CN108989092A - A kind of wireless network predicting method, electronic equipment and storage medium - Google Patents
A kind of wireless network predicting method, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN108989092A CN108989092A CN201810671779.8A CN201810671779A CN108989092A CN 108989092 A CN108989092 A CN 108989092A CN 201810671779 A CN201810671779 A CN 201810671779A CN 108989092 A CN108989092 A CN 108989092A
- Authority
- CN
- China
- Prior art keywords
- value
- predicted
- independent variable
- cross
- unknown parameter
- 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
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/14—Network analysis or design
- H04L41/145—Network analysis or design involving simulating, designing, planning or modelling of a network
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/14—Network analysis or design
- H04L41/147—Network analysis or design for predicting network behaviour
Abstract
The invention discloses a kind of wireless network predicting methods, comprise the following steps that choosing several network communication target variables and corresponding target indicator variable establishes Gaussian process regression model, determine the covariance function of gaussian kernel function form, the reasonable value for seeking the parameters in covariance function, the first cross-section data vector of the first value to be predicted and required independent variable that set any point-in-time obtain corresponding joint distribution function so that the valuation of the first value to be predicted is calculated;The second value to be predicted of any time period and the second cross-section data vector of required independent variable are set, the valuation of the described second value to be predicted is calculated.The present invention uses Gaussian process homing method, law mining is carried out to the network direct broadcasting data of historical accumulation, construct data model, and situation of change of the network index of target within following certain time can be effectively predicted, more effective data reference is provided for the allotment of optimization wireless network resource and performance optimization.
Description
Technical field
The present invention relates to a kind of wireless network communication technique more particularly to a kind of wireless network predicting methods, electronic equipment
And storage medium.
Background technique
Currently, cordless communication network gradually realizes extensive universal, daily life with the fast development of communication information technology
User gradually increases wireless network Dependent Demand degree in work.How accurately and efficiently to predict the user of different zones to wireless
Network using vigorous degree, avoid the occurrence of the situations such as network blockage, effectively improve wireless network resource allocative efficiency, be communication
One of the key content of the daily O&M optimization of operator, to further increase the satisfaction that user uses wireless network.Currently
To wireless network usually fed back afterwards by user using vigorous degree analyzing or the analyses such as network index simple statistics are found, and
And existing wireless network requirement analysis method is based primarily upon the simple one-dimensional linear analysis between index, can not effectively integrate
The data of multidimensional predict the wireless network demand fed back in following a period of time in advance
But existing being had the following deficiencies: using vigorous degree analyzing to wireless network
(1) the methods of common linear prediction method, neural network practicability are poor, and prediction result is unstable under small sample
It is fixed, it can not be predicted compatible with different network demand indexs, such as telephone traffic, GPRS service condition;
(2) the wireless network changes in demand rule in the following a period of time of certain region can not be effectively predicted;
(3) prediction that multiple-factor carries out wireless network index of correlation can not be integrated.
Summary of the invention
For overcome the deficiencies in the prior art, one of the objects of the present invention is to provide a kind of wireless network predicting method,
It more effectively predicts the wireless network demand in the following certain time period according to historical data.
The second object of the present invention is to provide a kind of electronic equipment, more effectively be predicted according to historical data following a certain
Wireless network demand in period.
The third object of the present invention is to provide a kind of computer storage medium, more effectively be predicted not according to historical data
Carry out the wireless network demand in certain time period.
An object of the present invention adopts the following technical scheme that realization:
A kind of wireless network predicting method, includes the following steps:
Data decimation step: several network communication target variables are chosen as independent variable, and choose the independent variable
Corresponding target indicator variable establishes the Gaussian process regression model of independent variable according to the independent variable as dependent variable;
Function determines step: the covariance function of gaussian kernel function form is determined according to the Gaussian process regression model,
There is the first unknown parameter, the second unknown parameter and third unknown parameter in the covariance function;
Parameter determination: the reasonable value of the first unknown parameter, the second unknown parameter and third unknown parameter is sought;
Time point model prediction step: it sets needed for the first value to be predicted and first value to be predicted of any point-in-time
First cross-section data vector of independent variable, according to the independent variable, dependent variable, first it is to be predicted value and the first cross-section data to
Amount obtains corresponding joint distribution function so that the valuation of the first value to be predicted is calculated;
Period model prediction step: it sets needed for the second value to be predicted and second value to be predicted of any time period
Second cross-section data vector of independent variable is calculated described according to joint distribution function described in the second cross-section data vector sum
The valuation of two values to be predicted.
Further, further include following steps:
Error analysis step: error point is carried out to the valuation of the second predicted value using standard deviation, relative error as measurement index
Analysis, within a preset range whether judgment criteria difference and relative error.
Further, in data decimation step, the argument list is shown as: Xn×s={ x1,x2,...xs, dependent variable indicates
For Y={ y1,y2,...yn, wherein the vector X in independent variablei={ xi1,xi2,...xis, i=1,2 ..., m are indicated from change
The cross-section data collection of amount, the n in dependent variable is data set sequence length, i.e. data capacity, the Gaussian process regression model isWhereinFor a gaussian random noise process,
Z=(2 π)n/2|K|1/2, μ is the mean vector of independent variable;K is covariance matrix, and
Further, the covariance function are as follows:Its
Middle xi,xjRespectively i-th of cross-section data vector sum independent variable i-th of cross-section data vector of index of independent variable index, s are input
The number of arguments, σfFor the first unknown parameter, l is the second unknown parameter, σnFor third unknown parameter.
Further, in parameter determination, selection meets maximum θ value under the conditions of P (θ | X, Y), passes through conjugation ladder
Degree method maximizes likelihood to P (θ | X, Y) to obtain the reasonable value of θ, wherein θ={ l, σf,σn}。
Further, in time point model prediction step, the first value to be predicted of setting isThe first value institute to be predicted
The the first cross-section data vector for needing independent variable isThen the joint distribution function isI.e.According to formula
The valuation that middle mean of probability distribution obtains the first value to be predicted isWherein,
Further, in period model prediction step, the second value to be predicted of setting is Second cross-section data vector isBy X*Gradually inputThe valuation of the second value to be predicted is calculated.
The second object of the present invention is implemented with the following technical solutions:
A kind of electronic equipment can be run on a memory and on a processor including memory, processor and storage
Computer program, the processor perform the steps of when executing the computer program
Data decimation step: several network communication target variables are chosen as independent variable, and choose the independent variable
Corresponding target indicator variable establishes the Gaussian process regression model of independent variable according to the independent variable as dependent variable;
Function determines step: the covariance function of gaussian kernel function form is determined according to the Gaussian process regression model,
There is the first unknown parameter, the second unknown parameter and third unknown parameter in the covariance function;
Parameter determination: the reasonable value of the first unknown parameter, the second unknown parameter and third unknown parameter is sought;
Time point model prediction step: it sets needed for the first value to be predicted and first value to be predicted of any point-in-time
First cross-section data vector of independent variable, according to the independent variable, dependent variable, first it is to be predicted value and the first cross-section data to
Amount obtains corresponding joint distribution function so that the valuation of the first value to be predicted is calculated;
Period model prediction step: it sets needed for the second value to be predicted and second value to be predicted of any time period
Second cross-section data vector of independent variable is calculated described according to joint distribution function described in the second cross-section data vector sum
The valuation of two values to be predicted.
It preferably, further include following steps after having executed period model prediction step: error analysis step: with standard
Whether the valuation progress error analysis of difference, relative error for measurement index to the second predicted value, judgment criteria difference and relative error
Within a preset range.
The third object of the present invention is implemented with the following technical solutions:
A kind of computer readable storage medium, is stored thereon with computer program, and the computer program is held by processor
The wireless network predicting method as described in one of the object of the invention is realized when row.
Compared with prior art, the beneficial effects of the present invention are:
The present invention uses Gaussian process homing method, carries out law mining, building to the network direct broadcasting data of historical accumulation
Data model, and situation of change of the network index of target within following certain time can be effectively predicted, to optimize wireless network
Network resource allocation and performance optimization provide more effective data reference.
Detailed description of the invention
Fig. 1 is a kind of flow chart of wireless network predicting method of the invention.
Specific embodiment
In the following, being described further in conjunction with attached drawing and specific embodiment to the present invention, it should be noted that not
Under the premise of conflicting, new implementation can be formed between various embodiments described below or between each technical characteristic in any combination
Example.
As shown in Figure 1, the present invention provides a kind of wireless network predicting method, it is based on Gaussian process regression model, specifically
Include the following steps:
S1: it is corresponding as independent variable, and the selection independent variable that several network communication target variables are chosen
Target indicator variable establishes the Gaussian process regression model of independent variable according to the independent variable as dependent variable;
In this step, s network communication target variable is chosen as prediction model impact factor (independent variable) and is used for model
Training, corresponding target indicator variable Y (dependent variable), model argument data set representations are Xn×s={ x1,x2,...xs,
Wherein vector Xi={ xi1,xi2,...xis, i=1,2 ..., m indicate the cross-section data collection of independent variable, model dependent variable vector
It is expressed as Y={ y1,y2,...yn, wherein n is data set sequence length, i.e. data capacity.
Gaussian process regression model in this method is assumed to beWhereinFor a height
This random noise processes, f (χ) are stochastic variable joint probability distribution density function f (χ)~GP (0, K) of Gaussian process, i.e.,Z=(2 π)n/2|K|1/2, μ is the mean vector of independent variable;K is covariance matrix, and
And
As unit of certain city, the time series historical data of several target variables of urban base acquisition is chosen,
Such as need predicting telephone traffic in one week future situation of change in certain region, then respectively using the whole network, network element or cell as region granularity,
It selects the time for the TCH/SDCCH congestion ratio of 5 minutes granularities, TCH/SDCCH percent of call completed, TCH/SDCCH cutting off rate, cut out into
The independents variable index such as power and switch success ratio is as training data.
For the influence for avoiding the independent variable of mode input or the index dimension of dependent variable, therefore need to be to mode input data set
It is normalized, formula are as follows:In formula: min (xi) and max (xi) be respectively sequence minimum
Value and maximum value, xiFor original series value.
S2: the covariance function of gaussian kernel function form, the covariance are determined according to the Gaussian process regression model
There is the first unknown parameter, the second unknown parameter and third unknown parameter in function;
This step can obtain covariance function according to covariance matrix, and covariance function uses the form of gaussian kernel function
It indicates, isWherein xi,xjRespectively i-th of independent variable index
Cross-section data vector sum independent variable i-th of cross-section data vector of index, s are the number of arguments of input, σfFor the first unknown ginseng
Number, l are the second unknown parameter, σnFor third unknown parameter.
S3: the reasonable value of the first unknown parameter, the second unknown parameter and third unknown parameter is sought;
It is last due to assuming that normal distribution mean value is zero using Gaussian process in above process in this step
Model result and the selection of covariance have much relations, the parameter that above-mentioned steps are related to has θ={ l, σf,σn, it is usually used
Method for parameter estimation is MAP estimation, and selection meets maximum θ value under the conditions of P (θ | X, Y), but due to not any to θ
Priori knowledge, then P (θ | X, Y) degenerates for maximum likelihood P (Y | X, θ) at this time, and following formula is set up:It is above-mentioned seemingly using conjugate gradient method maximization
It so, can be in the hope of the corresponding reasonable value of θ.
Conjugate gradient method calculating process is as follows:
A initializes iterative step: t=0, gives initial value: θ(0), give positive definite matrix Q;B, the f (θ) of calculation formula 3
Derivative, enableIf g(0)=0, meet function maxima condition at this time, has then stopped iterating to calculate.It is no
Then, the optimal value direction of search uses steepest descent method, i.e., function f (θ) is in θ(0)Gradient negative direction, i.e. d(0)=-g(0);C, meter
Calculate iteration step length:It calculates and generates next iteration point: θ(t+1)=θ(t)+αtd(t);E is calculatedIf g(t+1)=0, then stop iterating to calculate, in t+1 iteration, calculateSo that d(t+1)With d(0),d(1)...d(t)Form Q conjugate direction;F calculates d then in t+1 iteration(t +1)=-g(t+1)+βtd(t);G enables t=t+1, returns to step c), is iteratively repeated calculating;H, until reach condition of convergence threshold value (ε=
1e-6) or g(t+1)=0, obtained θ(t+1)Value is current reasonable estimated value.
S4: the first section of independent variable needed for setting the first value to be predicted and first value to be predicted of any point-in-time
Data vector obtains corresponding joint according to the independent variable, dependent variable, the first value to be predicted and the first cross-section data vector
Distribution function is to be calculated the valuation of the first value to be predicted;
Setting the first value to be predicted isFirst cross-section data vector of independent variable needed for first value to be predicted isDefine several mentioned-above network communication target variables (independent variable), its corresponding target
Target variable (dependent variable) is training data, and the first value reality to be predicted in this step is also dependent variable, the first value to be predicted
It is prediction data with the first cross-section data vector, according to the property of Gaussian process, training data is with prediction data from same
The joint distribution function of the characteristics of data sample, available training data and prediction data is the Gaussian Profile of higher-dimension, then institute
Stating joint distribution function isI.e.According to public affairs
FormulaThe valuation that middle mean of probability distribution obtains the first value to be predicted isWherein,
S5: the second section of independent variable needed for setting the second value to be predicted and second value to be predicted of any time period
Estimating for the described second value to be predicted is calculated according to joint distribution function described in the second cross-section data vector sum in data vector
Value.
Setting the second value to be predicted isSecond cross-section data vector isBy X*Gradually inputThe valuation of the second value to be predicted is calculated.
S6: error analysis, judgment criteria are carried out to the valuation of the second predicted value using standard deviation, relative error as measurement index
Within a preset range whether difference and relative error.
For model prediction as a result, two standard deviation, relative error measurement indexs are respectively adopted to be predicted
The valuation of value and actual value carry out error analysis, standard deviationRelative errorThe smaller expression prediction result of the value of RMSE, MAPE is more ideal.
The present invention also provides a kind of electronic equipment, including memory, processor and storage on a memory and can located
The computer program run on reason device, processor perform the steps of when executing the computer program
Data decimation step: several network communication target variables are chosen as independent variable, and choose the independent variable
Corresponding target indicator variable establishes the Gaussian process regression model of independent variable according to the independent variable as dependent variable;
Function determines step: the covariance function of gaussian kernel function form is determined according to the Gaussian process regression model,
There is the first unknown parameter, the second unknown parameter and third unknown parameter in the covariance function;
Parameter determination: the reasonable value of the first unknown parameter, the second unknown parameter and third unknown parameter is sought;
Time point model prediction step: it sets needed for the first value to be predicted and first value to be predicted of any point-in-time
First cross-section data vector of independent variable, according to the independent variable, dependent variable, first it is to be predicted value and the first cross-section data to
Amount obtains corresponding joint distribution function so that the valuation of the first value to be predicted is calculated;
Period model prediction step: it sets needed for the second value to be predicted and second value to be predicted of any time period
Second cross-section data vector of independent variable is calculated described according to joint distribution function described in the second cross-section data vector sum
The valuation of two values to be predicted;
Error analysis step: error point is carried out to the valuation of the second predicted value using standard deviation, relative error as measurement index
Analysis, within a preset range whether judgment criteria difference and relative error.
The present invention can also provide a kind of computer readable storage medium, be stored thereon with computer program, computer journey
A kind of wireless network predicting method of the invention is realized when sequence is executed by processor.
The model framework of this method building, in actual mechanical process, network optimization engineer can define according to forecast demand
Different zones granularity, the model data index of different time span progress model training and forecast analysis are inputted, model will be to future
Being effectively predicted for certain time, facilitates the planning in advance during network O&M, largely to formulate effective resource
Allotment strategy.
The above embodiment is only the preferred embodiment of the present invention, and the scope of protection of the present invention is not limited thereto,
The variation and replacement for any unsubstantiality that those skilled in the art is done on the basis of the present invention belong to institute of the present invention
Claimed range.
Claims (10)
1. a kind of wireless network predicting method, which comprises the steps of:
Data decimation step: choosing several network communication target variables as independent variable, and chooses the independent variable difference
Corresponding target indicator variable establishes the Gaussian process regression model of independent variable according to the independent variable as dependent variable;
Function determines step: the covariance function of gaussian kernel function form is determined according to the Gaussian process regression model, it is described
There is the first unknown parameter, the second unknown parameter and third unknown parameter in covariance function;
Parameter determination: the reasonable value of the first unknown parameter, the second unknown parameter and third unknown parameter is sought;
Time point model prediction step: it sets needed for the first value to be predicted and first value to be predicted of any point-in-time from change
First cross-section data vector of amount, is obtained according to the independent variable, dependent variable, the first value to be predicted and the first cross-section data vector
Corresponding joint distribution function is taken so that the valuation of the first value to be predicted is calculated;
Period model prediction step: it sets needed for the second value to be predicted and second value to be predicted of any time period from change
Amount the second cross-section data vector, according to joint distribution function described in the second cross-section data vector sum be calculated described second to
The valuation of predicted value.
2. wireless network predicting method as described in claim 1, which is characterized in that further include following steps:
Error analysis step: error analysis is carried out to the valuation of the second predicted value using standard deviation, relative error as measurement index, is sentenced
Within a preset range whether disconnected standard deviation and relative error.
3. wireless network predicting method as described in claim 1, which is characterized in that in data decimation step, the independent variable
It indicates are as follows: Xn×s={ x1,x2,...xs, dependent variable is expressed as Y={ y1,y2,...yn, wherein the vector X in independent variablei=
{xi1,xi2,...xis, i=1,2 ..., m indicate the cross-section data collection of independent variable, and the n in dependent variable is that data set sequence is long
Degree, i.e., data capacity, the Gaussian process regression model areWhereinFor a Gauss with
Machine noise process,Z=(2 π)n/2|K|1/2, μ is the mean vector of independent variable;K is association
Variance matrix, and
4. wireless network predicting method as claimed in claim 3, which is characterized in that the covariance function are as follows:Wherein xi,xjRespectively i-th of number of cross-sections of independent variable index
According to i-th of cross-section data vector of vector sum independent variable index, s is the number of arguments of input, σfFor the first unknown parameter, l is
Second unknown parameter, σnFor third unknown parameter.
5. wireless network predicting method as claimed in claim 4, which is characterized in that in parameter determination, selection meets P
Maximum θ value under the conditions of (θ | X, Y) maximizes likelihood to P (θ | X, Y) by conjugate gradient method to obtain the reasonable value of θ,
In, θ={ l, σf,σn}。
6. wireless network predicting method as claimed in claim 5, which is characterized in that in time point model prediction step, setting first
Value to be predicted isFirst cross-section data vector of independent variable needed for first value to be predicted is
Then the joint distribution function isI.e.According to public affairs
FormulaThe valuation that middle mean of probability distribution obtains the first value to be predicted isWherein,
7. wireless network predicting method as claimed in claim 6, which is characterized in that in period model prediction step, setting the
Two values to be predicted areSecond cross-section data vector is
By X*Gradually inputThe valuation of the second value to be predicted is calculated.
8. a kind of electronic equipment including memory, processor and stores the meter that can be run on a memory and on a processor
Calculation machine program, which is characterized in that the processor performs the steps of when executing the computer program
Data decimation step: choosing several network communication target variables as independent variable, and chooses the independent variable difference
Corresponding target indicator variable establishes the Gaussian process regression model of independent variable according to the independent variable as dependent variable;
Function determines step: the covariance function of gaussian kernel function form is determined according to the Gaussian process regression model, it is described
There is the first unknown parameter, the second unknown parameter and third unknown parameter in covariance function;
Parameter determination: the reasonable value of the first unknown parameter, the second unknown parameter and third unknown parameter is sought;
Time point model prediction step: it sets needed for the first value to be predicted and first value to be predicted of any point-in-time from change
First cross-section data vector of amount, is obtained according to the independent variable, dependent variable, the first value to be predicted and the first cross-section data vector
Corresponding joint distribution function is taken so that the valuation of the first value to be predicted is calculated;
Period model prediction step: it sets needed for the second value to be predicted and second value to be predicted of any time period from change
Amount the second cross-section data vector, according to joint distribution function described in the second cross-section data vector sum be calculated described second to
The valuation of predicted value.
9. electronic equipment as claimed in claim 8, which is characterized in that after having executed period model prediction step, further include
Following steps: error point error analysis step: is carried out to the valuation of the second predicted value using standard deviation, relative error as measurement index
Analysis, within a preset range whether judgment criteria difference and relative error.
10. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the computer program
Such as claim 1-7 described in any item wireless network predicting methods are realized when being executed by processor.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810671779.8A CN108989092B (en) | 2018-06-26 | 2018-06-26 | Wireless network prediction method, electronic equipment and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810671779.8A CN108989092B (en) | 2018-06-26 | 2018-06-26 | Wireless network prediction method, electronic equipment and storage medium |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108989092A true CN108989092A (en) | 2018-12-11 |
CN108989092B CN108989092B (en) | 2023-01-17 |
Family
ID=64538916
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810671779.8A Active CN108989092B (en) | 2018-06-26 | 2018-06-26 | Wireless network prediction method, electronic equipment and storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108989092B (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109996247A (en) * | 2019-03-27 | 2019-07-09 | 中国电子科技集团公司信息科学研究院 | Networked resource concocting method, device, equipment and storage medium |
CN110139288A (en) * | 2019-04-08 | 2019-08-16 | 深圳众赢维融科技有限公司 | A kind of Internet phone-calling method, apparatus, system and recording medium |
CN111126624A (en) * | 2019-12-20 | 2020-05-08 | 北京首汽智行科技有限公司 | Method for judging validity of model prediction result |
CN111385128A (en) * | 2018-12-29 | 2020-07-07 | 中兴通讯股份有限公司 | Method and device for predicting burst load, storage medium, and electronic device |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102542126A (en) * | 2011-10-10 | 2012-07-04 | 上海交通大学 | Soft measurement method based on half supervision learning |
CN102621953A (en) * | 2012-03-20 | 2012-08-01 | 天津大学 | Automatic online quality monitoring and prediction model updating method for rubber hardness |
US20160299938A1 (en) * | 2015-04-10 | 2016-10-13 | Tata Consultancy Services Limited | Anomaly detection system and method |
-
2018
- 2018-06-26 CN CN201810671779.8A patent/CN108989092B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102542126A (en) * | 2011-10-10 | 2012-07-04 | 上海交通大学 | Soft measurement method based on half supervision learning |
CN102621953A (en) * | 2012-03-20 | 2012-08-01 | 天津大学 | Automatic online quality monitoring and prediction model updating method for rubber hardness |
US20160299938A1 (en) * | 2015-04-10 | 2016-10-13 | Tata Consultancy Services Limited | Anomaly detection system and method |
Non-Patent Citations (1)
Title |
---|
王建民等: "基于高斯过程回归的变形智能预测模型及应用", 《武汉大学学报(信息科学版)》 * |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111385128A (en) * | 2018-12-29 | 2020-07-07 | 中兴通讯股份有限公司 | Method and device for predicting burst load, storage medium, and electronic device |
CN109996247A (en) * | 2019-03-27 | 2019-07-09 | 中国电子科技集团公司信息科学研究院 | Networked resource concocting method, device, equipment and storage medium |
CN109996247B (en) * | 2019-03-27 | 2022-03-11 | 中国电子科技集团公司信息科学研究院 | Networked resource allocation method, device, equipment and storage medium |
CN110139288A (en) * | 2019-04-08 | 2019-08-16 | 深圳众赢维融科技有限公司 | A kind of Internet phone-calling method, apparatus, system and recording medium |
CN110139288B (en) * | 2019-04-08 | 2022-07-01 | 简链科技(广东)有限公司 | Network communication method, device, system and recording medium |
CN111126624A (en) * | 2019-12-20 | 2020-05-08 | 北京首汽智行科技有限公司 | Method for judging validity of model prediction result |
Also Published As
Publication number | Publication date |
---|---|
CN108989092B (en) | 2023-01-17 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Nikravesh et al. | Mobile network traffic prediction using MLP, MLPWD, and SVM | |
CN108989092A (en) | A kind of wireless network predicting method, electronic equipment and storage medium | |
WO2020248228A1 (en) | Computing node load prediction method in a hadoop platform | |
CN112633316B (en) | Load prediction method and device based on boundary estimation theory | |
CN111294812B (en) | Resource capacity-expansion planning method and system | |
CN104899135A (en) | Software defect prediction method and system | |
CN107547154A (en) | A kind of method and device for establishing video traffic prediction model | |
CN107809766B (en) | Method and device for generating machine learning sample for network optimization | |
He et al. | Graph attention spatial-temporal network for deep learning based mobile traffic prediction | |
Luna et al. | Adaptive fuzzy system to forecast financial time series volatility | |
CN109034490A (en) | A kind of Methods of electric load forecasting, device, equipment and storage medium | |
CN115915364A (en) | Energy-saving method and device for communication base station, computer readable medium and equipment | |
CN106600042A (en) | Mobile phone user networking period prediction method | |
US20210099894A1 (en) | Forcasting time series data | |
CN107329887A (en) | A kind of data processing method and device based on commending system | |
CN113988558B (en) | Power grid dynamic security assessment method based on blind area identification and electric coordinate system expansion | |
He et al. | Network traffic prediction method based on multi-channel spatial-temporal graph convolutional networks | |
CN110659930A (en) | Consumption upgrading method and device based on user behaviors, storage medium and equipment | |
Sun et al. | Forecasting wireless demand with extreme values using feature embedding in gaussian processes | |
CN111800807A (en) | Method and device for alarming number of base station users | |
CN108808657A (en) | A kind of Short-Term Load Forecasting of Electric Power System | |
Ahani et al. | A feature weighting and selection method for improving the homogeneity of regions in regionalization of watersheds | |
Miao et al. | MSFS: multiple spatio-temporal scales traffic forecasting in mobile cellular network | |
CN111932302A (en) | Method, device, equipment and system for determining number of service sites in area | |
Kalra et al. | RETRACTED ARTICLE: Enduring data analytics for reliable data management in handling smart city services |
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 |