CN116347461A

CN116347461A - Method and device for adjusting antenna in 5G network

Info

Publication number: CN116347461A
Application number: CN202111601201.3A
Authority: CN
Inventors: 刘兵
Original assignee: Bright Oceans Inter Telecom Co Ltd
Current assignee: Bright Oceans Inter Telecom Co Ltd
Priority date: 2021-12-24
Filing date: 2021-12-24
Publication date: 2023-06-27

Abstract

The invention discloses a method for adjusting an antenna in a 5G network, which comprises the following steps: acquiring a 5G measurement report, and positioning the measurement report to a geographic grid by combining an electronic map and a cell industrial parameter; establishing a cell coverage model according to the cell project parameters and the geographic grid; acquiring the regional user number and the traffic volume in the geographic grid according to the measurement report, and establishing a regional user traffic model; establishing a cell comprehensive model according to the cell coverage model and the regional user service model in combination with a large-scale antenna parameter library, training the cell comprehensive model by using full data, and iterating parameters of the cell comprehensive model; and generating antenna adjustment parameters of the cell according to the cell comprehensive model by taking the number of cell users and the coverage area as targets. The invention also discloses an adjusting device of the antenna in the 5G network.

Description

Method and device for adjusting antenna in 5G network

Technical Field

The present invention relates to the mobile communication industry, and more particularly, to a mobile communication network optimization technique.

Background

The 5G enables all things to be interconnected or to be realized, application scenes such as smart cities, smart home, internet of vehicles, automatic driving, VR, AR and the like are supported by a network environment, the life of people and even all industries are influenced, and business upgrading is promoted.

The main frequency bands of the current Chinese 5G network construction are 2.6G frequency bands, 3.3-3.6GHz frequency bands and 4.8-5.0GHz frequency bands, and because of the characteristics of wireless electromagnetic wave propagation, a communication network is constructed on the frequency bands, and more base stations or cells are needed to meet the coverage requirements than the traditional 800M and 1800M frequency bands, so that the number of antenna feeds is greatly increased, and a great amount of manpower and material resources are consumed for acquiring an antenna feed adjustment scheme by virtue of engineer experience analysis, so that the operation and maintenance cost of operators is greatly increased.

In a 5G system, a Massive MIMO smart antenna with 64T64R will be used, and the NR protocol provides a broadcast beamforming/scanning means, and both horizontal and vertical dimensions provide dynamic narrow beams. The number of antenna elements is large, the wave beam mode is diversified, the traditional method for manually testing and analyzing to determine the adjustment scheme cannot determine the optimal scheme, and the auxiliary analysis cannot be fully performed by combining big data generated by a user.

Therefore, there is a need for an antenna adjustment technique that adjusts antenna parameters based on the large amount of network coverage data generated by mobile network users using the network, and the amount of users and traffic, in combination with the basic situation of the 5G cell. The 5G network coverage is optimized, and the user perception promotion and the input-output ratio of a network operator are improved.

Disclosure of Invention

The invention provides a method for adjusting an antenna in a 5G network, which comprises the following steps:

acquiring a 5G measurement report, and positioning the measurement report to a geographic grid by combining an electronic map and a cell industrial parameter;

establishing a cell coverage model according to the cell project parameters and the geographic grid;

acquiring the regional user number and the traffic volume in the geographic grid according to the measurement report, and establishing a regional user traffic model;

establishing a cell comprehensive model according to the cell coverage model and the regional user service model in combination with a large-scale antenna parameter library, training the cell comprehensive model by using full data, and iterating parameters of the cell comprehensive model;

and generating antenna adjustment parameters of the cell according to the cell comprehensive model by taking the number of cell users and the coverage area as targets.

Further, the method for establishing the cell coverage model according to the cell engineering parameters and the geographic grid specifically comprises the following steps:

acquiring a center point of a cell corresponding to the geographic grid;

establishing a corresponding relation between antenna parameters in the cell industrial parameters and the geographic grid center point;

and constructing a comprehensive cost function of the three dimensions of cell coverage, repeated coverage and service absorption based on the corresponding relation.

Further, the method for establishing the regional user service model specifically includes the steps of:

acquiring the measurement report in the geographic grid in a time period;

determining the total number of users and the total number of traffic in the geographic grid according to the acquired measurement report;

and correcting the regional user service model according to the historical measurement report and the current network measurement report.

Further, the method for establishing a cell comprehensive model according to the cell coverage model and the regional user service model in combination with a large-scale antenna parameter library, training the cell comprehensive model by using full data, and iterating parameters of the cell comprehensive model specifically comprises the following steps:

creating a sample set through the full data, and cleaning and standardizing the sample;

carrying out sample distribution modeling on each index in the sample set in combination with the cell coverage model, the regional user service model and a large-scale antenna parameter library, and representing the relation among different antenna parameters, regional coverage and regional service volumes of the cell;

substituting the total data into the cell comprehensive model for training, and taking the corresponding parameter value as an optimal solution when the comprehensive cost function of all the geographic grids in the coverage area of the single cell is maximum;

And iterating the parameters of the cell comprehensive model according to the parameter values corresponding to the optimal solution.

Further, the method for generating the antenna adjustment parameters of the cell according to the cell comprehensive model by taking the number of users, the traffic volume and the coverage area of the cell as targets specifically comprises the following steps:

and substituting the cell user quantity, the service quantity and the coverage area into the cell comprehensive model to generate antenna azimuth angle weight, horizontal wave width weight, downward inclination angle weight and vertical wave width weight antenna adjustment parameters.

The invention also discloses an adjusting device of the antenna in the 5G network, which is characterized in that the device comprises:

the data acquisition module is used for acquiring a 5G measurement report, an electronic map and a cell industrial parameter;

the positioning geographic grid module is used for positioning the measurement report to the geographic grid by combining the electronic map and the cell industrial parameter according to the 5G measurement report acquired by the data acquisition module;

the cell coverage model module is used for establishing a cell coverage model according to the cell engineering parameters and the geographic grids positioned by the positioning geographic grid module;

the user service model module is used for acquiring the regional user number and the service volume in the geographic grid according to the measurement report and establishing a regional user service model;

The cell comprehensive model module is used for establishing a cell comprehensive model according to the cell coverage model established by the cell coverage model module and the regional user service model established by the user service model module and combining a large-scale antenna parameter library, training the cell comprehensive model by using full data, and iterating parameters of the cell comprehensive model;

and the parameter tuning unit is used for generating antenna tuning parameters of the cell according to the cell comprehensive model of the cell comprehensive model module by taking the number of cell users and the coverage area as targets.

Specifically, the cell coverage model module further includes:

the geographical grid center point acquisition module is used for acquiring the center point of the geographical grid corresponding to the cell;

the corresponding relation establishing module is used for establishing a corresponding relation between the antenna parameters in the cell industrial parameters and the geographic grid center point;

and the comprehensive cost function construction module is used for constructing the comprehensive cost functions of the three dimensions of cell coverage, repeated coverage and service absorption based on the corresponding relation.

Specifically, the user service model module further includes:

the grid data acquisition module is used for acquiring the measurement report in the geographic grid positioned by the positioning geographic grid module in the time period;

The user quantity service determining module is used for determining the total quantity of the user quantity and the service quantity in the geographic grid according to the measurement report in the geographic grid acquired by the grid data acquiring module;

and the model correction module is used for correcting the regional user service model according to the historical measurement report and the current network measurement report.

Specifically, the cell comprehensive model module further includes:

the sample collection module is used for creating a sample collection through the full data and cleaning and standardizing the samples;

the sample distribution modeling module is used for carrying out sample distribution modeling on each index of the sample set created by the sample set module in combination with the cell coverage model, the regional user service model and the large-scale antenna parameter library, and representing the relationship among different antenna parameters, regional coverage and regional traffic of the cell;

the model training module is used for training the cell comprehensive model of the sample distribution modeling module by using the full data to be substituted into the cell comprehensive model;

a parameter iteration module; and the method is used for obtaining the parameter value corresponding to the maximum value of the comprehensive cost function of all the geographic grids in the coverage area of the single cell as an optimal solution, and iterating the parameters of the cell comprehensive model according to the parameter value corresponding to the optimal solution.

The parameter tuning unit is used for generating antenna tuning parameters of the cell according to the cell comprehensive model of the cell comprehensive model module by taking the number of cell users and coverage area as targets, and specifically comprises the following steps:

According to the technical scheme, the antenna adjustment technology in the 5G network disclosed by the embodiment of the invention is used for positioning the geographic grid based on MR data, accurately and objectively evaluating the coverage condition of the geographic grid area, and optimizing the antenna optimal weight according to the geographic grid area, so that the network is self-adaptively provided with optimization taking a value user as a direction, and the input-output ratio of an operator is accurately and reliably improved; and (3) evaluating the comprehensive cost function of the model by establishing a cell comprehensive model, screening an optimization scheme, and performing repeated cyclic iterative computation to finally obtain the optimization scheme, so that the optimization scheme is suitable for the current condition of the network, and can be adaptively adjusted according to the development of the network.

Drawings

In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

Fig. 1 is a flow chart of a method for adjusting an antenna in a 5G network according to an embodiment of the present invention;

FIG. 2 is a flow chart of a method according to a second embodiment of the present invention;

FIG. 2-1 is a diagram illustrating grid positioning according to a second embodiment of the present invention;

fig. 2-2 is a schematic diagram illustrating absorption at a sampling point according to a second embodiment of the present invention;

FIG. 3 is a flow chart of a method according to a third embodiment of the present invention;

FIG. 4 is a flow chart of a method according to a fourth embodiment of the present invention;

fig. 5 is a structural diagram of an adjusting device for an antenna in a 5G network according to a fifth embodiment of the present invention;

fig. 6 is a detailed structure diagram of a device according to a sixth embodiment of the present invention.

Detailed Description

The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without any inventive effort, are intended to be within the scope of the invention.

Referring to fig. 1, a first embodiment of the present invention is shown in fig. 1, and a method for adjusting an antenna in a 5G network in the present invention is described in detail, including the following steps:

Step S1: and acquiring a 5G measurement report, and positioning the measurement report to a geographic grid by combining an electronic map and a cell industrial parameter.

The 5G measurement report is the measurement of the network and reported to the network in the process of using the 5G network by a user, and can be acquired in a soft acquisition mode or in a gNB acquisition and reporting mode, and comprises UE information, cell information, longitude and latitude information, channel measurement information such as CSI-RSRP, SSB-RSRP and the like.

The 5G positioning comprises indoor and outdoor positioning, combines a high-precision electronic map and 5G network cell engineering parameters, positions the MR to a geographic grid of a proper area, and changes the coverage area of the grid according to actual conditions, wherein the geographic grid can be 5 x 5m or 10 x 10m (length, width and height), or other dimensions according with the actual conditions, so as to evaluate the coverage condition, the traffic density and the regional traffic volume of the 5G network based on the geographic positions. The outdoor positioning is mainly performed according to high-precision longitude and latitude information in MR, and a base station positioning mode is used in an auxiliary mode. Indoor positioning mainly uses schemes such as WLAN positioning technology, fingerprint library positioning and the like to perform indoor positioning, and the spatial position is used for assisting automatic adjustment of Massive MIMO antenna parameters.

And by combining the technologies of MR-OTT or MDT and the like, the geographic rasterization positioning of the sampling points can be realized.

Step S2: and establishing a cell coverage model according to the cell engineering parameters and the geographic grid.

The cell engineering parameters comprise cell marks, neighbor cell information, transmitting power, coverage area, base station information, antenna parameters (antenna height, longitude and latitude, antenna direction angle, downtilt angle) and the like.

The coverage model of the cell is established according to the cell engineering parameters and the geographic grids, namely, the corresponding relation between the direction angle and the downward inclination angle of the cell antenna and the relative direction angle and the relative downward inclination angle of the antenna at the center point of the physical grid is established.

Step S3: and acquiring the regional user number and the service volume in the geographic grid according to the measurement report, and establishing a regional user service model.

According to the MR rasterization positioning result, the number of users in the grid, the network model and the time length of the grid users can be calculated by combining with the measurement report, or the calculation is carried out by combining with the analysis of a universal NG-U interface. Accumulating the number of users, service types, service volume and the like in a long-time area, establishing an area user service model, wherein the user service model comprises the total number of users and the total number of service volumes in the area, and the specific service type is used as the basis for selecting antennas and adjusting antennas in a specific area, so that the interference in the network is reduced as much as possible. And the regional service model can be corrected according to the historical data and the current network data accumulated for a long time.

Step S4: and establishing a cell comprehensive model according to the cell coverage model and the regional user service model in combination with a large-scale antenna parameter library, training the cell comprehensive model by using full data, and iterating parameters of the cell comprehensive model.

The large-scale antenna parameter library (Massive MIMO antenna parameter library) is obtained from actual statistics of operator network construction departments, and specific parameters include cell names, antenna types, antenna manufacturers, antenna lobe width weights, antenna height weights, antenna azimuth weights, antenna downtilt angles weights, antenna numbers, coverage areas, levels and the like.

For a 5G cell, a cell comprehensive model is built by comprehensively considering three aspects of a cell coverage model, a regional service model and a Massive MIMO antenna parameter library.

Wherein α, β, γ are adjustment parameters of coverage, area traffic, antenna parameters.

The main characteristic parameters comprise a cell coverage model, the number of users, the traffic volume, the service duration, the number of MR stripes, an azimuth angle weight, a horizontal wave width weight, a downtilt angle weight, a vertical wave width weight and the like.

After modeling is completed, intelligent sample training is carried out through full data, and iteration and optimization are carried out on the model through continuous algorithm optimization.

Step S5: and generating antenna adjustment parameters of the cell according to the cell comprehensive model by taking the number of cell users and the coverage area as targets.

The output optimization scheme adjusting parameters such as azimuth angle weight, horizontal wave width weight, downtilt angle weight and vertical wave width weight are automatically modified, and the network traffic and business change are self-adapted.

The method aims at the number of the cell users and the coverage area, namely, according to the target number of the cell users and the target coverage area of the cell, an antenna adjustment parameter is generated by adopting a cell comprehensive model.

The output optimization scheme adjustment parameters such as azimuth angle weight, horizontal wave width weight, downtilt angle weight and vertical wave width weight are automatically modified by the network management system of the butt joint 5G equipment manufacturer, and the traffic flow and service change of the network are self-adapted.

Automatic adjustment type of optimization scheme: disposable, periodic. And the periodicity is selected to be repeatedly executed in association with a scheduling period, and the periodicity is associated with the cell scheduling type in scheduling management. Performing task management on cells to be adjusted according to groups and standards of various provinces

The invention firstly collects the measurement report data MR of the 5G user, and performs geographic position location to realize the association between the user and the geographic position. And then, according to the cell industrial parameters and the positioned data, the regional coverage, the user density and the traffic of the 5G network can be evaluated, and based on the regional coverage, the user density and the traffic, an integrated service model of the cell is established by combining antenna parameters, an artificial intelligent means is utilized, intelligent optimization is realized, an optimization scheme of regional network structure parameters is formed, and real-time adjustment can be carried out according to the traffic, the traffic and the like.

Referring to fig. 2, step S2 is described in detail, and the method for establishing a cell coverage model according to the cell parameter and the geographic grid specifically includes:

step S21: and acquiring a center point of the cell corresponding to the geographic grid. See fig. 2-1

In fig. 2-1, a is a cell antenna, B is a projection of the antenna on the ground, C is an intersection point of an antenna signal and the ground (BC is an antenna coverage radius), D is a physical grid center point (grid center point is determined by grid longitude and latitude coordinate values), and E is a perpendicular to the plane ABC.

x ₁ For cell antenna longitude, y ₁ For cell antenna latitude, z ₁ For the cell antenna height, x ₂ Longitude, y, which is the intersection of the antenna signal and ground ₂ Is the latitude, x of the intersection point of the antenna signal and the ground ₃ Is the longitude of the center point of the physical grid, y ₃ Is the physical grid center point latitude.

Step S22: and establishing a corresponding relation between the antenna parameters in the cell industrial parameters and the geographic grid center point.

The azimuth angle of the antenna is psi (the negative direction of the Y axis is positive north), the downward inclination angle of the antenna (mechanical downward inclination angle+electronic downward inclination angle) is phi BAC, the relative included angle of the grid center point to the antenna is theta (& DBC), and the relative downward inclination angle of the grid center point to the antenna is

In right triangle Δdeb, cos θ=be/BD, be=bd×cos θ

In right triangle Δabe, tan +.bae=be/AB, then +.bae=arctan (BE/AB)

Thus, +_bae=arctan (BD x cos θ/AB), where (BD and AB are both known or can be) the relative downtilt of the grid center point to the antenna

In right triangle Δdeb, the relative angle of the grid center point to the antenna

θ＝270-ψ-arctan((y ₃ -y ₁ )/(x ₁ -x ₃ )) ②

The two formulas (1) and (2) are combined to establish the relative azimuth angle theta and the relative downtilt angle of the antenna between the azimuth angle psi and the downtilt angle BAC of the antenna and the center point of the physical grid to the antenna

Corresponding relation of (3).

Step S23: and constructing a comprehensive cost function of the three dimensions of cell coverage, repeated coverage and service absorption based on the corresponding relation.

For performance evaluation of a cell, the performance evaluation can be performed from three dimensions of coverage, quality and capacity, and meanwhile, to reduce the complexity of the function, a unitary comprehensive cost function taking the reference signal level RSRP as an independent variable is constructed as follows:

Q(RSRP)＝αc1+βc2+γc3 ④

wherein, alpha (alpha > 0), beta (beta > 0) and gamma (gamma > 0) are the ratio coefficients of coverage rate, non-overlapping coverage rate and service absorption rate, and the weight of coverage rate, non-overlapping coverage rate and service absorption capacity in the comprehensive cost function is changed by adjusting alpha, beta and gamma.

Coverage rate c1

The coverage c1 is defined as: c1 =n1/N

Wherein n1 is the effective coverage point number, that is, in the physical grid, if the reference signal level RSRP of the MR sampling point of the primary cell is greater than the threshold K1, the sampling point is considered to be an effective MR sampling point; n is the total number of MR sampling points counted in single physical grid

From the above definition, the value range of c1 is (0.ltoreq.c1.ltoreq.1), and the value is related to RSRP.

Service absorptivity c3

Traffic absorption by a cell is related to the level RSRP of the cell at the physical location (to simplify analysis, factors of poor quality and load sharing are temporarily not considered)

Thus, the traffic absorbance c3 is defined as:

wherein N is the total number of MR sampling points counted in a single physical grid; q _i (i=1, 2 … …..n) is the number of MR samples absorbed by a cell within a single physical grid. Sample point absorption is divided into the following three cases, as shown in fig. 2-2:

it is assumed that there are 1 MR sampling points in a physical grid, the level of the primary serving cell CellA is-68 dbm, the level of the secondary strong same-frequency neighbor cell B is-75 dbm, and the level of the inter-frequency neighbor cell C is-70 dbm.

1) After antenna adjustment, the coverage level of the CellA in a certain physical grid is increased from-68 dbm to-66 dbm, and the sampling point is still marked as A cell absorption if CellA (-66 dbm) > CellB (-75 dbm);

2) Through antenna adjustment, the coverage level of the cell B of the same frequency cell on a certain physical grid is increased from-75 dbm to-65 dbm,

CellA (-68 dbm) < CellB (-65 dbm) +common frequency threshold K1 (2 db), the sampling point is updated and marked as B cell absorption;

3) Through antenna adjustment, the coverage level of the inter-frequency cell Cell C in a certain physical grid is increased from-70 dbm to-64 dbm,

CellA (-68 dbm) < CellC (-64 dbm) +different frequency threshold K2 (3 db), the sampling point is updated and marked as C cell absorption;

note that: the same (different) frequency threshold value K1, K2 needs to be set according to the actual reselection and switching threshold parameters.

From the above definition, the value range of c3 is (0.ltoreq.c2.ltoreq.1), and the value is related to RSRP.

In summary, for all physical grids within the coverage area of a single cell, when the comprehensive cost function Q takes the maximum value, the corresponding RSRP value is the optimal solution.

Referring to fig. 3, a step S3 is described in detail, where the method for obtaining the regional user number and the traffic in the geographic grid according to the measurement report and establishing the regional user traffic model specifically includes:

step S31: and acquiring the measurement report in the geographic grid in a time period.

Step S32: and determining the total number of users and the total number of traffic in the geographic grid according to the acquired measurement report.

Step S33: and correcting the regional user service model according to the historical measurement report and the current network measurement report.

The user service model is established mainly for conveniently acquiring the number of users and the total quantity of service in an area, and can distinguish specific service types, and the interference in a network can be reduced as far as possible according to the number of users and the total quantity of service as the basis for selecting antennas and adjusting antennas.

Referring to fig. 4, step S4 is described in detail, the method for establishing a cell comprehensive model according to the cell coverage model and the regional user service model in combination with a large-scale antenna parameter library, training the cell comprehensive model by using full data, and iterating parameters of the cell comprehensive model specifically includes:

step S41: and creating a sample set through the full data, and cleaning and standardizing the samples.

Full data acquisition, including MR data, antenna data, user and traffic, cell coverage data, etc.

A sample set is created through the full data, valid sample data is selected, and abnormal data (missing data, inaccurate data, out-of-range data and the like) is cleaned.

The interference samples are cleaned and standardized, and the LTE wireless network becomes a self-interference system. The more interference sources of the self-interference system, the larger the interference is, the worse the link quality is, and according to the antenna pattern, the antenna azimuth angle and the received power RSRP of the physical grid after the downtilt angle adjustment can be calculated, and the unified data format is the received power.

Identifying an abnormal sample, including (missing data, inaccurate data, out of range data, etc.), the criteria for the abnormal sample are: the coverage rate c1 is in the range of (0-1) and is related to RSRP; the value range of the non-overlapping coverage rate c2 is (0 is less than or equal to c2 is less than or equal to 1), and the value is related to RSRP; the value range of the service absorptivity c3 is (0 is less than or equal to c2 is less than or equal to 1), and the value is related to RSRP. If the values of the three indexes are not in the range, the values are regarded as abnormal standard data. The abnormal samples will be cleaned up.

The availability of sample data can be maintained to the maximum extent by cleaning and standardization processing of the samples, and more accurate model parameters can be obtained by training courses according to the processed samples.

Step S42: and carrying out sample distribution modeling on each index in the sample set in combination with the cell coverage model, the regional user service model and a large-scale antenna parameter library, and representing the relation among different antenna parameters, regional coverage and regional service volumes of the cell.

Firstly, establishing a coverage geometric model by adopting coordinate axis projection aiming at a single cell, and secondly, establishing a comprehensive cost function based on three dimensions of coverage, overlapping coverage and service absorption for evaluating a coverage optimization effect; thirdly, an antenna parameter optimizing method based on a genetic iterative algorithm is provided, and an optimal antenna parameter setting scheme is converted into an optimal solution problem for solving a comprehensive cost function.

Genetic algorithm is a randomized search method which is evolved by referring to evolution rules (survival of the right and superior and inferior genetic mechanisms) of the biology world. The method is mainly characterized by having inherent hidden parallelism and better global optimizing capability; by adopting the probabilistic optimizing method, the optimized searching space can be automatically acquired and guided, the searching direction can be adaptively adjusted, and the determined rule is not needed.

The intelligent optimizing algorithm of the 5G Massive MIMO antenna is used as the genetic algorithm, and the algorithm uses the crossing, variation and selection processes in the genetic algorithm to set a cost function according to the 5G network coverage optimizing target (coverage area and coverage requirement), and then uses the antenna weight, the downtilt angle, the azimuth angle, the antenna height and the like as limiting conditions. And combining regional people flow distribution and traffic, adjusting cell antenna parameters through a crossover and mutation process, evaluating by a cost function, screening an optimization scheme, repeating the loop and iterative calculation, and finally obtaining an optimal optimization scheme suitable for the current network people flow distribution and traffic, and timely outputting a more reasonable optimization scheme according to the changes of the people flow distribution and the traffic.

Step S43: and substituting the full data into the cell comprehensive model for training, and taking the corresponding parameter value as the optimal solution when the comprehensive cost function of all the geographic grids in the coverage area of the single cell is maximum.

Training phase: in the Massive MIMO tuning configuration, the input data of the tree root and the tree trunk of the decision tree in the training stage is a set of beam flow information, the tree root is the total antenna lobe width and the antenna coverage distance weight, and the tree trunk is the antenna lobe width and the antenna weight obtained according to the data change. The traffic information of 64 beams is acquired from the MR data in the Massive MIMO antenna of 64T64R, and the leaf node of the decision tree is a Pattern value corresponding to the leaf node. The final Massive MIMO Pattern optimized decision tree is obtained through training of a random forest algorithm based on massive historical data, the training of the random forest algorithm is mainly based on an antenna analysis model, and training is carried out according to the numerical variation among antenna lobe width, antenna weight, antenna coverage distance and the number of users.

Reasoning: after the Massive MIMO cell is started and operated, a Pattern value is set according to a default value (planning design) of a node in a decision tree and is put into use, the value is a default antenna weight coefficient, and the parameter combination is not necessarily optimal. After the system is put into use, parameters of the cell antenna are collected, mainly, flow information of each wave beam is collected and sent to a trained Massive MIMO Pattern tuning model deployed on an AI reasoning platform, and calculation is carried out by using a random forest model. Different Pattern values are obtained through a large number of random trees, wherein the highest probability is the initial value of the result Pattern, and the highest probability refers to the initial value of the Pattern with the most common coverage model and the most reasonable coverage distance under the condition that the antenna lobe is set.

In the above, it has been discussed that for all physical grids within the coverage area of a single cell, when the comprehensive cost function Q takes the maximum value, the corresponding RSRP value is the optimal solution, and then the optimal solution is found when the model is trained.

Coding, interleaving, mutation and decoding of input variables in the model:

encoding

Genetic algorithms cannot directly deal with parameters of the problem space and must convert them into chromosomes or individuals of genetic space that consist of genes in a certain structure. This conversion operation is called encoding.

The coding strategy needs to meet the following specifications:

completeness of

All points in the problem space (candidate solutions) can behave as points in genetic space (chromosomes)

Soundness of health

Chromosomes in genetic space can correspond to candidate solutions in all spaces.

Non-redundancy

Chromosome and candidate solution are in one-to-one correspondence.

Several common encoding techniques currently exist, such as binary encoding, floating point encoding, character encoding, programming encoding, and the like. Binary coding is the most common coding method in genetic algorithms, i.e. a common 0,1 string is generated from the binary character set {0,1} to represent a candidate solution of a problem. The method has the characteristics of simplicity, easiness, compliance with the minimum character set coding principle, convenience in analysis by using a mode theorem and the like. The algorithm adopts a binary coding scheme.

In combination with the foregoing, the point X constituting the solution space of the comprehensive evaluation function is composed of four variables of the antenna azimuth angle, the electronic downtilt angle, the mechanical downtilt angle, and the antenna transmission power.

The practical field optimization is combined, and the range and the step length of the antenna azimuth angle, the electronic declination angle, the mechanical declination angle and the power are set as follows:

1) The adjustment range of the azimuth angle of the antenna is 0-120 degrees, and the adjustment step length is 5 degrees;

2) The mechanical downtilt angle of the antenna is adjusted to be 0-12 degrees, and the step length is adjusted to be 1 degree;

3) The electronic downtilt angle of the antenna is adjusted to be 0-14 degrees, and the step length is adjusted to be 0.5 degrees;

4) The adjustment range 92-152 of the cell transmitting power, adjust step length 1;

the 2-ary codewords set to 5 bits/4 bits/5 bits/6 bits respectively encode each chromosome as follows:

chromosome of the human body	Azimuth angle of antenna	Mechanical downtilt angle	Electron downtilt angle	Cell transmit power
					Encoding	11111	1111	11111	111111
Number of coding bits	5	4	5	6

After chromosome coding, a total of 20 binary codewords are adopted to correspond to genes in genetic space, namely solution of problem space.

Selection operator

In the genetic and natural evolution of organisms, species with higher adaptation to living environment will have more chance to be inherited to the next generation; and species with a low degree of adaptation to the living environment will have relatively little chance to be transmitted to the next generation. Mimicking this process, genetic algorithms use selection operators to perform a superior and inferior elimination operation on individuals in a population: individuals with lower fitness have less probability of being inherited into the next generation population; individuals with high fitness have a greater probability of being inherited to the next generation.

The most common selection operator is the scale selection operator. The scaling operator is a method of playback random sampling. The basic idea is as follows: the probability that each individual is selected is proportional to its fitness size. Let the population size be M. The fitness of the individual i is F, and the probability that the individual i is selected is:

crossover

The crossover operation in the genetic algorithm means that genes are exchanged in some way for two paired chromosomes to form a new individual. Cross-over operation is an important feature of genetic algorithms, which plays a key role in genetic algorithms, and is the primary method of generating new individuals.

The most common crossover operator is the single point crossover algorithm, also known as simple crossover, which refers to randomly placing a crossover point in an individual code string and then exchanging part of the chromosomes of two paired individuals at that point with each other. The algorithm adopts a single-point crossover algorithm.

Variation of

Mutation is an operation method in which a gene value at a certain locus is changed with a small probability, and it is also an operation method in which a new individual is generated. The algorithm adopts a basic mutation method to carry out mutation operation, and the specific operation process comprises the steps of firstly determining the gene mutation position of each individual, and then inverting the original genes of the mutation points according to a certain probability.

Decoding

And restoring the new code word into an antenna azimuth angle, an electronic downtilt angle, a mechanical downtilt angle and power, and respectively correcting the obtained mutated antenna azimuth angle, electronic downtilt angle, mechanical downtilt angle and power again according to the adjustment range and the adjustment step length set by the cell or all antennas so as to meet the adjustment requirement.

Step S44: and iterating the parameters of the cell comprehensive model according to the parameter values corresponding to the optimal solution.

Execution/iterative optimization phase: mapping the Pattern value obtained by reasoning to corresponding horizontal wave width, vertical wave width, downward inclination angle and horizontal angle, mapping according to a random forest model library according to the beam flow characteristic value of the antenna to obtain the Pattern value with the highest frequency, and adjusting the antenna parameters according to the parameters. After the reasoning result is executed, evaluating the network area coverage and the people flow improvement condition, and if the result is poor, returning to the previous parameter configuration; if optimization is brought, the configuration is kept, and repeated iteration and further optimization are continued until the network keeps a stable optimized result. The whole process is automatically and efficiently completed.

The invention also discloses an adjusting device of the antenna in the 5G network, an embodiment five is given below, and the structure of the device is described in detail with reference to the accompanying figure 5, wherein the device comprises a data acquisition module 1, a positioning geographic grid module 2, a cell coverage model module 3, a user service model module 4, a cell comprehensive model module 5 and a parameter adjusting module 6, wherein:

The data acquisition module 1 is used for acquiring a 5G measurement report, an electronic map and a cell industrial parameter;

the positioning geographic grid module 2 is used for positioning the measurement report to the geographic grid according to the 5G measurement report acquired by the data acquisition module and combining an electronic map and a cell industrial parameter;

a cell coverage model module 3, configured to establish a cell coverage model according to the cell parameter and the geographic grid located by the location geographic grid module;

the user service model module 4 is used for acquiring the regional user number and the service volume in the geographic grid according to the measurement report and establishing a regional user service model;

the cell comprehensive model module 5 is used for building a cell comprehensive model according to the cell coverage model built by the cell coverage model module and the regional user service model built by the user service model module and combining a large-scale antenna parameter library, training the cell comprehensive model by using full data, and iterating parameters of the cell comprehensive model;

and the parameter tuning unit 6 is used for generating antenna tuning parameters of the cell according to the cell comprehensive model of the cell comprehensive model module by taking the number of cell users and the coverage area as targets.

In order to better explain the structure and the working principle of each unit, the following provides a sixth embodiment of the present application, and the detailed description is given with reference to fig. 6:

the data acquisition module 1 is used for acquiring a 5G measurement report, an electronic map and a cell industrial parameter.

The cell coverage model module 2 further includes:

a geographical grid center point acquisition module 21, configured to acquire a center point of the geographical grid corresponding to a cell;

a correspondence establishing module 22, configured to establish a correspondence between antenna parameters in the cell parameter and the geographic grid center point;

and the comprehensive cost function construction module 23 is configured to construct a comprehensive cost function of three dimensions of cell coverage, repeated coverage and service absorption based on the correspondence.

The user traffic model module 3 further comprises:

a raster data acquisition module 31, configured to acquire the measurement report in the geographic grid located by the location geographic grid module in a time period;

a user traffic determining module 32, configured to determine the total number of users and the total number of traffic in the geographic grid according to the measurement report in the geographic grid acquired by the raster data acquiring module;

and the model correction module 33 is configured to correct the regional user service model according to the historical measurement report and the current network measurement report.

The cell synthesis model module 4 further comprises:

a sample collection module 41, configured to create a sample collection from the full data, and clean and normalize the sample;

a sample distribution modeling module 42, configured to perform sample distribution modeling on each index of the sample set created by the sample set module in combination with the cell coverage model, the regional user service model, and a large-scale antenna parameter library, and characterize a relationship between different antenna parameters, regional coverage, and regional traffic of the cell;

a model training module 43 for training a cell comprehensive model of the sample distribution modeling module using full data substitution;

a parameter iteration module 44; and the method is used for obtaining the parameter value corresponding to the maximum value of the comprehensive cost function of all the geographic grids in the coverage area of the single cell as an optimal solution, and iterating the parameters of the cell comprehensive model according to the parameter value corresponding to the optimal solution.

And the parameter tuning unit 5 is used for generating antenna tuning parameters of the cell according to the cell comprehensive model of the cell comprehensive model module by taking the number of cell users and the coverage area as targets.

It will be clearly understood by those skilled in the art that, for convenience and brevity, the corresponding process in the above-described apparatus embodiment may refer to the specific working process of the foregoing method, and will not be described in detail.

In the present specification, each embodiment is described in a progressive manner, and each embodiment is mainly described in a different point from other embodiments, and identical and similar parts between the embodiments are all enough to refer to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant points refer to the description of the method section.

The steps of a method or algorithm described in connection with the embodiments disclosed herein may be embodied directly in hardware, in a software unit executed by a processor, or in a combination of the two. The software elements may be disposed in Random Access Memory (RAM), memory, read Only Memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

The terms "first," "second," "third," "fourth" and the like in the description and in the claims and in the above drawings, if any, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments of the present application described herein may be capable of operation in sequences other than those illustrated herein.

The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for adjusting an antenna in a 5G network, the method comprising:

2. The method according to claim 1, wherein the method for establishing a cell coverage model according to cell engineering parameters and the geographic grid comprises the following steps:

acquiring a center point of a cell corresponding to the geographic grid;

3. The method according to claim 2, wherein the method for obtaining the regional user number and the traffic in the geographic grid according to the measurement report and establishing the regional user traffic model specifically comprises the following steps: acquiring the measurement report in the geographic grid in a time period;

Determining the total number of users and the total number of traffic in the geographic grid according to the acquired measurement report; and correcting the regional user service model according to the historical measurement report and the current network measurement report.

4. The method of claim 3, wherein the method for building a cell integrated model according to the cell coverage model and the regional user service model in combination with a large-scale antenna parameter library, training the cell integrated model by using full data, and iterating parameters of the cell integrated model specifically comprises:

5. The method according to claim 4, wherein the method for generating the antenna adjustment parameters of the cell according to the cell integrated model, targeting cell user number, traffic volume, coverage area, specifically comprises:

6. An apparatus for adjusting an antenna in a 5G network, the apparatus comprising:

7. The apparatus of claim 6, wherein the cell coverage model module further comprises:

8. The apparatus of claim 7, wherein the user traffic model module further comprises:

9. The apparatus of claim 8, wherein the cell synthesis model module further comprises:

10. The apparatus of claim 9, wherein the parameter tuning unit is configured to target a number of cell users and a coverage area, and the method for generating the antenna tuning parameter of the cell according to the cell synthesis model of the cell synthesis model module specifically includes: