Communication network performance evaluation method and system

Abstract

The present invention discloses a communication network performance evaluation method and system, the method comprising: respectively collecting performance data and cell information data of each cell; classifying the cells according to the cell information data to obtain different types of cells; sorting the cells of the same type to obtain a sequence of cells to be clustered; for each cell in the sequence of cells to be clustered, generating a performance evaluation array of the cell according to the cell performance data; clustering the cells in the sequence of cells to be clustered using the performance evaluation array to obtain a cell performance clustering result. By using the scheme of the present invention, the performance evaluation of the communication network can be realized efficiently, conveniently and accurately.

Description

Technical Field

[0001] The present invention relates to the technical field of network performance evaluation, and in particular to a communication network performance evaluation method and system. Background Art

[0002] At present, the 5G network is still in the rapid construction stage. The number of 5G users is relatively small compared to the number of 4G users. The traffic absorption and service carrying capacity of 5G network cells cannot be fully reflected, and the performance cannot be evaluated. In this case, it is particularly difficult to accurately evaluate the performance of 5G cells of different manufacturers, station types, transceiver modes, and bandwidths under different user conditions, find the best performance point, make accurate estimates for future user and business explosions, and find targeted optimization strategies.

[0003] Traditionally, there are two main ways to evaluate the performance of communication networks: 1. Manual processing, by extracting performance indicators, grouping and summarizing them using Excel or database tools, finding the average, and then finding the best performance point by drawing a graph; 2. Simulation through large business tests.

[0004] The above-mentioned traditional method 1 has high classification complexity and poor accuracy; the traditional method 2 has large resource investment and cost consumption, and this method cannot reflect the performance trends of cells in different scenarios and configurations, especially in the process of continuous network changes. It not only cannot accurately reflect the performance trends of cells, but also cannot effectively track the performance trends of cells. Summary of the invention

[0005] The present invention provides a communication network performance evaluation method and system, which can efficiently, conveniently and accurately evaluate the communication network performance.

[0006] To this end, the present invention provides the following technical solutions:

[0007] In one aspect, the present invention provides a communication network performance evaluation method, the method comprising:

[0008] Collecting the efficiency data and information data of each cell respectively;

[0009] Classifying cells according to the cell information data to obtain different types of cells;

[0010] Sort the cells of the same type to obtain a sequence of cells to be clustered;

[0011] For each cell in the sequence of cells to be clustered, generating a performance evaluation array of the cell according to the cell performance data;

[0012] The cells in the sequence of cells to be clustered are clustered using the performance evaluation array to obtain a cell performance clustering result.

[0013] Optionally, the cell performance data includes: time, area name, cell identifier CGI, number of activated users, PRB utilization, and traffic; the cell information data includes: area name, manufacturer name, cell identifier CGI, communication standard, cell frequency band bandwidth, number of cell channels, and cell type.

[0014] Optionally, the classifying the cell information data includes: classifying the cell information data according to one or more combinations of data types included in the cell information data.

[0015] Optionally, the sorting of cells of the same type to obtain a sequence of cells to be clustered includes:

[0016] For each cell, the interval is divided according to the number of activated users, and multiple intervals of the number of activated users in the cell are obtained;

[0017] Determining the cell data enrichment degree according to the cell activated user number interval of the cell;

[0018] The cells of the same type are sorted according to the cell data richness and the number of cell data samples to obtain a sequence of cells to be clustered.

[0019] Optionally, determining the cell data fullness according to the cell activated user number intervals of the cell includes: calculating the cell data fullness, the cell data fullness=the number of cell activated user number intervals for which performance data exists / the total number of cell activated user number intervals.

[0020] Optionally, the sorting of cells of the same type according to the cell data fullness and the number of cell data samples to obtain a sequence of cells to be clustered includes: sorting the cells of the same type from high to low according to the cell data fullness, and then sorting the cells with the same data fullness from high to low according to the number of cell data samples.

[0021] Optionally, the method further includes: eliminating cells whose cell data enrichment is lower than a set value.

[0022] Optionally, generating the cell performance evaluation array according to the cell performance data includes:

[0023] Calculating the performance mean of the interval of the number of activated users in each cell in the cell;

[0024] According to the intervals of the number of activated users in each area of ​​the cell and the mean values ​​of their performance, a performance evaluation array of the cell is generated.

[0025] Optionally, clustering the cells in the sequence of cells to be clustered by using the performance evaluation array to obtain a cell performance clustering result includes:

[0026] Calculating the performance fitting degree between different cells in the sequence of cells to be clustered in pairs under the same interval of the number of activated users;

[0027] Two cells whose bidirectional effectiveness fitting degrees both reach the set threshold are included in the same type of trend characteristic cell set;

[0028] If the bidirectional performance fitting degree of a cell with all cells in a certain type of trend characteristic cell set reaches the set threshold, the cell is included in the trend characteristic cell set.

[0029] Optionally, the cell performance clustering result includes one or more types of trend feature cell sets; and the method further includes:

[0030] According to the performance data of all cells in the trend characteristic cell cluster, a common performance curve of the corresponding type of cells is generated.

[0031] On the other hand, the present invention also provides a communication network performance evaluation system, the system comprising:

[0032] A data collection module, used to collect the efficiency data and information data of each cell respectively;

[0033] A classification module, used to classify cells according to the cell information data to obtain different types of cells;

[0034] A sorting module is used to sort cells of the same type to obtain a sequence of cells to be clustered;

[0035] An efficiency evaluation array generation module, used for generating an efficiency evaluation array of each cell in the sequence of cells to be clustered according to the cell efficiency data;

[0036] The clustering module is used to cluster the cells in the sequence of cells to be clustered using the performance evaluation array to obtain a cell performance clustering result.

[0037] Optionally, the cell performance clustering result includes one or more types of trend feature cell sets; and the system further includes:

[0038] The performance curve generation module is used to generate a common performance curve of the corresponding type of cells based on the performance data of all cells in the various trend characteristic cell sets.

[0039] The communication network performance evaluation method and system provided by the present invention respectively collect the performance data and cell information data of each cell, and for each cell of the same type, generate a performance evaluation array of the cell according to the cell performance data, and use the performance evaluation array to cluster the cells in the sequence of cells to be clustered to obtain the cell performance clustering results, so as to efficiently, conveniently and accurately evaluate the performance of the communication network, and then classify the high and low performance of cells in the entire network, formulate high and low performance trend standards, explore the differences and problems of high and low performance and abnormalities, and better support actual network production. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 It is a flow chart of the communication network effectiveness evaluation method of the present invention;

[0041] Figure 2 It is an example of common performance curves of three types of cells obtained by using the communication network performance evaluation method of the present invention;

[0042] Figure 3 It is a structural schematic diagram of the communication network effectiveness evaluation system of the present invention;

[0043] Figure 4 It is another structural schematic diagram of the communication network effectiveness evaluation system of the present invention. DETAILED DESCRIPTION

[0044] In order to enable persons skilled in the art to better understand the solutions of the embodiments of the present invention, the embodiments of the present invention are further described in detail below in conjunction with the accompanying drawings and implementation modes.

[0045] In view of the current problems of difficulty in evaluating the performance of 5G networks and discovering performance problems, the embodiments of the present invention provide a communication network performance evaluation method and system, which evaluates the performance of cells according to different manufacturers, coverage types, frequency bands, standards, bandwidths, scenarios, etc., classifies the performance of cell communication networks through clustering fitting, and uses performance curves to mine low-performance and abnormal performance curves, and conducts more accurate performance positioning, analysis and judgment, so as to better guide actual network production.

[0046] The communication network performance evaluation method and system provided by the present invention divide and classify cell performance data and cell information into intervals according to the number of activated users and cell frequency bands, standards, manufacturers and other information. For each cell, the corresponding performance mean in the interval is calculated according to the interval of the number of activated users. By intelligently learning the cell performance data, polynomial regression is used to calculate the fit of the performance curve between two cells, and the cell curves of different cells are classified according to the performance fitting situation. By analyzing the performance curves of different categories, a comparative analysis of the performance trends of cells of different manufacturers, frequency bands, and standards can be completed.

[0047] like Figure 1 FIG. 1 is a flow chart of a method for evaluating the effectiveness of a communication network according to the present invention, comprising the following steps:

[0048] Step 101, collect the performance data and cell information data of each cell respectively.

[0049] The cell performance data includes but is not limited to any one or more of the following data: time, area name, cell identifier CGI, number of activated users, PRB utilization, and traffic;

[0050] The cell information data includes but is not limited to any one or more of the following data: area name, manufacturer name, cell identifier CGI, communication standard, cell frequency bandwidth, cell channel number, and cell type.

[0051] In a specific application, the collected data may be summarized in a table form, such as being filled into a corresponding cell performance table and a cell information table. For example, the cell performance table and the cell information table may be in the form of the following Table 1 and Table 2, respectively.

[0052] Table 1

[0053]

[0054]

[0055] Table 2

[0056]

[0057] It should be noted that the above parameters can be obtained from the relevant system platform and OMC (Operation and Maintenance Center). In addition, in order to enable more accurate cluster analysis, the corresponding parameter data in the cell performance table can be extracted according to a certain granularity for a certain time period, such as extracting cell-level data with a granularity of 15 minutes for one month.

[0058] Step 102: Classify cells according to the cell information data to obtain different types of cells.

[0059] Considering that factors such as network equipment, standard and configuration have different effects on communication network performance, only cells with the same hardware, physical configuration, and scenario are comparable. To this end, in the embodiment of the present invention, the cells to be compared may be classified first, and then cluster analysis may be performed for cells of each type.

[0060] Specifically, in response to requirements of different applications, the cell information data may be classified according to one or more combinations of the data types included in the cell information data, which is not limited in the embodiment of the present invention.

[0061] Step 103: sort the cells of the same type to obtain a sequence of cells to be clustered.

[0062] In the embodiment of the present invention, for all cells of the same type, these cells may be sorted according to the cell data richness and the number of cell data samples.

[0063] For each cell, the interval can be divided according to the number of activated users to obtain multiple intervals of the number of activated users in the cell. The number of activated users refers to the number of users who are doing business in the cell at the same time. Specifically, for each cell, the average number of activated users that appear in a certain time step (such as 5 seconds) can be counted each time, and the average value can be calculated after counting for a period of time (such as 15 minutes), and the average value is used as the number of activated users in the cell. For example, the maximum number of activated users supported by the cell network device is 200. When the interval granularity RangeSize = 1, it is equivalent to dividing the number of activated users in the cell into 200 intervals, that is, [0,1), [1,2).... [199,200); if RangeSize = 0.5, it means that it is divided into 400 intervals of activated users in the cell, that is, [0,0.5), [0.1,1).... [399.5,400). The smaller the interval, the finer the granularity, and the more accurate the clustering result, but if there is less original data, the finer the interval division granularity, which may lead to more interval null values. The selection of the interval granularity can be set according to application requirements, and the embodiment of the present invention does not limit this.

[0064] The calculation of the cell data richness can be based on the above-mentioned division of the cell interval. For the convenience of description, the cell data richness is represented by celldataFRate, which can be calculated according to the following formula:

[0065] celldataFRate=number of cell active user number intervals for which performance data exists / total number of cell active user number intervals.

[0066] The cell activated user number interval with performance data refers to an interval in which data in the cell activated user number interval is not empty.

[0067] The cell data sampling quantity refers to the total number of data items recorded in the cell performance table corresponding to the cell.

[0068] When sorting all cells of the same type, the cells are mainly sorted from high to low according to the cell data richness, and the cells with the same data richness are sorted from high to low according to the number of cell data samples.

[0069] It should be noted that in specific applications, data in intervals with a small amount of data or a small number of sampling points can be eliminated to avoid affecting the accuracy of clustering results due to the sparse sampling points.

[0070] Step 104: for each cell in the sequence of cells to be clustered, generate a performance evaluation array of the cell according to the cell performance data.

[0071] In the embodiment of the present invention, the efficiency mean value corresponding to each interval divided into the above-mentioned intervals of the number of activated users in each cell can be calculated respectively.

[0072] The performance mean corresponding to the interval = avg1 × avg2;

[0073] Among them, avg1 represents the average traffic value corresponding to all activated users in the interval, and avg2 represents the average PRB average utilization value corresponding to all activated users in the interval.

[0074] It should be noted that in the calculation process, in order to avoid the influence of errors caused by large data fluctuations, abnormal data can be processed accordingly. For example, it can be processed according to the following principles:

[0075] 1) If the number of samples involved in the interval of the number of activated users in the cell is ≥ the first threshold (for example, 100), the maximum value of n% and the minimum value of n% are respectively eliminated, and the value of n can be set as needed, for example, it can be set to 2 to 5;

[0076] 2) If the first threshold involved in the interval of the number of activated users in the cell is greater than the number of samples ≥ the second threshold (for example, 5), then one maximum value and one minimum value are eliminated;

[0077] 3) If the number of records involved in the interval of the number of activated users in the cell is less than the second threshold, the value of the interval is set to empty.

[0078] The above is only an example of abnormal data processing. Of course, there may be other processing methods, which are not limited in the embodiment of the present invention.

[0079] In a specific application, a cell data processing table - number of activated users vs. performance table can be generated according to the above ranking and calculation of the performance average, and the calculation results can be added to the table. The specific form of the cell data processing table is shown in Table 3 below.

[0080] Table 3

[0081]

[0082] Each row in the above Table 3 corresponds to a cell.

[0083] Read the cell data processing table - number of activated users VS performance table, and get the cell performance evaluation array according to the data format of machine learning.

[0084] The cell performance evaluation array may adopt the following format: [[activated user number interval 1, activated user number interval 1 performance mean], [activated user number interval 2, activated user number interval 2 performance mean], [], [], [], ... [activated user number interval N, activated user number interval N performance mean]].

[0085] Step 105: clustering the cells in the sequence of cells to be clustered using the performance evaluation array to obtain a cell performance clustering result.

[0086] The above-mentioned cell performance evaluation arrays are used to cluster the cells in the sequence of cells to be clustered to obtain a cell performance clustering result. The specific process is as follows:

[0087] By using machine learning and polynomial regression, the fitting degree of the performance of each cell in the to-be-clustered cell sequence under the same activated user number interval as other cells is calculated;

[0088] Two cells whose bidirectional fitting degrees both reach a set threshold (for example, the threshold may be set to 95%) are included in the same type of trend feature cell set;

[0089] If the bidirectional fit between a cell and all cells under a certain type of trend feature reaches a set threshold, the cell is included in the cell set of this type of trend feature.

[0090] According to the above principle, the above process is repeated for all cells in the sequence of cells to be clustered, and finally cluster fitting is completed to obtain a plurality of trend feature cell sets of different categories.

[0091] It should be noted that when calculating the performance fit of the same activated user number interval of the two cells, it is necessary to delete the abnormal activated user number interval of each cell and regenerate the cell performance evaluation array so that the data of the two cells can effectively correspond. For example, if there is no invalid performance data in a certain interval involved in the cell performance evaluation array A[] or B[] of the two cells A and B, then delete the data of A[] and B[] in the interval respectively, and regenerate the cell performance evaluation arrays A[] and B[].

[0092] In the embodiment of the present invention, the cell performance clustering result may include one or more types of trend characteristic cell sets, and each type of trend characteristic cell set may include one or more cells.

[0093] The communication network performance evaluation method provided by the present invention collects the performance data and cell information data of each cell respectively, generates a performance evaluation array of the cell according to the cell performance data for each cell of the same type, clusters the cells in the sequence of cells to be clustered using the performance evaluation array, and obtains the cell performance clustering result, so as to efficiently, conveniently and accurately evaluate the performance of the communication network, and further classify the high and low performance of cells in the entire network, formulate high and low performance trend standards, explore the differences and problems of high and low performance and abnormalities, and better support actual network production.

[0094] Furthermore, in another embodiment of the method of the present invention, the performance data of each cell in the same type of trend characteristic cell set can be used as a sample to fit the common performance curve of the cells of this type. For example, the python machine learning scikit-learn library is used to perform polynomial curve fitting, and finally the common performance curve of the cells of this type is fitted.

[0095] like Figure 2 As shown, an example of common performance curves of three types of cells obtained by using the communication network performance evaluation method of the present invention is shown. The common performance curve is used to evaluate the performance of various types of cells more conveniently, intuitively and accurately.

[0096] By analyzing the trend of performance curves of different categories, we can compare and analyze the performance of cells with high and low performance, performance of different manufacturers, and performance in different scenarios, explore the causes of low performance, extract high performance features, and further accurately evaluate and optimize network performance. In addition, we can also accurately estimate the capacity performance trend of the cell based on the performance curve to better support network planning and deployment.

[0097] Accordingly, an embodiment of the present invention also provides a communication network performance evaluation system, such as Figure 3 The figure shows a structural diagram of the system.

[0098] In this embodiment, the system includes the following modules:

[0099] The data collection module 301 is used to collect the efficiency data and the information data of each cell respectively;

[0100] A classification module 302, configured to classify cells according to the cell information data to obtain different types of cells;

[0101] A sorting module 303 is used to sort cells of the same type to obtain a sequence of cells to be clustered;

[0102] The performance evaluation array generation module 304 is used to generate the performance evaluation array of each cell in the sequence of cells to be clustered according to the cell performance data;

[0103] The clustering module 305 is used to cluster the cells in the sequence of cells to be clustered using the performance evaluation array to obtain a cell performance clustering result.

[0104] The specific parameters included in the cell performance data and the cell information data can refer to the description in the previous embodiment of the method of the present invention, which will not be repeated here. Accordingly, the classification module 302 can classify the cell information data according to one or more combinations of the data types included in the cell information data.

[0105] The above-mentioned sorting module 303 can specifically sort the cells of the same type according to the cell data richness and the number of cell data samples to obtain a sequence of cells to be clustered. Among them, the calculation of the cell data richness can refer to the description in the previous embodiment of the method of the present invention, that is, for each cell, the interval is divided according to the number of activated users to obtain multiple cell activated user number intervals; the cell data richness is determined according to the cell activated user number interval of the cell. When sorting the cells, the cells of the same type are mainly sorted from high to low according to the cell data richness, and the cells with the same data richness are sorted from high to low according to the number of cell data samples. Furthermore, the cells whose cell data richness is lower than the set value can also be eliminated to avoid affecting the accuracy of subsequent clustering results.

[0106] The above-mentioned performance evaluation array generation module 304 also needs to obtain the performance data corresponding to each interval based on the division of the number of activated users in each cell, and then generate the performance evaluation array of each cell. Specifically, the performance mean of the number of activated users in each cell in the cell is calculated, and then the performance evaluation array of the cell is generated according to the number of activated users in each cell in the cell and its performance mean. The calculation of the performance mean can refer to the description in the above-mentioned method embodiment of the present invention, which will not be repeated here.

[0107] The clustering module 305 can cluster the cells in the to-be-clustered cell sequence by calculating the performance fit under the same activated user number interval between different cells. Specifically, the clustering process can be completed by adopting the following principles:

[0108] 1) calculating the performance fitting degree between different cells in the sequence of cells to be clustered under the same interval of the number of activated users;

[0109] 2) Two cells whose bidirectional effectiveness fitting degrees both reach the set threshold are included in the same type of trend characteristic cell set;

[0110] 3) If the bidirectional performance fitting degree of a cell with all cells in a certain type of trend characteristic cell set reaches the set threshold, the cell is included in the trend characteristic cell set.

[0111] Through the above clustering process, the final cell performance clustering result can be obtained. The cell performance clustering result may include one or more types of trend characteristic cell sets, and each type of trend characteristic cell set may include one or more cells.

[0112] It should be noted that for all the cells to be evaluated, by classifying these cells, a sequence of cells to be clustered corresponding to different types can be obtained. For each sequence of cells to be clustered, cluster fitting calculation is performed on the cells therein to determine the trend characteristic cell set to which the cell belongs.

[0113] The communication network performance evaluation system provided by the present invention collects the performance data and cell information data of each cell respectively, generates a performance evaluation array of the cell according to the cell performance data for each cell of the same type, and clusters the cells in the sequence of cells to be clustered using the performance evaluation array to obtain the cell performance clustering results, thereby efficiently, conveniently and accurately evaluating the performance of the communication network, and further classifies the high and low performance of cells in the entire network, formulates high and low performance trend standards, explores the differences and problems of high and low performance and abnormalities, and better supports actual network production.

[0114] Furthermore, if Figure 4 As shown, in another embodiment of the communication network performance evaluation system of the present invention, it can further include: a performance curve generation module 306, which is used to generate a common performance curve of the corresponding type of cell based on the performance data of all cells in each type of trend characteristic cell set. Specifically, the performance data of each cell in the same type of trend characteristic cell set can be used as a sample to fit the common performance curve of the cell of this type. For example, the python machine learning scikit-learn library is used to perform polynomial curve fitting, and finally the common performance curve of the cell of this type is fitted.

[0115] In this way, by analyzing the trend of performance curves of different categories, we can compare and analyze the performance of cells with high and low performance, performance of different manufacturers, and performance in different scenarios, explore the causes of low performance, extract high performance features, and further accurately evaluate and optimize network performance. In addition, we can also accurately estimate the capacity performance trend of the cell based on the performance curve to better support network planning and deployment.

[0116] It should be noted that the terms "including" and "having" and any variations thereof in the specification and claims of the present invention and the above-mentioned drawings are intended to cover non-exclusive inclusions. For example, a process, method, system, product or apparatus comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or apparatus.

[0117] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. Moreover, the system embodiments described above are merely schematic, in which the modules and units described as separate components may or may not be physically separated, that is, they may be located on one network unit, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without creative work.

[0118] The embodiments of the present invention are described in detail above. The present invention is described in detail using specific implementation methods herein. The description of the above embodiments is only used to help understand the method and device of the present invention. It is only a part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should belong to the scope of protection of the present invention, and the content of this specification should not be understood as limiting the present invention. Therefore, any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

[0119] The communication network performance evaluation method provided by the present invention collects the performance data and cell information data of each cell respectively, generates a performance evaluation array of the cell according to the cell performance data for each cell of the same type, clusters the cells in the sequence of cells to be clustered using the performance evaluation array, and obtains the cell performance clustering result, so as to efficiently, conveniently and accurately evaluate the performance of the communication network, and further classify the high and low performance of cells in the entire network, formulate high and low performance trend standards, explore the differences and problems of high and low performance and abnormalities, and better support actual network production.

Claims

1. A communication network performance evaluation method, characterized in that: The method comprises: Collecting the efficiency data and information data of each cell respectively; Classifying cells according to the cell information data to obtain different types of cells; Sort the cells of the same type to obtain a sequence of cells to be clustered; For each cell in the sequence of cells to be clustered, generating a performance evaluation array of the cell according to the cell performance data; Clustering cells in the sequence of cells to be clustered using the performance evaluation array to obtain a cell performance clustering result; The ordering of cells of the same type to obtain a sequence of cells to be clustered comprises: For each cell, the interval is divided according to the number of activated users, and multiple intervals of the number of activated users in the cell are obtained; Determining the cell data enrichment degree according to the cell activated user number interval of the cell; The cells of the same type are sorted according to the cell data richness and the number of cell data samples to obtain a sequence of cells to be clustered.

2. The method according to claim 1, characterized in that The cell performance data includes: time, area name, cell identifier CGI, number of activated users, PRB utilization, and traffic; The cell information data includes: area name, manufacturer name, cell identifier CGI, communication standard, cell frequency bandwidth, cell channel number, and cell type.

3. The method according to claim 2, characterized in that The classifying the cell information data comprises: The cell information data is classified according to one or more combinations of data types included in the cell information data.

4. The method according to claim 1, characterized in that The determining of the cell data enrichment degree according to the cell activated user number interval of the cell comprises: The cell data richness is calculated, where the cell data richness=the number of cell activated user number intervals for which performance data exists / the total number of cell activated user number intervals.

5. The method according to claim 1, characterized in that The process of sorting the cells of the same type according to the cell data richness and the cell data sampling quantity to obtain the cell sequence to be clustered includes: The cells of the same type are sorted from high to low according to the cell data richness, and the cells of the same data richness are sorted from high to low according to the number of cell data sampling.

6. The method according to claim 1, characterized in that The method further comprises: Eliminate cells whose cell data enrichment is lower than the set value.

7. The method according to claim 6, characterized in that The step of generating the cell performance evaluation array according to the cell performance data comprises: Calculating the performance mean of the interval of the number of activated users in each cell in the cell; According to the intervals of the number of activated users in each area of ​​the cell and the mean values ​​of their performance, a performance evaluation array of the cell is generated.

8. The method according to claim 7, characterized in that The clustering of cells in the sequence of cells to be clustered using the performance evaluation array to obtain a cell performance clustering result includes: Calculating the performance fitting degree between different cells in the sequence of cells to be clustered in pairs under the same interval of the number of activated users; Two cells whose bidirectional effectiveness fitting degrees both reach the set threshold are included in the same type of trend characteristic cell set; If the bidirectional performance fitting degree of a cell with all cells in a certain type of trend characteristic cell set reaches the set threshold, the cell is included in the trend characteristic cell set.

9. The method according to any one of claims 1 to 8, characterized in that: The cell performance clustering result includes one or more types of trend feature cell sets; the method further includes: According to the performance data of all cells in the trend characteristic cell cluster, a common performance curve of the corresponding type of cells is generated.

10. A communication network performance evaluation system, characterized in that: The system comprises: A data collection module, used to collect the efficiency data and information data of each cell respectively; A classification module, used to classify cells according to the cell information data to obtain different types of cells; A sorting module is used to sort cells of the same type to obtain a sequence of cells to be clustered; An efficiency evaluation array generation module, used for generating an efficiency evaluation array of each cell in the sequence of cells to be clustered according to the cell efficiency data; A clustering module, used for clustering cells in the sequence of cells to be clustered using the performance evaluation array to obtain a cell performance clustering result; The ordering of cells of the same type to obtain a sequence of cells to be clustered comprises: For each cell, the interval is divided according to the number of activated users, and multiple intervals of the number of activated users in the cell are obtained; Determining the cell data enrichment degree according to the cell activated user number interval of the cell; The cells of the same type are sorted according to the cell data richness and the number of cell data samples to obtain a sequence of cells to be clustered.

11. The system according to claim 10, characterized in that The cell performance clustering result includes one or more types of trend feature cell sets; the system also includes: The performance curve generation module is used to generate a common performance curve of the corresponding type of cells based on the performance data of all cells in the various trend characteristic cell sets.

Images