CN112749745A - Automatic 5G base station cell service scene identification method based on machine learning - Google Patents
Automatic 5G base station cell service scene identification method based on machine learning Download PDFInfo
- Publication number
- CN112749745A CN112749745A CN202110024060.7A CN202110024060A CN112749745A CN 112749745 A CN112749745 A CN 112749745A CN 202110024060 A CN202110024060 A CN 202110024060A CN 112749745 A CN112749745 A CN 112749745A
- Authority
- CN
- China
- Prior art keywords
- traffic data
- telephone traffic
- cell
- data
- service
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 27
- 238000010801 machine learning Methods 0.000 title claims abstract description 13
- 239000000203 mixture Substances 0.000 claims abstract description 7
- 238000004458 analytical method Methods 0.000 claims abstract description 5
- 238000000513 principal component analysis Methods 0.000 claims abstract description 4
- 238000010606 normalization Methods 0.000 claims description 7
- 230000033228 biological regulation Effects 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 3
- 238000004134 energy conservation Methods 0.000 description 2
- 238000005265 energy consumption Methods 0.000 description 2
- 238000007726 management method Methods 0.000 description 2
- 238000013459 approach Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000007621 cluster analysis Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000009440 infrastructure construction Methods 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
-
- G06Q50/40—
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W24/00—Supervisory, monitoring or testing arrangements
- H04W24/08—Testing, supervising or monitoring using real traffic
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W52/00—Power management, e.g. TPC [Transmission Power Control], power saving or power classes
- H04W52/02—Power saving arrangements
- H04W52/0203—Power saving arrangements in the radio access network or backbone network of wireless communication networks
- H04W52/0206—Power saving arrangements in the radio access network or backbone network of wireless communication networks in access points, e.g. base stations
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D30/00—Reducing energy consumption in communication networks
- Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks
Abstract
The invention relates to the field of big data, in particular to a method for automatically identifying a cell service scene of a 5G base station based on machine learning, which realizes the automatic identification of the cell service scene and greatly improves the accuracy of the cell service scene identification. The technical scheme is summarized as acquiring hourly traffic data, PRB utilization rate and service switching times of each cell; then, the PRB utilization rate and the service switching times are normalized or labeled; integrating the daily telephone traffic data into telephone traffic data of a plurality of time periods, and adding the processed PRB utilization rate and the service switching times to form multidimensional telephone traffic data; reducing the multidimensional telephone traffic data into two-dimensional telephone traffic data through principal component analysis; then, profile coefficient analysis and service logic are introduced to gather the two-dimensional telephone traffic data into a plurality of categories; and finally, carrying out classification clustering on the two-dimensional telephone traffic data through a Gaussian mixture model, and automatically identifying a cell scene according to a clustering result. The method and the device are suitable for identifying the cell service scene.
Description
Technical Field
The invention relates to the field of big data, in particular to a method for automatically identifying a cell service scene of a 5G base station based on machine learning.
Background
In the field of industrial internet, the aims of energy conservation and consumption reduction of industrial equipment are fulfilled, green development is realized, cost reduction and efficiency improvement are realized, and the method is a key direction for enterprise development.
As a 5G network for new infrastructure construction, the network support system supports various complex network applications by bearing tasks of the new infrastructure, provides higher bandwidth and shorter time delay for traditional network data application, thereby assisting in application upgrade and reducing network construction cost; there is a network slice management channel support that provides customization for innovative product applications.
Therefore, in the present day of various application wind and cloud surge, the traditional cell scene identification and classification method for the 3G and 4G service model is difficult to express the new development trend of the current 5G network.
The prior art mainly performs the following classifications by data analysis:
(1) based on the purpose of the specified cell capacity expansion planning standard, the ratio of the service volume to the number of users is divided into cells of 'big packet transmission', 'small packet transmission' and 'middle packet transmission'. Therefore, the set capacity expansion threshold is as follows: small packet < medium packet < large packet.
(2) The classification of the cell scenes is mainly divided according to the geographic area where the wireless base station is located, and the cell scenes comprise business super scenes, school scenes, office building scenes, dense urban areas, suburban areas and the like.
(3) And classifying the user behavior characteristics into indoor users, outdoor users, users with low mobility and users with high mobility according to the planning requirement of the wireless cell, so that the scenes covered by the wireless cell are divided into indoor scenes, outdoor scenes, low mobility areas and high mobility areas.
The defects of the technology are as follows:
(1) from the perspective of wireless network planning and wireless network optimization, the scene division is carried out, and the cell characteristics are not classified from the purposes of energy consumption management and energy conservation.
(2) The classification of the size packet takes the average value of the day, week and month of the statistical data as the calculation basis, and cannot reflect the change condition of the granularity of the finer date.
(3) Due to the lack of statistical methods, for a mixed type scene, such as a cell covering both an office building and a highway, the scene type cannot be accurately identified, and the corresponding cell service characteristics cannot be reflected.
Disclosure of Invention
The invention aims to provide a method for automatically identifying a cell service scene of a 5G base station based on machine learning, which realizes the automatic identification of the cell service scene and greatly improves the accuracy of cell service scene identification.
The invention adopts the following technical scheme to realize the aim, and the method for automatically identifying the cell service scene of the 5G base station based on machine learning comprises the following steps:
step (1), acquiring hourly telephone traffic data, PRB (physical Resource Block) utilization rate and service switching times of each cell;
step (2), the PRB utilization rate and the service switching times are normalized or labeled;
step (3), integrating the daily telephone traffic data into telephone traffic data of a plurality of time periods, and then adding the processed PRB utilization rate and the service switching times to form multidimensional telephone traffic data;
step (4), reducing the multidimensional telephone traffic data into two-dimensional telephone traffic data through principal component analysis;
and (6) carrying out classification clustering on the two-dimensional telephone traffic data through a Gaussian mixture model, and automatically identifying a cell scene according to a clustering result.
Further, in the step (1), after the corresponding data is acquired, the data missing values are filled with the mean values, and the data values larger than the set threshold are removed.
Further, in step (2), the formula of the normalization process is x ═ x-u)/δ, where x is sample data, μ is a mean value of all sample data, and δ is a standard deviation of all sample data.
Further, in step (2), the normalization process is formulated as max (x) is the maximum value of the sample data, and min (x) is the minimum value of the sample data.
Further, before integrating the daily traffic data in step (3), the method further includes: and calculating the percentage of the traffic data in each hour to the traffic data in the corresponding day.
Further, in the step (6), the result obtained by performing classification clustering on the two-dimensional telephone traffic data through the gaussian mixture model comprises the probability of belonging to each cell service scene, and the threshold value regulation and control of the energy-saving threshold value are performed on each cell service site according to the corresponding probability.
The invention counts the telephone traffic of each cell per hour, integrates the traffic data in different periods, reflects the cell service scene type from the data change of more detailed dimension, adds PRB utilization rate and service switching frequency, the PRB utilization rate can obtain the service condition of each cell service site, the service switching frequency can distinguish and connect more service sites, forms multidimensional data, introduces related models for cluster analysis through dimension reduction processing, automatically identifies the cell service scene according to the analysis result, greatly improves the accuracy of cell service scene identification, performs threshold regulation of energy-saving threshold on each cell service site according to the corresponding probability obtained after clustering, and also realizes energy-saving control.
Drawings
FIG. 1 is a 24 hour trend of traffic for different scenarios of the present invention.
FIG. 2 is a flowchart of a method for automatically identifying a cell service scenario of a 5G base station based on machine learning according to the present invention.
Detailed Description
The invention relates to a method for automatically identifying a 5G base station cell service scene based on machine learning, wherein the flow chart of the method is shown in figure 2, and the method comprises the following steps:
105, introducing contour coefficient analysis and service logic to gather two-dimensional telephone traffic data into a plurality of categories;
and 106, carrying out classification clustering on the two-dimensional telephone traffic data through a Gaussian mixture model, and automatically identifying a cell scene according to a clustering result.
In step 101, after acquiring corresponding data, filling the missing data values with a mean value, and removing data values larger than a set threshold.
In step 102, the formula of the normalization process is x ═ x-u)/δ, where x is sample data, μ is the mean of all sample data, and δ is the standard deviation of all sample data.
In step 102, the normalization process has the formula that max (x) is the maximum value of the sample data, and min (x) is the minimum value of the sample data.
Before integrating the daily traffic data in step 103, the method further includes: and calculating the percentage of the traffic data in each hour to the traffic data in the corresponding day.
In step 106, the result obtained by performing classification clustering on the two-dimensional traffic data through the gaussian mixture model includes the probability of belonging to each cell service scene, and the threshold regulation of the energy-saving threshold is performed on each cell service site according to the corresponding probability, so that the energy-saving control is realized.
Fig. 1 is a 24-hour traffic ratio trend chart in different scenarios, which shows the change of traffic flow in 24 hours in different scenarios, where a traffic flow in scenario one gradually decreases and approaches after 0 point and gradually recovers after 0 point 5 point, and an obvious peak appears at 7 point 8 point, and then a small-amplitude drop is maintained at a higher level; the traffic flow of a scene II is gradually reduced to 0 from 0 point and then increased to 16 points in a small amplitude, and is rapidly increased after 16 points and reaches a peak from 19 points to 20 points; the traffic flow of scene three telephone traffic is gradually reduced to 0 from 0 point, gradually recovered from 5 points and gradually reduced after reaching a peak from 17 points to 18 points; in the scene four, the traffic flow drops slightly after 0 point, gradually rises after 3 points, reaches a first peak after 8 to 9 points, and then gradually falls after 21 points.
The traffic flow data change of each hour of the whole day is monitored, the cell service scene type can be embodied from smaller dimension, the utilization rate of the PRB can approximately obtain the station use condition of each time period, and the switching times are added to better distinguish the VIP stations such as hospitals, high-speed rail stations and the like, so that the mixed scene can be better distinguished.
In conclusion, the invention realizes the automatic identification of the cell service scene, and greatly improves the accuracy of the cell service scene identification.
Claims (6)
1. The method for automatically identifying the service scene of the 5G base station cell based on machine learning is characterized by comprising the following steps:
step (1), acquiring hourly telephone traffic data, PRB utilization rate and service switching times of each cell;
step (2), the PRB utilization rate and the service switching times are normalized or labeled;
step (3), integrating the daily telephone traffic data into telephone traffic data of a plurality of time periods, and then adding the processed PRB utilization rate and the service switching times to form multidimensional telephone traffic data;
step (4), reducing the multidimensional telephone traffic data into two-dimensional telephone traffic data through principal component analysis;
step 5, introducing contour coefficient analysis and service logic to gather the two-dimensional telephone traffic data into a plurality of categories;
and (6) carrying out classification clustering on the two-dimensional telephone traffic data through a Gaussian mixture model, and automatically identifying a cell scene according to a clustering result.
2. The method for automatically identifying a cell service scene of a 5G base station based on machine learning as claimed in claim 1, wherein in the step (1), after the corresponding data is obtained, the missing data values are filled with the mean value, and the data values larger than the set threshold are removed.
3. The method according to claim 1 or 2, wherein in step (2), the formula of the normalization process is x' ═ x-u)/δ, x is sample data, μ is the mean of all sample data, and δ is the standard deviation of all sample data.
5. The method for automatically identifying 5G base station cell service scenes based on machine learning of claim 4, wherein in the step (3), before integrating the traffic data of each day, further comprising: and calculating the percentage of the traffic data in each hour to the traffic data in the corresponding day.
6. The method for automatically identifying a cell service scenario of a 5G base station based on machine learning as claimed in claim 5, wherein in step (6), the result obtained by performing classification clustering on the two-dimensional traffic data through a Gaussian mixture model comprises a probability of belonging to each cell service scenario, and the threshold of the energy-saving threshold is adjusted and controlled for each cell service site according to the corresponding probability.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110024060.7A CN112749745B (en) | 2021-01-08 | 2021-01-08 | Automatic 5G base station cell service scene identification method based on machine learning |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110024060.7A CN112749745B (en) | 2021-01-08 | 2021-01-08 | Automatic 5G base station cell service scene identification method based on machine learning |
Publications (2)
Publication Number | Publication Date |
---|---|
CN112749745A true CN112749745A (en) | 2021-05-04 |
CN112749745B CN112749745B (en) | 2021-10-12 |
Family
ID=75650453
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110024060.7A Active CN112749745B (en) | 2021-01-08 | 2021-01-08 | Automatic 5G base station cell service scene identification method based on machine learning |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN112749745B (en) |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106211194A (en) * | 2016-07-28 | 2016-12-07 | 武汉虹信技术服务有限责任公司 | The outer separation method of a kind of MR data room based on statistical model |
CN109635858A (en) * | 2018-12-03 | 2019-04-16 | 北京工业大学 | A kind of physical area identification distribution method based on fuzzy layered cluster |
CN109699035A (en) * | 2017-10-20 | 2019-04-30 | 中国移动通信集团浙江有限公司 | A kind of subway network scene cell recognition method and device |
CN109996246A (en) * | 2017-12-30 | 2019-07-09 | 中国移动通信集团辽宁有限公司 | Power-economizing method, device, equipment and the medium of base station cell |
CN110493803A (en) * | 2019-09-17 | 2019-11-22 | 南京邮电大学 | A kind of cell scenario division methods based on machine learning |
KR102066018B1 (en) * | 2019-10-21 | 2020-01-14 | 주식회사 한일엔지니어링 | Database construction system for realty information using network |
CN110996377A (en) * | 2019-11-25 | 2020-04-10 | 宜通世纪科技股份有限公司 | Base station energy saving method, system, device and storage medium |
CN111918319A (en) * | 2019-05-08 | 2020-11-10 | 中国移动通信集团福建有限公司 | Busy hour busy area prediction method and device |
-
2021
- 2021-01-08 CN CN202110024060.7A patent/CN112749745B/en active Active
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106211194A (en) * | 2016-07-28 | 2016-12-07 | 武汉虹信技术服务有限责任公司 | The outer separation method of a kind of MR data room based on statistical model |
CN109699035A (en) * | 2017-10-20 | 2019-04-30 | 中国移动通信集团浙江有限公司 | A kind of subway network scene cell recognition method and device |
CN109996246A (en) * | 2017-12-30 | 2019-07-09 | 中国移动通信集团辽宁有限公司 | Power-economizing method, device, equipment and the medium of base station cell |
CN109635858A (en) * | 2018-12-03 | 2019-04-16 | 北京工业大学 | A kind of physical area identification distribution method based on fuzzy layered cluster |
CN111918319A (en) * | 2019-05-08 | 2020-11-10 | 中国移动通信集团福建有限公司 | Busy hour busy area prediction method and device |
CN110493803A (en) * | 2019-09-17 | 2019-11-22 | 南京邮电大学 | A kind of cell scenario division methods based on machine learning |
KR102066018B1 (en) * | 2019-10-21 | 2020-01-14 | 주식회사 한일엔지니어링 | Database construction system for realty information using network |
CN110996377A (en) * | 2019-11-25 | 2020-04-10 | 宜通世纪科技股份有限公司 | Base station energy saving method, system, device and storage medium |
Non-Patent Citations (2)
Title |
---|
JINGJIN WU ET AL.: "Energy-efficient base-stations sleep-mode techniques in green cellular networks:a survey", 《IEEE COMMUNICATION SURVEYS & TUTORIALS》 * |
张青: "基于AI的5G基站节能场景智能识别", 《5G网络规划》 * |
Also Published As
Publication number | Publication date |
---|---|
CN112749745B (en) | 2021-10-12 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106912015B (en) | Personnel trip chain identification method based on mobile network data | |
WO2021209024A1 (en) | Energy-saving method, base station, control unit, and storage medium | |
CN112488322A (en) | Federal learning model training method based on data feature perception aggregation | |
CN1333984A (en) | Determining subscriber demands on communiations system | |
CN101287269B (en) | Mobile communication network optimizing method, device and system | |
CN112001829B (en) | Population distribution judging method based on mobile phone signaling data | |
CN111314940A (en) | Wireless network deployment method for 5G NSA networking mode | |
CN108243435B (en) | Parameter optimization method and device in LTE cell scene division | |
CN109462853B (en) | Network capacity prediction method based on neural network model | |
CN202134048U (en) | Scenic area visitor distribution statistical system | |
CN107222871B (en) | TD-LTE 230 wireless private network power base station planning method | |
CN112399458A (en) | Big data analysis method for mobile communication network flow | |
CN109348501B (en) | Indoor and outdoor distinguishing method based on LTE (Long term evolution) signals | |
CN112749745B (en) | Automatic 5G base station cell service scene identification method based on machine learning | |
CN108271160B (en) | Network resource optimization method and device | |
CN103037375B (en) | Method and device for dividing community telephone traffic scenes | |
CN111275073A (en) | Regional people flow analysis method based on mobile phone signaling data | |
CN111194045A (en) | Energy-saving method based on user group aggregation behavior model | |
CN115955698A (en) | 5G network slicing system based on smart power grid | |
WO2020215282A1 (en) | Method and apparatus for evaluate data traffic depressed by radio issues | |
Jiang et al. | Research on location planning of 5G base station based on DBSCAN clustering algorithm | |
CN112954715B (en) | Wireless service node capacity estimation method based on transfer learning | |
Wang et al. | Multi-dimensional prediction model for cell traffic in city scale | |
CN102752776B (en) | The acquisition methods of wireless network utilization and device | |
CN110991791B (en) | Power distribution station resource fusion and allocation method and system based on fog calculation model |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |