throughput class 9	(45～70]
Throughput class 10	>70

Further, on the basis of the above method embodiment, model training may be performed based on a training set, and model accuracy may be detected based on a test set, and corresponding processing may be as follows: performing model training based on the training set to generate a first quality assessment model, and detecting whether the accuracy of the first quality assessment model is greater than or equal to a preset minimum accuracy based on the test set; and if the first quality evaluation model is greater than or equal to the preset minimum accuracy, determining the first quality evaluation model as a preset quality evaluation model.

Wherein the first quality evaluation model refers to a quality evaluation model generated based on test set training.

The preset minimum accuracy refers to the minimum accuracy to be met by a preset quality evaluation model.

In implementation, when the preset quality assessment model is generated by a classification prediction algorithm based on the historical cell switching data information training, firstly, model training may be performed based on a training set to generate a first quality assessment model. The accuracy of the first quality assessment model may then be tested based on the test set to determine whether the accuracy of the first quality assessment model is greater than or equal to a preset minimum accuracy. If the accuracy of the first quality assessment model is greater than or equal to the predetermined minimum accuracy, the first quality assessment model may be determined as the predetermined quality assessment model. Therefore, the accuracy of the preset quality evaluation model can be effectively improved, the accuracy of the determined target cell can be further improved, the network service quality can be further improved, and the user experience can be improved.

Taking a classification prediction algorithm as KNN and taking the preset lowest accuracy as 95% as an example, the KNN can be adopted for model training, namely under the condition that feature data and labels in a training set are known, test set data is input, the features of the test set data and the corresponding features in the training set data are compared with each other, the first K data which are most similar to the test set data in the training set data are found, the grade category corresponding to the test set data is the grade with the largest occurrence frequency in the K data, and the algorithm can be described as follows:

firstly, the Euclidean distance between the test set data and each training set data is calculated, and the data are sorted according to the increasing relation of the distance. And then, selecting K points with the minimum distance, and determining the occurrence frequency of the class of the first K points. And then, returning the grade class with the highest occurrence frequency in the previous K points as the predicted grade classification of the test set data.

Then, after the first quality evaluation model is obtained, a preset minimum accuracy rate can be adopted as an evaluation function of the model to select a proper value of K to tune parameters of the first quality evaluation model. The formula for determining the actual accuracy of the first quality assessment model may be: accuracy is the number of model predictions in the test set data/total number of test set data. When the actual accuracy of the first quality assessment model reaches the preset minimum accuracy, that is, more than 95%, the currently adopted K value may be considered to be relatively accurate, and the first quality assessment model at this time may be determined as the preset quality assessment model.

Further, on the basis of the above method embodiment, the cell corresponding to the maximum value of the service quality value may be determined as the target cell, and the corresponding processing may be as follows: and determining the maximum value of all the service quality values, and determining the cell corresponding to the maximum value as the target cell.

In implementation, after the service quality value corresponding to each cell reported by the UE is determined, the service quality values corresponding to each cell may be sorted according to the size order, for example, may be sorted in a descending order or in an ascending order. Then, the maximum value of the quality of service values in the sequence may be determined, and if the values are sorted in descending order, the first value in the sequence may be selected as the maximum value; if ascending, the last in the sequence may be selected as the maximum. Then, the cell corresponding to the maximum value may be determined as a target cell, so that the UE may switch to the target cell. Therefore, the cell corresponding to the maximum value of the service quality value is selected as the target cell, the condition that the network service quality is influenced due to cell switching can be further avoided, the user experience can be further improved, and complaints can be reduced.

It is to be understood that the above quality of service value may be only throughput, or may be determined according to one or more quality of service parameters, for example, according to quality of service parameters such as throughput, delay, packet loss rate, and the like. Specifically, first, the throughput rate, the time delay, and the packet loss rate corresponding to each cell may be determined by a preset quality evaluation model. Then, the throughput rate, the time delay, and the packet loss rate corresponding to each cell may be subjected to comprehensive weighted scoring, for example, the throughput rate may account for 50%, the packet loss rate may account for 30%, and the time delay may account for 20%, and finally a comprehensive scoring result (i.e., a service quality value) is obtained, where a higher score of the service quality value indicates better service quality, and the ranking is further forward.

Further, on the basis of the above method embodiment, the cell corresponding to the highest-level quality of service value may be determined as the target cell, and the corresponding processing may be as follows: and performing grade division on all the service quality values according to a preset grade division rule, and determining the cell corresponding to the service quality value with the highest grade as a target cell.

The preset grade division rule refers to a preset rule for dividing the service quality values in different ranges into different grades, and if the grade is higher, the service quality value is also higher.

In implementation, after the service quality value corresponding to each cell is determined, the service quality value corresponding to each cell may be classified according to a preset classification rule. Specifically, first, each quality of service value may be ranked according to a preset ranking rule, and the quality of service values may be sorted in ascending/descending order of rank. Then, the service quality value with the highest grade can be selected from the sequence, and if the service quality values are sorted according to the ascending order of the grade, the last service quality value in the sequence is the service quality value with the highest grade; if the quality of service values are sorted according to a descending order of rank, the first quality of service value in the sequence is the highest quality of service value. Then, the cell corresponding to the highest quality of service value of the rank may be determined as the target cell. Therefore, the cell corresponding to the service quality value with the highest grade is determined as the target cell, and the network service quality and the user experience can be further improved.

It should be noted that, when the maximum value of the service quality values is multiple or the service quality value with the largest level is multiple, the reporting time of the MR of the cell corresponding to each service quality value may be obtained, and the cell with the earliest reporting time is determined as the target cell.

Fig. 2 shows a complete implementation of an embodiment of the invention. First, historical cell switching data information of a preset number of cell pairs corresponding to each cell in the coverage area of the base station can be acquired. Then, data preprocessing and feature tag processing may be performed on all the historical cell handover data information to obtain a handover data set. Then, data splitting can be performed on the switching data set according to a preset training proportion, and a training set and a test set are obtained. And then, training based on the training set through the KNN to obtain a first quality evaluation model, evaluating the first quality evaluation model based on the test set, adjusting parameters of the model until the accuracy of the first quality evaluation model is greater than or equal to the preset minimum accuracy, and finishing training to obtain the preset quality evaluation model.

After the preset quality evaluation model is trained, the target cell can be determined through the preset quality evaluation model. First, the cell reported by the UE may be determined, and the measurement data of each cell may be obtained. Then, the service quality value corresponding to each cell may be determined based on the measurement data of each cell through a preset quality evaluation model, that is, the measurement data of each cell is sequentially input to the preset quality evaluation model, so as to obtain the service quality value corresponding to each cell. Then, a target cell corresponding to the UE may be determined based on the quality of service value corresponding to each cell.

Fig. 3 shows a target cell determination apparatus provided in this embodiment, which includes a quality value determination module 301 and a cell determination module 302, where:

the quality value determining module 301 is configured to obtain measurement data of each cell reported by the UE, and determine a service quality value corresponding to each cell based on the measurement data of each cell through a preset quality evaluation model;

the cell determining module 302 is configured to determine a target cell corresponding to the UE based on the service quality value corresponding to each cell.

Optionally, the quality value determining module 301 is configured to:

the quality value determination module 301 is configured to:

Optionally, the model training module is configured to:

Optionally, the cell determining module 302 is configured to:

and performing grade division on all the service quality values according to a preset grade division rule, and determining the cell corresponding to the service quality value with the highest grade as a target cell.

The determining apparatus of the target cell in this embodiment may be used to perform the above method embodiments, and the principle and technical effect are similar, which are not described herein again.

Referring to fig. 4, the electronic device includes: a processor (processor)401, a memory (memory)402, and a bus 403;

wherein the content of the first and second substances,

the processor 401 and the memory 402 complete communication with each other through the bus 403;

the processor 401 is configured to call program instructions in the memory 402 to perform the methods provided by the above-described method embodiments.

The present embodiments disclose a computer program product comprising a computer program stored on a non-transitory computer readable storage medium, the computer program comprising program instructions which, when executed by a computer, enable the computer to perform the methods provided by the above-described method embodiments.

The present embodiments provide a non-transitory readable storage medium storing computer instructions that cause the computer to perform the methods provided by the method embodiments described above.

The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.

Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in the embodiments or some parts of the embodiments.

It should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.

Claims

1. A method for determining a target cell, comprising:

2. The method of claim 1, wherein the determining the quality of service value corresponding to each cell based on the measurement data of each cell through a preset quality assessment model comprises:

3. The method according to claim 2, wherein the preset qos evaluation parameter includes a throughput rate, a delay, and a packet loss rate corresponding to the current cell;

4. The method for determining a target cell according to claim 1, further comprising:

5. The method of claim 4, wherein before the generating the preset quality assessment model based on the historical cell handover data information by the classification prediction algorithm, the method further comprises:

6. The method for determining a target cell according to claim 5, wherein the performing data preprocessing on the historical cell handover data information to obtain a preprocessed data set includes:

7. The method for determining a target cell according to claim 5, wherein the performing feature tag processing on the preprocessed data set to obtain a handover data set includes:

8. The method for determining a target cell according to claim 5, wherein the generating a preset quality evaluation model based on the historical cell handover data information through a classification prediction algorithm comprises:

9. The method of claim 1, wherein the determining the target cell corresponding to the UE based on the quality of service value corresponding to each cell comprises:

10. The method of claim 7, wherein the determining the target cell corresponding to the UE based on the quality of service value corresponding to each cell comprises:

11. A target cell determination apparatus, comprising a quality value determination module and a cell determination module, wherein:

12. The apparatus of claim 11, wherein the quality value determination module is configured to:

13. The apparatus for determining a target cell according to claim 12, wherein the preset qos evaluation parameter includes a throughput rate, a delay and a packet loss rate corresponding to the current cell;

the quality value determination module is to:

14. The apparatus for determining a target cell of claim 11, further comprising a model training module configured to:

15. The apparatus for determining the target cell of claim 14, wherein the model training module is configured to:

16. The apparatus for determining a target cell of claim 15, wherein the model training module is configured to:

17. The apparatus for determining a target cell of claim 15, wherein the model training module is configured to:

18. The apparatus for determining a target cell of claim 15, wherein the model training module is configured to:

19. The apparatus for determining the target cell of claim 11, wherein the cell determining module is configured to:

20. The apparatus for determining the target cell of claim 17, wherein the cell determining module is configured to:

21. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor when executing the program implements the method for determining a target cell according to any of claims 1 to 10.

22. A non-transitory readable storage medium having stored thereon a computer program, wherein the computer program when executed by a processor implements the method for determining a target cell according to any one of claims 1 to 10.