CN110213774A - A kind of 5G network automatic evaluation system neural network based, method and device - Google Patents
A kind of 5G network automatic evaluation system neural network based, method and device Download PDFInfo
- Publication number
- CN110213774A CN110213774A CN201910436086.5A CN201910436086A CN110213774A CN 110213774 A CN110213774 A CN 110213774A CN 201910436086 A CN201910436086 A CN 201910436086A CN 110213774 A CN110213774 A CN 110213774A
- Authority
- CN
- China
- Prior art keywords
- data
- processing
- information
- network
- user
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2135—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on approximation criteria, e.g. principal component analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- 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
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L43/00—Arrangements for monitoring or testing data switching networks
- H04L43/08—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
- H04L43/0805—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters by checking availability
- H04L43/0811—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters by checking availability by checking connectivity
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L43/00—Arrangements for monitoring or testing data switching networks
- H04L43/08—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
- H04L43/0805—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters by checking availability
- H04L43/0817—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters by checking availability by checking functioning
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W16/00—Network planning, e.g. coverage or traffic planning tools; Network deployment, e.g. resource partitioning or cells structures
- H04W16/22—Traffic simulation tools or models
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W24/00—Supervisory, monitoring or testing arrangements
- H04W24/06—Testing, supervising or monitoring using simulated traffic
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Health & Medical Sciences (AREA)
- Software Systems (AREA)
- Environmental & Geological Engineering (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention belongs to 5G technical fields, and in particular to a kind of 5G network automatic evaluation system neural network based, method and device.System includes front end acquisition module, is sent for acquiring the data information of 5G front end applications equipment, and by collected data information；Data information includes: user's master data, scene classification data, communications device data, communication quality data and user's score data；Data loading module is used for the collected data information of receiving front-end acquisition module, and stores to the data information received；Data processing and automatic evaluation module carry out data modeling according to the result of data processing, generate the data model of 5G network automatic evaluation system for carrying out data processing according to the data information received；Using supporting module, for the data model according to the 5G network automatic evaluation system of generation, the data information of the collected front end applications equipment of front end acquisition module；Intelligence degree is high, and application field is extensive.
Description
Technical field
The invention belongs to 5G technical fields, and in particular to a kind of 5G network automatic evaluation system neural network based, side
Method and device.
Background technique
5th third-generation mobile communication technology (English: 5th generation mobile networks or 5th
Generation wireless systems, abbreviation 5G) it is latest generation cellular mobile communication technology, it is 4G (English: The
4th generation of mobile phone mobile communication technology standards, referred to as
4G), 3G (The 3rd generation mobile communication technology, abbreviation 3G) and 2G (The 2nd
Generation mobile phone communication technology specification, abbreviation 2G) after system
Extension.The performance objective of 5G is high data rate, reduces delay, save the energy, reduce cost, improving power system capacity and big rule
The connection of mould equipment.The first stage of 5G specification in Release-15 is to adapt to the business deployment of early stage.Release-16
Second stage will be completed the year two thousand twenty April, the candidate as IMT-2020 technology submits to International Telecommunication Union (ITU).
ITU IMT-2020 code requirement speed is up to 20Gbit/s, and wide channels bandwidth and large capacity MIMO may be implemented.
Third generation partner program (3GPP) will submit 5G NR (new radio) to be used as its 5G communication standard motion.5G NR may include
Low frequency (FR1) is lower than 6GHz and higher frequency (FR2), is higher than 24GHz and millimeter wave range.However, in early deployment,
Only slightly better than new 4G system using the speed of 5G NR software and delay on 4G hardware (dependent), estimation 15% will arrive
50%.The emulation of independent eMBB deployment shows that within the scope of FR1, handling capacity improves 2.5 times, improves within the scope of FR2 close
20 times.
It is extremely urgent for popularizing on a large scale for 5G network, although having been taken for the technology of construction and the operation of 5G network
Tremendous development is obtained, but after 5G network application, how to improve the experience of user, if the system huge in data volume and user volume
In, the defect of network is found in time, improves network structure and targetedly network is adjusted, still belong to the true of the field
It is empty.
Artificial neural network (English: Artificial Neural Network, ANN), abbreviation neural network (Neural
Network, NN) or neural network in machine learning and cognitive science field be a kind of mimic biology neural network (animal
Central nervous system, especially brain) structure and function mathematical model or computation model, for estimating function
Or it is approximate.Neural network is coupled by a large amount of artificial neuron to be calculated.In most cases artificial neural network can be outside
Change internal structure on the basis of boundary's information, be a kind of Adaptable System, popular saying is exactly to have learning functionality.Modern neuro
It is usually to pass through based on mathematical statistics type that network, which is a kind of Nonlinear Statistical data modeling artificial neural networks,
Learning method (Learning Method) is optimised, so and mathematical statistics method a kind of practical application, pass through statistics
We can obtain the partial structurtes space that can be largely expressed with function to standard mathematical techniques, on the other hand in people
The human perception field that work is intelligently learned, we can be come the decision problem for work perceptible aspect of conducting oneself by the application of mathematical statistics
(that is by statistical method, artificial neural network can similar people equally have simple deciding ability and simple
Judgement), this method is more advantageous compared with formal logistics reasoning calculation.
From the above, it can be seen that being combined by neural network and 5G service, to the data during 5G Web vector graphic
Information carries out comprehensive collect and analyzes with data, can analyze out the defect of network and proposes improved suggestion.Promote 5G network
Performance and promote the network optimization and technological progress.
Summary of the invention
In view of this, being assessed automatically the main purpose of the present invention is to provide a kind of 5G network neural network based and being
System, method and device establish the neural network mould of these data informations by the data information during the collection 5G network operation
Type, and then realize that analyzes the operating status of the 5G network automatically, and propose improved scheme and discovery defect, it can be applicable in
In all network models based on 5G, intelligence degree is high, and application field is extensive.
In order to achieve the above objectives, the technical scheme of the present invention is realized as follows:
A kind of 5G network automatic evaluation system neural network based, the system comprises:
Front end acquisition module, for acquiring the data information of 5G front end applications equipment, and by collected data information into
Row is sent；The data information includes: user's master data, scene classification data, communications device data, communication quality data and
User's score data；
Data loading module is used for the collected data information of receiving front-end acquisition module, and believes the data received
Breath is stored；
Data processing and automatic evaluation module, for carrying out data processing according to the data information received, according to data
The result of processing carries out data modeling, generates the data model of 5G network automatic evaluation system；
Using supporting module, for the data model and front-end collection according to the 5G network automatic evaluation system of generation
The data information of the collected front end applications equipment of module, the automatic network quality for assessing current network, carries out usertracking, property
It can optimize, the processing of customer service and effect of optimization analysis.
Further, the front end acquisition module includes: data acquisition unit, front-end local database, front end data biography
Defeated server and DEU data encryption unit；The data acquisition unit acquires user's master data, scene classification data, communication equipment
Data, communication quality data and user's score data, send collected data in front-end local database and store；
The front end data transmission server sends the data stored in front-end local database；The DEU data encryption unit exists
When front end data transmission server carries out data transmission, data are encrypted.
Further, user's master data includes at least: user's registration essential information, address area division information,
5G set meal type information and connection network number information；The row that the scene classification data are engaged in specific time by user
For data；The communications device data includes: base station number in 5 kilometers of circumference of region locating for the user, base station performance
With base station distance data；The communication quality data includes: switching rate, cutting off rate, drop rate, switch rate and the connection of 5G network
Rate；User's score data is the marking data that user carries out in specific time.
Further, the data loading module includes: data decryption unit, remote data transmission server, interface clothes
Business device and remote database；The data decryption unit solves the data information for the encryption that front end acquisition module sends over
It is close, the data information after decryption is then sent to remote data transmission server；The interface server and remote database
It is connected directly, the data information after decryption that remote data server receives is called, the data information hair called
Remote database is sent to be stored.
Further, the data processing and automation evaluation module include: data processing and modeling processor；The number
The data information of remote database is called directly according to processing and modeling processor, carries out data modeling, is generated 5G network and is commented automatically
Estimate the data model of system；The data processing and modeling processor include: data pre-processing unit, hough transformation unit, number
According to Standardisation Cell, algorithm predicting unit and modeling analysis unit；The data prediction is successively removed data information
Unique attribute, processing missing values and rejecting outliers processing；The hough transformation unit, for by the data after data prediction
Specification processing is carried out, so that treated that data are irrelevant two-by-two for specification, but is able to maintain original information；The data normalization
Unit is allowed to specification treated data bi-directional scaling to fall into a specific sections；The algorithm predicting unit, with you
Data after poly- data normalization cell processing carry out data modeling；The modeling analysis unit is used to generate by computation model
Customer satisfaction data and existing client satisfaction data carry out precision calculating.
A kind of 5G network automatic evaluation method neural network based, the method execute following steps:
Step S1: the data information of acquisition 5G front end applications equipment, and collected data information is sent；
Step S2: the collected data information of receiving front-end acquisition module, and the data information received is stored；
Step S3: carrying out data processing according to the data information received, carries out data according to the result of data processing and builds
Mould generates the data model of 5G network automatic evaluation system；
Step S4: according to the data model of the 5G network automatic evaluation system of generation, and: collected front end applications are set
Standby data information, the automatic network quality for assessing current network carry out usertracking, performance optimization, customer service and optimization
The processing of effect analysis.
Further, the step S3: data processing is carried out according to the data information received, according to the knot of data processing
Fruit carries out data modeling, and the method for generating the data model of 5G network automatic evaluation system executes following steps:
Step S3.1: data prediction is carried out, comprising: removal unique attribute, processing missing values and rejecting outliers and place
Reason；
Step S3.2: hough transformation processing is carried out, comprising: remove average value, calculate covariance matrix, calculate covariance matrix
Eigen vector, characteristic value is sorted from large to small, maximum k feature vector is retained, data are transformed into k
In the new space of feature vector building；Finally treated new data, it is irrelevant two-by-two between these data；
Step S3.3: it carries out data normalization processing and is allowed to data bi-directional scaling to fall into a specific sections；Its
In, using following transfer function, linear transformation is carried out to data, so that result is fallen on [0,1] section, transfer function is as follows:
Wherein, x*For data normalization treated result；X is data to be processed；Min is
Minimum value in data；Max is the maximum value in data；
Step S3.4: data modeling is carried out；
Step S3.5: effect analysis is carried out, comprising: after model training, using following formula, computation model is generated
Customer satisfaction data and existing client satisfaction data carry out precision calculate to get arrive R2Score, score is higher, indicates
Model precision is better；
The wherein customer satisfaction data (predicted value) that y representative model generates；
Represent existing client satisfaction data；
nsamplesRepresent the sample size size for entering model.
Further, the step S3.4: the method for carrying out data modeling executes following steps:
Step S3.4.1: it obtains and uses x as input variable for the data of modelingiIt indicates, wherein i represents the data
In i-th of variable；The xiIt includes at least: user's registration essential information, address area division information, 5G set meal type information
With connection network number information；The behavioral data that the scene classification data are engaged in specific time by user；It is described logical
Letter device data includes: base station number, base station performance and base station distance number in 5 kilometers of circumference of region locating for the user
According to；The communication quality data includes: switching rate, cutting off rate, drop rate, switch rate and the percent of call completed of 5G network；The user
Score data is the marking data that user carries out in specific time；
Step S3.4.2: one weighting function of setting uses wiIt indicates, by each input variable and corresponding weighting function
Convolution algorithm is carried out, the first intermediate result is obtained；
Step S3.4.3: one excitation function of setting, the excitation function are as follows:Set neural network
Neuron threshold value are as follows: Θ；First intermediate result and the excitation function and neuron threshold value are subjected to operation, before obtaining Godwards
Result through network are as follows:
Step S3.4.5: the training error of feedforward neural network is calculated；Since the output variable E of this training is " client
To the impression score of Web vector graphic ", but it is O that a predicted value can be generated after model training, therefore obtain error function are as follows:
Wherein m represents the quantity for inputting this modeling sample, and i indicates i-th of variable；
Step S3.4.6: backpropagation updates weight w
To make error smaller and smaller, the accuracy of model prediction is improved, neural network can be from output layer backpropagation data
To input layer, the value of weight w is readjusted, the deconditioning after model error reaches minimum completes model creation.
Further, the step S1 and step S2 further include: step S1 carries out data encryption when data are transmitted, and
When receiving data, the method for carrying out data deciphering, the method executes following steps to step S2:
Step 1: data encryption, comprising: carry out coded communication in systems, it is initial to generate AES using SAES Encryption Algorithm
Key；
AES initial key is unfolded to obtain AES encryption key；According to AES encryption algorithm, encrypted using AES encryption key
Confidential information to be added；
Wherein, during generating AES initial key using SAES Encryption Algorithm: configuration SAES initial key and SAES
In plain text；SAES key handling is carried out to SAES initial key, obtains processing result；Cipher key spreading is carried out to processing result, to obtain
Obtain SAES encryption key；According to SAES Encryption Algorithm, in plain text using SAES encryption keys SAES, AES initial key is generated；
Wherein, executing SAES key handling to SAES initial key includes: by several M groups n in SAES initial key
Data are combined into M*n Bits Serial stream；Cyclic shift processing is carried out to M*n Bits Serial stream, generates new M*n Bits Serial stream；To new
M*n Bits Serial stream carry out key selection processing, several L group adjacent datas are selected from new M*n Bits Serial stream, wherein by phase
The input that the L group of adjacent data is unfolded as SAES encryption key；
Step 2: carrying out the inverse process of data encryption.
A kind of 5G network neural network based assesses device automatically, which is characterized in that a kind of described device are as follows: nonvolatile
The computer readable storage medium of property, the storage medium store computations comprising: for acquiring 5G front end applications equipment
Data information, and the code segment that collected data information is sent；It is collected for receiving front-end acquisition module
Data information, and to the code segment that the data information received is stored；For being counted according to the data information received
According to processing, data modeling is carried out according to the result of data processing, generates the code of the data model of 5G network automatic evaluation system
Section；For according to the data model of the 5G network automatic evaluation system of generation and the data of collected front end applications equipment
Information, the automatic network quality for assessing current network carry out usertracking, performance optimization, customer service and effect of optimization analysis
Processing code segment.
A kind of of the invention 5G network automatic evaluation system neural network based, method and device, have following beneficial
Effect: the present invention establishes nerve by collecting the data information of 5G network in application process comprehensively, according to these data informations
Network model is analyzed 5G network, most using the neural network model according to the data information in user's use process
The defects of discovery network operation process eventually, and improvement project is proposed to network, it can be adapted for the big network structure of data volume
In, meanwhile, carrying out analysis automatically reduces cost brought by manual analysis；In addition, being used during client data transfers
Complete AES encryption and decryption, guarantee the safety of data, improve the safety of whole system.
Detailed description of the invention
Fig. 1 is that the system structure of 5G network automatic evaluation system neural network based provided in an embodiment of the present invention is illustrated
Figure；
Fig. 2 is the content structure schematic diagram of the data information of the 5G front end applications equipment in the embodiment of the present invention；
Fig. 3 is the method flow schematic diagram of 5G network automatic evaluation method neural network based of the invention.
Fig. 4 is the system structure diagram of 5G network automatic evaluation system neural network based provided by the invention.
Specific embodiment
With reference to the accompanying drawing and the embodiment of the present invention is described in further detail method of the invention.
Embodiment 1
As shown in Figure 1, Figure 2 and Figure 4, a kind of 5G network automatic evaluation system neural network based, the system comprises:
Front end acquisition module, for acquiring the data information of 5G front end applications equipment, and by collected data information into
Row is sent；The data information includes: user's master data, scene classification data, communications device data, communication quality data and
User's score data；
Data loading module is used for the collected data information of receiving front-end acquisition module, and believes the data received
Breath is stored；
Data processing and automatic evaluation module, for carrying out data processing according to the data information received, according to data
The result of processing carries out data modeling, generates the data model of 5G network automatic evaluation system；
Using supporting module, for the data model and front-end collection according to the 5G network automatic evaluation system of generation
The data information of the collected front end applications equipment of module, the automatic network quality for assessing current network, carries out usertracking, property
It can optimize, the processing of customer service and effect of optimization analysis.
Specifically, the present invention can extract in 5G Operation Network, the operation data of mobile client, then according to operation number
According to data progress data forecast analysis processing, according to analysis processing as a result, being optimized to 5G network.
Embodiment 2
On the basis of a upper embodiment, the front end acquisition module includes: data acquisition unit, front-end local data
Library, front end data transmission server and DEU data encryption unit；The data acquisition unit acquires user's master data, scene classification
Collected data are sent front-end local data by data, communications device data, communication quality data and user's score data
It is stored in library；The front end data transmission server sends the data stored in front-end local database；It is described
DEU data encryption unit encrypts data when front end data transmission server carries out data transmission.
Specifically, DEU data encryption unit is as follows to the process of data encryption: expansion AES initial key is to obtain AES encryption
Key；According to AES encryption algorithm, confidential information to be added is encrypted using AES encryption key；
Wherein, during generating AES initial key using SAES Encryption Algorithm: configuration SAES initial key and SAES
In plain text；SAES key handling is carried out to SAES initial key, obtains processing result；Cipher key spreading is carried out to processing result, to obtain
Obtain SAES encryption key；According to SAES Encryption Algorithm, in plain text using SAES encryption keys SAES, AES initial key is generated.
Wherein, executing SAES key handling to SAES initial key includes: by several M groups n in SAES initial key
Data are combined into M*n Bits Serial stream；Cyclic shift processing is carried out to M*n Bits Serial stream, generates new M*n Bits Serial stream；To new
M*n Bits Serial stream carry out key selection processing, several L group adjacent datas are selected from new M*n Bits Serial stream, wherein by phase
The input that the L group of adjacent data is unfolded as SAES encryption key.
Embodiment 3
On the basis of a upper embodiment, user's master data is included at least: user's registration essential information, address area
Domain division information, 5G set meal type information and connection network number information；The scene classification data are user in specific time
Interior be engaged in behavioral data, the behavior include: to see TV, play game, see video, browsing webpage and listen to music；It is described logical
Letter device data includes: base station number, base station performance and base station distance number in 5 kilometers of circumference of region locating for the user
According to；The communication quality data includes: switching rate, cutting off rate, drop rate, switch rate and the percent of call completed of 5G network；The user
Score data is the marking data that user carries out in specific time.
Specifically, user is divided into different groups of users in user in 5G network data transmission.
First, it is assumed that two groups of users, some user equipment in the first groups of users is had received from base station
After information, framing is carried out to the information, then from the multiple sub-frame informations being separated into, uplink ginseng can be transmitted by filtering out
Examine the subframe of signal；
Then, which continues to receive the information from base station, in the subsequent information received, screening
Out with the sequence of transmission uplink reference signals.
Wherein, because different groups receive the base station information that base station information is all different, and the first groups of users receives
In screen can be used to transmit uplink reference signals subframe and second user group receive base station information sieve
What is selected can be different for transmitting the subframe of uplink reference information.Therefore certain in first user's groups of users
The information for the base station that a user receives for the first time, can be used to identify the first groups of users and other users group is used to transmit
The difference of the subframe of uplink signal.
After first groups of users receives the information for receiving base station for the second time, joined using filtering out with transmission uplink
The sequence for examining signal carries the subframe for receiving information sifting for the first time and going out by the sequence, then passes through the subframe for uplink
Link reference signal transmission is to base station.
Above scheme can effectively reduce the interference between different user group between uplink reference signals.
Embodiment 4
On the basis of a upper embodiment, the data loading module includes: data decryption unit, remote data transmission clothes
Business device, interface server and remote database；The number for the encryption that the data decryption unit sends over front end acquisition module
It is believed that breath is decrypted, the data information after decryption is then sent to remote data transmission server；The interface server
It is connected directly, the data information after decryption that remote data server receives is called, calling is arrived with remote database
Data information be sent to remote database and stored.
Embodiment 5
On the basis of a upper embodiment, the data processing and automation evaluation module include: data processing and modeling
Processor；The data processing and modeling processor call directly the data information of remote database, carry out data modeling, generate
The data model of 5G network automatic evaluation system；The data processing and modeling processor include: data pre-processing unit, data
Specification unit, data normalization unit, algorithm predicting unit and modeling analysis unit；The data prediction to data information according to
It is secondary to be removed unique attribute, processing missing values and rejecting outliers processing；The hough transformation unit, for locating data in advance
Data after reason carry out specification processing, so that treated that data are irrelevant two-by-two for specification, but can keep original letter as far as possible
Breath；The data normalization unit is allowed to specification treated data bi-directional scaling to fall into a small specific sections；
The algorithm predicting unit carries out data modeling with the data after your poly- data normalization cell processing；The modeling analysis list
Customer satisfaction data and existing client satisfaction data of the member for being generated by computation model carry out precision calculating.
Embodiment 6:
As shown in figure 3, a kind of 5G network automatic evaluation method neural network based, the method executes following steps:
Step S1: the data information of acquisition 5G front end applications equipment, and collected data information is sent；
Step S2: the collected data information of receiving front-end acquisition module, and the data information received is stored；
Step S3: carrying out data processing according to the data information received, carries out data according to the result of data processing and builds
Mould generates the data model of 5G network automatic evaluation system；
Step S4: according to the data model of the 5G network automatic evaluation system of generation, and: collected front end applications are set
Standby data information, the automatic network quality for assessing current network carry out usertracking, performance optimization, customer service and optimization
The processing of effect analysis.
Embodiment 7:
On the basis of a upper embodiment, the step S3: carrying out data processing according to the data information received, according to
The result of data processing carries out data modeling, and the method for generating the data model of 5G network automatic evaluation system executes following step
It is rapid:
Step S3.1: data prediction is carried out, comprising: removal unique attribute, processing missing values and rejecting outliers and place
Reason；
Step S3.2: hough transformation processing is carried out, comprising: remove average value, calculate covariance matrix, calculate covariance matrix
Eigen vector, characteristic value is sorted from large to small, maximum k feature vector is retained, data are transformed into k
In the new space of feature vector building；Finally treated new data, it is irrelevant two-by-two between these data, but can to the greatest extent may be used
It is able to maintain original information.
Step S3.3: it carries out data normalization processing and is allowed to data bi-directional scaling to fall into a small given zone
Between；Wherein, using following transfer function, linear transformation is carried out to data, so that result is fallen on [0,1] section, transfer function is such as
Under:
Wherein, x*For data normalization treated result；X is data to be processed；Min is
Minimum value in data；Max is the maximum value in data；
Step S3.4: data modeling is carried out；
Step S3.5: effect analysis is carried out, comprising: after model training, using following formula, computation model is generated
Customer satisfaction data and existing client satisfaction data carry out precision calculate to get arrive R2Score, score is higher, indicates
Model precision is better；
The wherein customer satisfaction data (predicted value) that y representative model generates；
Represent existing client satisfaction data；
nsamplesRepresent the sample size size for entering model.
Specifically, unique attribute is usually some id attributes, these attributes can not portray sample when removal unique attribute
The regularity of distribution of itself, so simply deleting these attributes.
When handling missing values, missing values are filled up using the method for mean value interpolation: if the distance of sample attribute is
Mensurable, then carry out the value of interpolation missing using the average value of the Attribute Valid；If distance be it is immeasurable, make
The value lacked with the mode of the Attribute Valid come interpolation.
When rejecting outliers and processing, being described property of variable is statisticallyd analyze, using box map analysis, obtains sample
Average value, maximum value, minimum value and standard deviation.If sample value and the deviation of average value are greater than 3 times of standard deviations, it is detected as exception
It is worth (3 σ rule)；The method that the processing of exceptional value is equally filled up using mean value.
Wherein, the purpose of hough transformation is: being to have certain correlativity between many situations, variable, when two
When having certain correlativity between variable, the two variables can be construed to and reflect that the information of this project has certain overlapping.It is main
Constituent analysis is all variables for originally proposing, it is extra that duplicate variable (variable of close relation) is left out, and is established to the greatest extent
The new variables that may lack, so that these new variables are incoherent two-by-two, and these new variables are in the information of reflection original sample
Aspect keeps original information as far as possible.Dimensionality reduction specification is carried out to data using Principal Component Analysis.Principal component analysis (PCA) is
A kind of more common dimensionality reduction technology, the thought of PCA is that dimensional feature is mapped in dimension, this dimension is completely new orthogonal characteristic.This
Dimensional feature is known as pivot, is the dimensional feature for reconfiguring out.In PCA, data are transformed into new seat from original coordinate system
Under mark system, the selection of new coordinate system and data itself are closely related.Wherein, what first new reference axis selected is original
The maximum direction of variance in data, what second new reference axis was chosen is orthogonal with first reference axis and has maximum variance
Direction, and so on, we can get such reference axis.
Data normalization
By data bi-directional scaling, it is allowed to fall into a small specific sections.In the processing of the index of certain comparisons and evaluation
In often use, remove the unit limitation of data, nondimensional pure values be translated into, convenient for not commensurate or magnitude
Index, which is able to carry out, to be compared and weights.
Embodiment 8
On the basis of a upper embodiment, the step S3.4: the method for carrying out data modeling executes following steps:
Step S3.4.1: it obtains and uses x as input variable for the data of modelingiIt indicates, wherein i represents the data
In i-th of variable；The xiIt includes at least: user's registration essential information, address area division information, 5G set meal type information
With connection network number information；The behavioral data that the scene classification data are engaged in specific time by user, the row
To include: to see TV, play game, see that video, browsing and listen to music webpage；The communications device data includes: locating for the user
5 kilometers of circumference of region in base station number, base station performance and base station distance data；The communication quality data includes: 5G
Switching rate, cutting off rate, drop rate, switch rate and the percent of call completed of network；User's score data is user in specific time
The marking data of progress.
Step S3.4.2: one weighting function of setting uses wiIt indicates, by each input variable and corresponding weighting function
Convolution algorithm is carried out, the first intermediate result is obtained；
Step S3.4.3: one excitation function of setting, the excitation function are as follows:Set neural network
Neuron threshold value are as follows: Θ；First intermediate result and the excitation function and neuron threshold value are subjected to operation, before obtaining Godwards
Result through network are as follows:
Step S3.4.5: the training error of feedforward neural network is calculated；Since the output variable E of this training is " client
To the impression score of Web vector graphic ", but it is O that a predicted value can be generated after model training, therefore obtain error function are as follows:
Wherein m represents the quantity for inputting this modeling sample, and i indicates i-th of variable.
Step S3.4.6: backpropagation updates weight w
To make error smaller and smaller, the accuracy of model prediction is improved, neural network can be from output layer backpropagation data
To input layer, the value of weight w is readjusted, the deconditioning after model error reaches minimum completes model creation.
Embodiment 9
On the basis of a upper embodiment, the step S1 and step S2 further include: step S1 is carried out when data are transmitted
When receiving data, the method for carrying out data deciphering, the method executes following steps by data encryption and step S2:
Step 1: data encryption, comprising: carry out coded communication in systems, it is initial to generate AES using SAES Encryption Algorithm
Key；
AES initial key is unfolded to obtain AES encryption key；According to AES encryption algorithm, encrypted using AES encryption key
Confidential information to be added；
Wherein, during generating AES initial key using SAES Encryption Algorithm: configuration SAES initial key and SAES
In plain text；SAES key handling is carried out to SAES initial key, obtains processing result；Cipher key spreading is carried out to processing result, to obtain
Obtain SAES encryption key；According to SAES Encryption Algorithm, in plain text using SAES encryption keys SAES, AES initial key is generated.
Wherein, executing SAES key handling to SAES initial key includes: by several M groups n in SAES initial key
Data are combined into M*n Bits Serial stream；Cyclic shift processing is carried out to M*n Bits Serial stream, generates new M*n Bits Serial stream；To new
M*n Bits Serial stream carry out key selection processing, several L group adjacent datas are selected from new M*n Bits Serial stream, wherein by phase
The input that the L group of adjacent data is unfolded as SAES encryption key；
Step 2: carrying out the inverse process of data encryption.
Embodiment 10
A kind of 5G network neural network based assesses device, described device automatically are as follows: a kind of computer of non-transitory
Readable storage medium storing program for executing, the storage medium store computations comprising: the data for acquiring 5G front end applications equipment are believed
Breath, and the code segment that collected data information is sent；For the collected data information of receiving front-end acquisition module,
And to the code segment that the data information received is stored；For carrying out data processing, root according to the data information received
Data modeling is carried out according to the result of data processing, generates the code segment of the data model of 5G network automatic evaluation system；For root
According to the data model of the 5G network automatic evaluation system of generation and the data information of collected front end applications equipment, automatically
The network quality of current network is assessed, the generation of usertracking, performance optimization, customer service and the processing of effect of optimization analysis is carried out
Code section.
Person of ordinary skill in the field can be understood that, for convenience and simplicity of description, foregoing description
The specific work process of system and related explanation, can refer to corresponding processes in the foregoing method embodiment, details are not described herein.
It should be noted that system provided by the above embodiment, only illustrate with the division of above-mentioned each functional module
It is bright, in practical applications, it can according to need and complete above-mentioned function distribution by different functional modules, i.e., it will be of the invention
Module or step in embodiment are decomposed or are combined again, for example, the module of above-described embodiment can be merged into a module,
It can also be further split into multiple submodule, to complete all or part of the functions described above.The present invention is implemented
Module, the title of step involved in example, it is only for distinguish modules or step, be not intended as to of the invention improper
It limits.
Person of ordinary skill in the field can be understood that, for convenience and simplicity of description, foregoing description
The specific work process and related explanation of storage device, processing unit, can refer to corresponding processes in the foregoing method embodiment,
Details are not described herein.
Those skilled in the art should be able to recognize that, mould described in conjunction with the examples disclosed in the embodiments of the present disclosure
Block, method and step, can be realized with electronic hardware, computer software, or a combination of the two, software module, method and step pair
The program answered can be placed in random access memory (RAM), memory, read-only memory (ROM), electrically programmable ROM, electric erasable and can compile
Any other form of storage well known in journey ROM, register, hard disk, moveable magnetic disc, CD-ROM or technical field is situated between
In matter.In order to clearly demonstrate the interchangeability of electronic hardware and software, in the above description according to function generally
Describe each exemplary composition and step.These functions are executed actually with electronic hardware or software mode, depend on technology
The specific application and design constraint of scheme.Those skilled in the art can carry out using distinct methods each specific application
Realize described function, but such implementation should not be considered as beyond the scope of the present invention.
Term " first ", " second " etc. are to be used to distinguish similar objects, rather than be used to describe or indicate specific suitable
Sequence or precedence.
Term " includes " or any other like term are intended to cover non-exclusive inclusion, so that including a system
Process, method, article or equipment/device of column element not only includes those elements, but also including being not explicitly listed
Other elements, or further include the intrinsic element of these process, method, article or equipment/devices.
So far, it has been combined preferred embodiment shown in the drawings and describes technical solution of the present invention, still, this field
Technical staff is it is easily understood that protection scope of the present invention is expressly not limited to these specific embodiments.Without departing from this
Under the premise of the principle of invention, those skilled in the art can make equivalent change or replacement to the relevant technologies feature, these
Technical solution after change or replacement will fall within the scope of protection of the present invention.
The foregoing is only a preferred embodiment of the present invention, is not intended to limit the scope of the present invention.
Claims (10)
1. a kind of 5G network automatic evaluation system neural network based, which is characterized in that the system comprises:
Front end acquisition module is sent out for acquiring the data information of 5G front end applications equipment, and by collected data information
It send；The data information includes: user's master data, scene classification data, communications device data, communication quality data and user
Score data；
Data loading module is believed for receiving the collected data information of the front end acquisition module, and to the data received
Breath is stored；
Data processing and automatic evaluation module, for carrying out data processing according to the data information received, according to data processing
Result carry out data modeling, generate 5G network automatic evaluation system data model；
Using supporting module, for the data model and the front-end collection according to the 5G network automatic evaluation system of generation
The data information of the collected front end applications equipment of module, the automatic network quality for assessing current network, carries out usertracking, property
It can optimize, the processing of customer service and effect of optimization analysis.
2. the system as claimed in claim 1, which is characterized in that the front end acquisition module includes: data acquisition unit, front end
Local data base, front end data transmission server and DEU data encryption unit；Data acquisition unit acquisition user's master data,
Scene classification data, communications device data, communication quality data and user's score data send collected data to described
It is stored in front-end local database；The number that the front end data transmission server will store in the front-end local database
According to being sent；The DEU data encryption unit carries out data when the front end data transmission server carries out data transmission
Encryption.
3. system as claimed in claim 2, which is characterized in that user's master data includes at least: user's registration is basic
Information, address area division information, 5G set meal type information and connection network number information；The scene classification data are user
The behavioral data being engaged in specific time；The communications device data includes: in 5 kilometers of circumference of region locating for user
Base station number, base station performance and base station distance data；The communication quality data include: the switching rate of 5G network, cutting off rate,
Drop rate, switch rate and percent of call completed；User's score data is the marking data that user carries out in specific time.
4. the system as claimed in claim 1, which is characterized in that the data loading module includes: data decryption unit, distal end
Data transfer server, interface server and remote database；The data decryption unit sends over front end acquisition module
The data information of encryption be decrypted, the data information after decryption is then sent to the remote data transmission server；
The interface server and remote database are connected directly, the data letter after the decryption that the remote data server is received
Breath is called, and the data information called is sent to the remote database and is stored.
5. system as claimed in claim 4, which is characterized in that the data processing and automation evaluation module include: data
Processing and modeling processor；The data processing and modeling processor call directly the data information of remote database, are counted
According to modeling, the data model of 5G network automatic evaluation system is generated；The data processing and modeling processor include: that data are located in advance
Manage unit, hough transformation unit, data normalization unit, algorithm predicting unit and modeling analysis unit；The data prediction
Unique attribute, processing missing values and rejecting outliers processing are successively removed to data information；The hough transformation unit is used
In the data after data prediction are carried out specification processing, so that treated that data are irrelevant two-by-two for specification, but it is able to maintain original
Some information；The data normalization unit is allowed to specification treated data bi-directional scaling to fall into a given zone
Between；The algorithm predicting unit carries out data modeling with the data after your poly- data normalization cell processing；The modeling analysis
The customer satisfaction data and existing client satisfaction data that unit is used to generate by computation model carry out precision calculating.
6. a kind of based on method described in one of claim 1 to 5, which is characterized in that the method executes following steps:
Step S1: the data information of acquisition 5G front end applications equipment, and collected data information is sent；
Step S2: the collected data information of receiving front-end acquisition module, and the data information received is stored；
Step S3: carrying out data processing according to the data information received, carries out data modeling according to the result of data processing, raw
At the data model of 5G network automatic evaluation system；
Step S4: according to the data model of the 5G network automatic evaluation system of generation and collected front end applications equipment
Data information, the automatic network quality for assessing current network carry out usertracking, performance optimization, customer service and effect of optimization
The processing of analysis.
7. method as claimed in claim 6, which is characterized in that the step S3: counted according to the data information received
According to processing, data modeling is carried out according to the result of data processing, the method for generating the data model of 5G network automatic evaluation system
Execute following steps:
Step S3.1: data prediction is carried out, comprising: removal unique attribute, processing missing values and rejecting outliers and processing；
Step S3.2: hough transformation processing is carried out, comprising: the spy for removing average value, calculating covariance matrix, calculating covariance matrix
Value indicative and feature vector sort from large to small characteristic value, retain maximum k feature vector, data are transformed into k feature
In the new space of vector building；Finally treated new data, it is irrelevant two-by-two between these data；
Step S3.3: it carries out data normalization processing and is allowed to data bi-directional scaling to fall into a specific sections；Wherein, make
With following transfer function, linear transformation is carried out to data, so that result is fallen on [0,1] section, transfer function is as follows:
Wherein, x*For data normalization treated result；X is data to be processed；Min is in data
Minimum value；Max is the maximum value in data；
Step S3.4: data modeling is carried out；
Step S3.5: effect analysis is carried out, comprising: after model training, using following formula, the visitor of computation model generation
Family satisfaction data and existing client satisfaction data carry out precision and calculate to get R is arrived2Score, score is higher, indicates model
Precision is better；
The wherein customer satisfaction data (predicted value) that y representative model generates；
Represent existing client satisfaction data；
nsamplesRepresent the sample size size for entering model.
8. the method for claim 7, which is characterized in that the step S3.4: carry out data modeling method execute with
Lower step:
Step S3.4.1: it obtains and uses x as input variable for the data of modelingiIt indicates, wherein i represents the in the data
I variable；The xiIt includes at least: user's registration essential information, address area division information, 5G set meal type information and connection
Network number information；The behavioral data that the scene classification data are engaged in specific time by user；The communication equipment
Data include: base station number, base station performance and base station distance data in 5 kilometers of circumference of region locating for the user；It is described
Communication quality data includes: switching rate, cutting off rate, drop rate, switch rate and the percent of call completed of 5G network；User's score data
The marking data carried out in specific time for user；
Step S3.4.2: one weighting function of setting uses wiIt indicates, each input variable is carried out with corresponding weighting function
Convolution algorithm obtains the first intermediate result；
Step S3.4.3: one excitation function of setting, the excitation function are as follows:Set the mind of neural network
Through first threshold value are as follows: Θ；First intermediate result and the excitation function and neuron threshold value are subjected to operation, obtain BP Neural Network
The result of network are as follows:
Step S3.4.5: the training error of feedforward neural network is calculated；Since the output variable E of this training is that " client is to net
The impression score that network uses ", but it is O that a predicted value can be generated after model training, therefore obtain error function are as follows:
Wherein m represents the quantity for inputting this modeling sample, and i indicates i-th of variable；
Step S3.4.6: backpropagation updates weight w；
To make error smaller and smaller, the accuracy of model prediction is improved, neural network can be from output layer backpropagation data to defeated
Enter layer, readjust the value of weight w, the deconditioning after model error reaches minimum completes model creation.
9. method as claimed in claim 6, which is characterized in that the step S1 and step S2 further include: step S1 is in data
Data encryption and step S2 are carried out when transmission when receiving data, the method for carrying out data deciphering, the method executes following
Step:
Step 1: data encryption, comprising: carry out coded communication in systems, generate AES initial key using SAES Encryption Algorithm；
AES initial key is unfolded to obtain AES encryption key；According to AES encryption algorithm, encrypted using AES encryption key to be added
Confidential information；
Wherein, during generating AES initial key using SAES Encryption Algorithm: configuration SAES initial key and SAES are bright
Text；SAES key handling is carried out to SAES initial key, obtains processing result；Cipher key spreading is carried out to processing result, to obtain
SAES encryption key；According to SAES Encryption Algorithm, in plain text using SAES encryption keys SAES, AES initial key is generated；
Wherein, executing SAES key handling to SAES initial key includes: by several M group n-bit datas in SAES initial key
It is combined into M*n Bits Serial stream；Cyclic shift processing is carried out to M*n Bits Serial stream, generates new M*n Bits Serial stream；To new M*n
Bits Serial stream carries out key selection processing, several L group adjacent datas is selected from new M*n Bits Serial stream, wherein by consecutive number
According to the input that is unfolded as SAES encryption key of L group；
Step 2: carrying out the inverse process of data encryption.
10. a kind of device of the method based on one of claim 6 to 9, which is characterized in that a kind of described device are as follows: nonvolatile
The computer readable storage medium of property, the storage medium store computations comprising: for acquiring 5G front end applications equipment
Data information, and the code segment that collected data information is sent；It is collected for receiving front-end acquisition module
Data information, and to the code segment that the data information received is stored；For being counted according to the data information received
According to processing, data modeling is carried out according to the result of data processing, generates the code of the data model of 5G network automatic evaluation system
Section；For according to the data model of the 5G network automatic evaluation system of generation and the data of collected front end applications equipment
Information, the automatic network quality for assessing current network carry out usertracking, performance optimization, customer service and effect of optimization analysis
Processing code segment.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910436086.5A CN110213774A (en) | 2019-05-23 | 2019-05-23 | A kind of 5G network automatic evaluation system neural network based, method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910436086.5A CN110213774A (en) | 2019-05-23 | 2019-05-23 | A kind of 5G network automatic evaluation system neural network based, method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110213774A true CN110213774A (en) | 2019-09-06 |
Family
ID=67788377
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910436086.5A Pending CN110213774A (en) | 2019-05-23 | 2019-05-23 | A kind of 5G network automatic evaluation system neural network based, method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110213774A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110617804A (en) * | 2019-09-25 | 2019-12-27 | 浙江海洋大学 | Marine ecological environment detection system and method based on remote sensing technology |
| CN110672534A (en) * | 2019-10-18 | 2020-01-10 | 湖南国天电子科技有限公司 | Ocean water quality testing system, method and device |
| CN111131149A (en) * | 2019-11-13 | 2020-05-08 | 江苏飞搏软件股份有限公司 | Method for acquiring data of mobile terminal in cross-domain mode and analyzing abnormal access |
| CN111650898A (en) * | 2020-05-13 | 2020-09-11 | 大唐七台河发电有限责任公司 | Distributed control system and method with high fault tolerance performance |
| CN112070224A (en) * | 2020-08-26 | 2020-12-11 | 成都品果科技有限公司 | Revision system and method of sample for neural network training |
| CN114257523A (en) * | 2021-12-08 | 2022-03-29 | 中国联合网络通信集团有限公司 | User perception prediction method, system, device and computer storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103152396A (en) * | 2013-02-05 | 2013-06-12 | 华南师范大学 | Data placement method and device applied to content distribution network system |
| CN107634866A (en) * | 2017-10-27 | 2018-01-26 | 朱秋华 | A kind of distribution network communication system performance estimating method and device |
| CN110113214A (en) * | 2019-05-16 | 2019-08-09 | 青岛博展智能科技有限公司 | A kind of 5G network automatic evaluation system neural network based, method and device |
-
2019
- 2019-05-23 CN CN201910436086.5A patent/CN110213774A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103152396A (en) * | 2013-02-05 | 2013-06-12 | 华南师范大学 | Data placement method and device applied to content distribution network system |
| CN107634866A (en) * | 2017-10-27 | 2018-01-26 | 朱秋华 | A kind of distribution network communication system performance estimating method and device |
| CN110113214A (en) * | 2019-05-16 | 2019-08-09 | 青岛博展智能科技有限公司 | A kind of 5G network automatic evaluation system neural network based, method and device |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110617804A (en) * | 2019-09-25 | 2019-12-27 | 浙江海洋大学 | Marine ecological environment detection system and method based on remote sensing technology |
| CN110672534A (en) * | 2019-10-18 | 2020-01-10 | 湖南国天电子科技有限公司 | Ocean water quality testing system, method and device |
| CN111131149A (en) * | 2019-11-13 | 2020-05-08 | 江苏飞搏软件股份有限公司 | Method for acquiring data of mobile terminal in cross-domain mode and analyzing abnormal access |
| CN111650898A (en) * | 2020-05-13 | 2020-09-11 | 大唐七台河发电有限责任公司 | Distributed control system and method with high fault tolerance performance |
| CN111650898B (en) * | 2020-05-13 | 2023-10-20 | 大唐七台河发电有限责任公司 | Distributed control system and method with high fault tolerance performance |
| CN112070224A (en) * | 2020-08-26 | 2020-12-11 | 成都品果科技有限公司 | Revision system and method of sample for neural network training |
| CN112070224B (en) * | 2020-08-26 | 2024-02-23 | 成都品果科技有限公司 | Revision system and method of samples for neural network training |
| CN114257523A (en) * | 2021-12-08 | 2022-03-29 | 中国联合网络通信集团有限公司 | User perception prediction method, system, device and computer storage medium |
| CN114257523B (en) * | 2021-12-08 | 2023-07-07 | 中国联合网络通信集团有限公司 | User perception prediction method, system, device and computer storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110113214A (en) | A kind of 5G network automatic evaluation system neural network based, method and device | |
| CN110213774A (en) | A kind of 5G network automatic evaluation system neural network based, method and device | |
| CN111190939A (en) | User portrait construction method and device | |
| Liu et al. | Image recognition technology based on machine learning | |
| CN108052639A (en) | Industry user based on carrier data recommends method and device | |
| CN107809766B (en) | Method and device for generating machine learning sample for network optimization | |
| CN113379176B (en) | Method, device, equipment and readable storage medium for detecting abnormal data of telecommunication network | |
| CN117436724B (en) | Multi-source data visual analysis method and system based on smart city | |
| CN114386514B (en) | Unknown flow data identification method and device based on dynamic network environment | |
| CN111062520B (en) | Hostname feature prediction method based on random forest algorithm | |
| Zhang et al. | Optimization of image transmission in cooperative semantic communication networks | |
| CN112926429B (en) | Machine-check model training and video machine-check method, device, equipment and storage medium | |
| CN112214677B (en) | Point of interest recommendation method and device, electronic equipment and storage medium | |
| CN108462544A (en) | A kind of frequency spectrum sensing method and device | |
| CN113591674B (en) | Edge environment behavior recognition system for real-time video stream | |
| CN109862587A (en) | Mobile network quality appraisal procedure based on multiple features time series and self-encoding encoder | |
| CN105873119A (en) | Method for classifying flow use behaviors of mobile network user groups | |
| Navarro et al. | GraphMAD: Graph mixup for data augmentation using data-driven convex clustering | |
| CN114282607A (en) | Double-sieve model-based dispersion trajectory analysis method and system | |
| CN110855474B (en) | Network feature extraction method, device, equipment and storage medium of KQI data | |
| CN110489665B (en) | Microblog personalized recommendation method based on scene modeling and convolutional neural network | |
| CN115906927A (en) | Data access analysis method and system based on artificial intelligence and cloud platform | |
| CN109412882B (en) | Feature extraction method, device, server and computer-readable storage medium | |
| CN114385876B (en) | Model search space generation method, device and system | |
| CN102129575A (en) | Pornographic image analysis system based on color space skin color 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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190906 |
|
| RJ01 | Rejection of invention patent application after publication |