CN110113119A - A kind of Wireless Channel Modeling method based on intelligent algorithm - Google Patents
A kind of Wireless Channel Modeling method based on intelligent algorithm Download PDFInfo
- Publication number
- CN110113119A CN110113119A CN201910343878.8A CN201910343878A CN110113119A CN 110113119 A CN110113119 A CN 110113119A CN 201910343878 A CN201910343878 A CN 201910343878A CN 110113119 A CN110113119 A CN 110113119A
- Authority
- CN
- China
- Prior art keywords
- model
- wireless channel
- radio wave
- intelligent algorithm
- channel
- 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
- 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
- G06N3/044—Recurrent networks, e.g. Hopfield 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
- G06N3/04—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
- G06N3/06—Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons
- G06N3/061—Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons using biological neurons, e.g. biological neurons connected to an integrated circuit
-
- 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
- H04B—TRANSMISSION
- H04B17/00—Monitoring; Testing
- H04B17/30—Monitoring; Testing of propagation channels
- H04B17/391—Modelling the propagation channel
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B17/00—Monitoring; Testing
- H04B17/30—Monitoring; Testing of propagation channels
- H04B17/391—Modelling the propagation channel
- H04B17/3912—Simulation models, e.g. distribution of spectral power density or received signal strength indicator [RSSI] for a given geographic region
Abstract
The Wireless Channel Modeling method based on intelligent algorithm that the invention discloses a kind of, belongs to wireless communication field.All kinds of radio wave measurement data that Wireless Channel Modeling is used under default test scene, training intelligent algorithm model are acquired first；Then according to test scene and intelligent algorithm model is preset, combination obtains the wireless channel large-scale fading model and wireless channel multipath fading model under default test scene.Under identical default test scene, input the radio wave measurement data of different geographic regions, training wireless channel size scale fading model, the basic law that the wireless channel model and radio wave propagation characteristic for directly obtaining other geographic areas change with characteristic parameter.Finally according to training sample, basic law and wireless channel size scale fading model, the wireless channel model based on characteristic parameter is established.The present invention more refines and intelligence, has good universality and flexible scalability.
Description
Technical field
The invention belongs to wireless communication field, it is related to Wireless Channel Modeling, it is specifically a kind of based on intelligent algorithm
Wireless Channel Modeling method.
Background technique
According to the difference of radio communication service and frequency range and application thereof, wireless channel model is also different, than if any being suitable for
The channel model of cellular mobile communication, the channel model of railway vehicle-ground wireless communication, the channel model of satellite communication and ground
The channel model of line-of-sight microwave link etc..
Traditional Wireless Channel Modeling method is broadly divided into three classes: method based on theoretical calculation, based on actual measurement
Method and the theoretical synthesis combined with calculating.
Method based on theoretical calculation, such as ray casting, the Correlation Moment tactical deployment of troops etc. generally require to propose some hypothesis items
Part simplifies the mathematical model of wireless channel, such as barriers some in ray casting, the electrical parameter of reflecting surface and scattering surface
It needs to obtain by experiment；The wireless channel model obtained to the method based on theoretical calculation be all it is approximate, and it is actual
Channel condition often has larger difference.
Method based on actual measurement is the main stream approach of current radio channel modeling, such as based on the random of space geometry
Theory method, but test job is especially time-consuming.Collecting test datamation is only in some fixed points or using movement at present
Mode carry out on a few streets (line), it is difficult to obtain the magnanimity number based on face (full wafer test default scene) or network level
Mathematical modeling is carried out according to wireless channel.
Method based on actual measurement may be only available for specifically according to the channel model that actual measurement data is established every time
Environment, once environment changes, original model is not just available.For example it is widely used in forth generation mobile communication system
What the random theory method that the wireless channel model of (4G) is namely based on space geometry obtained, telecom operators' network planning network optimization it
Before, channel model needs to carry out according to scene a large amount of tests and is localized that correction rear can be used.
Furthermore wireless channel model and radio propagation environment (such as topography and geomorphology) are closely bound up, and traditional modeling algorithm is not
Fusion treatment can be carried out to multivariate data.
Therefore the either method based on theoretical calculation is still based on the method for actual measurement, the standard of the channel model of acquisition
Exactness is lower.Therefore, most conventional channel modeling method be unable to do without the test of actual scene, the universality of channel model and can expand
Malleability is poor.Once having converted application scenarios, it is necessary to carry out radio wave test data collection and Channel Modeling work again
Make, causes the waste of a large amount of human and material resources, financial resources.
Based on intelligent algorithm, the Channel Modeling to technical research such as wireless communication system transmission is one and brand-new grinds
Study carefully field, the report that at home and abroad few systems for carrying out this respect work at present.
Summary of the invention
The present invention is for the problems of above-mentioned existing Wireless Channel Modeling method, in order to solve modeling procedure time-consuming
Long, complexity and at high cost and existing channel model scalability and the problem that universality is poor, accuracy is low, provide
A kind of Wireless Channel Modeling method based on intelligent algorithm.
Specific step is as follows:
Step 1: acquiring all kinds of radio wave measurement data for being used for Wireless Channel Modeling under default test scene；
Data include radio wave measurement data, terrain data and hydrometeorological data；
Step 2: utilizing all kinds of radio wave measurement data training intelligent algorithm model of acquisition；
Intelligent algorithm model includes reverse transmittance nerve network algorithm and radial basis function neural network algorithm；
Step 3: combination obtains the nothing under default test scene according to default test scene and intelligent algorithm model
Line channel large-scale fading model；
Large-scale fading model is based on reverse transmittance nerve network algorithm, the specific steps are as follows:
Step 301: converting reverse transmittance nerve network modeling institute for the radio wave measurement data of default test scene
The data format needed；
Step 302: selection reverse transmittance nerve network implies the number of plies, every layer of neuron or interstitial content；
Step 303: the randomizing data after conversion format is divided into training sample and test sample；
Step 304: inputting training sample and train back propagation artificial neural network model with Parameters in Mathematical Model, obtain reversed
The weight of required each layer neuron interconnections and input bias in Propagation Neural Network model；
Parameters in Mathematical Model includes: reception and transmission range, and time-domain sampling interval receives level etc.；
Step 305: according to the weight of each layer neuron interconnections and input bias, obtaining the wireless channel under default scene
Large-scale fading model；
Step 306: using test sample to the wireless channel large-scale fading model based on reverse transmittance nerve network into
Row verifying.
Step 4: simultaneously, according to default test scene and intelligent algorithm model, combination is obtained under default test scene
Wireless channel multipath fading model；
Multipath fading model is based on radial basis function neural network algorithm, the specific steps are as follows:
Step 401: converting the radio wave measurement data of default scene to needed for radial basis function neural network modeling
The data format wanted；
Step 402: selection radial basis function neural network implies the number of plies, every layer of neuron or interstitial content；
Step 403: the data format after conversion is divided into training sample and test sample at random；
Step 404: training sample being pre-processed using conventional high resolution algorithm；
Step 405: pretreated training sample and channel parameter are inputted, radial basis function neural network model is trained,
Obtain the weight and input bias of each layer neuron interconnections needed in radial basis function neural network model；
Channel parameter includes: mobile station speed, time-domain sampling interval, time delay and angle spread, Rice factor etc.,
Step 406: according to the weight of each layer neuron interconnections and input bias, obtaining the wireless channel under default scene
Multipath fading model；
Step 407: using test sample to the wireless channel multipath fading based on radial basis function neural network algorithm
Model is verified.
Step 5: inputting the radio wave measurement data of different geographic regions, training nothing under identical default test scene
Line channel size scale fading model, directly obtains the wireless channel model of other geographic areas.
Step 6: simultaneously, using trained wireless channel size scale fading model, obtaining radio wave propagation characteristic with spy
Levy the basic law of Parameters variation；
Characteristic parameter includes time domain, frequency domain, airspace (influence of topography and geomorphology) and scene type etc.；
The radio wave propagation characteristic refers to the basic law that preset condition changes: the time domain changing rule of the characteristic of channel,
The frequency domain changing rule of the characteristic of channel, the airspace changing rule of the characteristic of channel, the characteristic of channel are advised with the variation of default scene type
Rule.
Step 7: establishing and being joined based on feature according to training sample, basic law and wireless channel size scale fading model
Several wireless channel models.
The present invention has the advantages that
A kind of Wireless Channel Modeling method based on intelligent algorithm, breaches traditional modeling method and is difficult to obtain time-varying
The bottleneck of channel model has good universality and flexibly may be used so that wireless channel model more refines and intelligence
Scalability.
Detailed description of the invention
Fig. 1 is a kind of flow diagram of the Wireless Channel Modeling method based on intelligent algorithm of the present invention；
Fig. 2 is that the present invention is based on the processes of the wireless channel large-scale fading modeling method of reverse transmittance nerve network algorithm
Figure；
Fig. 3 is three layers of back propagation artificial neural network model schematic diagram of the invention；
Fig. 4 is that the present invention is based on the streams of the wireless channel multipath fading modeling method of radial basis function neural network algorithm
Cheng Tu；
Fig. 5 is three layers of radial basis function neural network model schematic diagram of the invention.
Specific embodiment
Specific implementation method of the invention is described in detail with reference to the accompanying drawing.
A kind of Wireless Channel Modeling method based on intelligent algorithm of the present invention, is capable of handling off line based on specific environment
The obtained magnanimity radio wave data of the distributed testing of network layers grade and the multi-source datas such as topography and geomorphology and hydrometeorology.
Due to including huge sample in data, more comprehensive radio wave propagation characteristic data information is enumerated, is covered under the environment
Various Radio Links, obtained model will have better universality and higher accuracy.A such as good path damage
Consumption model is difficult the small data based on actual test and obtains, but if using operator in different base station ten hundreds of or million
The user data (information such as reception level and position comprising user) of meter, is built using the wireless channel based on intelligent algorithm
Mould method, available pervasive path loss model and shadow attenuation rule model suitable for this area's environment are other similar
Area surroundings only need to acquire low volume data to the path loss model and shadow fading rule model training obtained before.
As shown in Figure 1, comprising the following steps:
Step 1: acquiring all kinds of radio wave measurement data for being used for Wireless Channel Modeling under default test scene；
Default test scene is different in the description of different radio operation system；Specifically, cell mobile communication systems
Default scene includes dense city macrocell, open Office Area, indoor hot spot and suburb etc., the test of Measurement of Railway Radio Communication System
Scene includes ditch, tunnel, open ground and overpass etc.；
Data must include radio wave measurement data, may also include terrain data and hydrometeorological data etc..
Step 2: utilizing all kinds of radio wave measurement data training intelligent algorithm model of acquisition；
Intelligent algorithm model includes reverse transmittance nerve network algorithm and radial basis function neural network algorithm；
Step 3: according to the intelligent algorithm model that default test scene and training obtain, combination obtains default test
Wireless channel large-scale fading model under scene；
Large-scale fading model is based on reverse transmittance nerve network algorithm, as shown in Figure 2, the specific steps are as follows:
Step 301: converting reverse transmittance nerve network modeling institute for the radio wave measurement data of default test scene
The data format needed；
Step 302: selection reverse transmittance nerve network implies the number of plies, every layer of neuron or interstitial content；
Reverse transmittance nerve network generally comprises three layers or four layers of implicit nervous layer, and every layer of neuron number need to be according to reality
The accuracy of simulation result determines.
Step 303: the randomizing data after conversion format is divided into training sample and test sample；
Step 304: inputting training sample and train back propagation artificial neural network model with Parameters in Mathematical Model, obtain reversed
The weight of required each layer neuron interconnections and input bias in Propagation Neural Network model；
Parameters in Mathematical Model includes: reception and transmission range, and time-domain sampling interval receives level etc.；
Step 305: according to the weight of each layer neuron interconnections and input bias, obtaining the wireless channel under default scene
Large-scale fading model；
Wireless channel large-scale fading model, as shown in figure 3, in modeling process assume number of training be it is N number of, it is each
A sample includes reception and transmission range, base station antenna height, and the relevant parameters such as receiver height are input to multisample x matrix parallel
Reverse transmittance nerve network, then with model parameter matrix θ1Be multiplied generate hidden state matrix a, a again with model parameter matrix θ2
It is multiplied and calculates the output h of neural networkθ(x), θ is obtained in the training process1And θ2Etc. model parameters.
Training to reverse transmittance nerve network includes following four step:
(1) random initializtion Parameters in Mathematical Model θ；
(2) reverse transmittance nerve network algorithm is executed；
(3) cost function related with θ is calculated；
(4) (2)~(3) are repeated until restraining or reaching desired the number of iterations.
Correlation of the path loss with multiple parameters, traditional path loss can be more accurately predicted using reverse transmittance nerve network
It is difficult to establish the path loss model changed with multi-parameter.
Step 306: using test sample to the wireless channel large-scale fading model based on reverse transmittance nerve network into
Row verifying.
Step 4: simultaneously, according to default test scene and intelligent algorithm model, combination is obtained under default test scene
Wireless channel multipath fading model；
Multipath fading model is based on radial basis function neural network algorithm, as shown in Figure 4, the specific steps are as follows:
Step 401: converting radial basis function mind for the radio wave measurement data of default scene and other related datas
Through data format required for network modelling；
Step 402: selection radial basis function neural network implies the number of plies, every layer of neuron or interstitial content；
Specifically, in practical applications, radial primary function network is by up of three-layer.Input layer transmits input signal,
Hidden node is described by gaussian kernel function (radial basis function), and is exported node layer and usually described by simple linear function.It is hidden
The excitation function (kernel function) of layer neuron (sension unit) locally generates response to input signal, i.e., when input signal is close to core
When the central range of function, hidden node will generate biggish output.
Step 403: the data format after the radio wave measurement data of default scene and conversion is divided into training at random
Sample and test sample；
Step 404: training sample being pre-processed using conventional high resolution algorithm；It can be based on letter to each sample
The statistical distribution of road parameter is randomly generated；
Step 405: pretreated training sample and channel parameter are inputted, radial basis function neural network model is trained,
Obtain the weight and input bias of each layer neuron interconnections needed in radial basis function neural network model；
Channel parameter includes: mobile station speed, time-domain sampling interval, time delay and angle spread, Rice factor etc.,
As shown in figure 5, radial basis function neural network model and mimo channel model structure are completely the same, first layer is
MIMO input layer, the last layer are MIMO output layer, and it includes multiple neuron sections in hidden layer that centre, which is neural network hidden layer,
Point.It is connected with each other between each neuron node and (is different from reverse transmittance nerve network), the upper complete and MIMO letter of connection in space
As connection between road.Input layer x1,x2...xnCorresponding MIMO tests every a line of training sample matrix and believes with tradition
The relevant channel parameter of road model, such as root mean square time delay and angle spread, Rice factor, Doppler frequency shift, antenna polarization identification
Degree etc. establishes the wireless channel model based on radial basis function neural network by the input of MIMO training sample, determines input
With first layer neuron, the models ginseng such as output and the weight W that is respectively linked between the last layer neuron and hidden layer neuron
Number.
Step 406: inclined according to trained radial basis function neural network model, the weight of each layer neuron interconnections and input
Value is set, the wireless channel multipath fading model under default scene is obtained；
Step 407: using test sample to the wireless channel multipath fading based on radial basis function neural network algorithm
Model is verified.
Step 5: inputting the radio wave measurement data of different geographic regions, training nothing under identical default test scene
Line channel size scale fading model, directly obtains the wireless channel model of other geographic areas.
Step 6: simultaneously, using trained wireless channel size scale fading model, obtaining radio wave propagation characteristic with spy
Levy the basic law of Parameters variation；
Characteristic parameter includes time domain, frequency domain, airspace (influence of topography and geomorphology) and scene type etc.；
The radio wave propagation characteristic refers to the basic law that preset condition changes: the time domain changing rule of the characteristic of channel,
The frequency domain changing rule of the characteristic of channel, the airspace changing rule of the characteristic of channel, the characteristic of channel are advised with the variation of default scene type
Rule.
Step 7: according to training sample characteristic parameter, the basic law of variation and wireless channel size scale fading model,
Establish the wireless channel model based on characteristic parameter.
Claims (6)
1. a kind of Wireless Channel Modeling method based on intelligent algorithm, which is characterized in that specific step is as follows:
Step 1: acquiring all kinds of radio wave measurement data for being used for Wireless Channel Modeling under default test scene；
Step 2: utilizing all kinds of radio wave measurement data training intelligent algorithm model of acquisition；
Intelligent algorithm model includes reverse transmittance nerve network algorithm and radial basis function neural network algorithm；
Step 3: combination obtains the wireless communication under default test scene according to default test scene and intelligent algorithm model
Road large-scale fading model；
Step 4: simultaneously, according to default test scene and intelligent algorithm model, combination obtains the nothing under default test scene
Line channel multipath fading model；
Step 5: inputting the radio wave measurement data of different geographic regions, training wireless communication under identical default test scene
Road size scale fading model, directly obtains the wireless channel model of other geographic areas；
Step 6: simultaneously, using trained wireless channel size scale fading model, obtaining radio wave propagation characteristic and joining with feature
The basic law of number variation；
The radio wave propagation characteristic refers to the basic law that preset condition changes: time domain changing rule, the channel of the characteristic of channel
The frequency domain changing rule of characteristic, the airspace changing rule of the characteristic of channel, the characteristic of channel with default scene type changing rule；
Step 7: being established according to training sample, basic law and wireless channel size scale fading model based on characteristic parameter
Wireless channel model.
2. a kind of Wireless Channel Modeling method based on intelligent algorithm as described in claim 1, which is characterized in that step
Data described in a kind of include radio wave measurement data, terrain data and hydrometeorological data.
3. a kind of Wireless Channel Modeling method based on intelligent algorithm as described in claim 1, which is characterized in that step
Large-scale fading model described in three is based on reverse transmittance nerve network algorithm, the specific steps are as follows:
Step 301: converting the radio wave measurement data of default test scene to needed for reverse transmittance nerve network modeling
Data format；
Step 302: selection reverse transmittance nerve network implies the number of plies, every layer of neuron or interstitial content；
Step 303: the randomizing data after conversion format is divided into training sample and test sample；
Step 304: inputting training sample and train back propagation artificial neural network model with Parameters in Mathematical Model, obtain backpropagation
The weight of required each layer neuron interconnections and input bias in neural network model；
Parameters in Mathematical Model includes: reception and transmission range, time-domain sampling interval and reception level；
Step 305: according to the weight of each layer neuron interconnections and input bias, obtaining the big ruler of wireless channel under default scene
Spend fading model；
Step 306: the wireless channel large-scale fading model based on reverse transmittance nerve network being tested using test sample
Card.
4. a kind of Wireless Channel Modeling method based on intelligent algorithm as described in claim 1, which is characterized in that step
Multipath fading model described in four is based on radial basis function neural network algorithm, the specific steps are as follows:
Step 401: converting the radio wave measurement data of default scene to required for radial basis function neural network modeling
Data format；
Step 402: selection radial basis function neural network implies the number of plies, every layer of neuron or interstitial content；
Step 403: the data format after conversion is divided into training sample and test sample at random；
Step 404: training sample being pre-processed using conventional high resolution algorithm；
Step 405: inputting pretreated training sample and channel parameter, training radial basis function neural network model obtains
The weight and input bias of each layer neuron interconnections needed in radial basis function neural network model；
Step 406: according to the weight of each layer neuron interconnections and input bias, obtaining the small ruler of wireless channel under default scene
Spend fading model；
Step 407: using test sample to the wireless channel multipath fading model based on radial basis function neural network algorithm
It is verified.
5. a kind of Wireless Channel Modeling method based on intelligent algorithm as claimed in claim 4, which is characterized in that step
Channel parameter described in 405 includes: mobile station speed, time-domain sampling interval, time delay and angle spread and Rice factor.
6. a kind of Wireless Channel Modeling method based on intelligent algorithm as described in claim 1, which is characterized in that step
Characteristic parameter described in six includes time domain, frequency domain, airspace and scene type.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910343878.8A CN110113119A (en) | 2019-04-26 | 2019-04-26 | A kind of Wireless Channel Modeling method based on intelligent algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910343878.8A CN110113119A (en) | 2019-04-26 | 2019-04-26 | A kind of Wireless Channel Modeling method based on intelligent algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110113119A true CN110113119A (en) | 2019-08-09 |
Family
ID=67486863
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910343878.8A Pending CN110113119A (en) | 2019-04-26 | 2019-04-26 | A kind of Wireless Channel Modeling method based on intelligent algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110113119A (en) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111212003A (en) * | 2020-02-14 | 2020-05-29 | 南京工程学院 | Channel root mean square delay spread prediction method suitable for intelligent medical treatment |
| CN111313987A (en) * | 2020-02-14 | 2020-06-19 | 南京工程学院 | Neural network-based path loss prediction method suitable for smart home |
| CN113037668A (en) * | 2021-05-20 | 2021-06-25 | 武汉科技大学 | Millimeter wave point-to-point communication channel equalization method |
| CN113055111A (en) * | 2021-03-04 | 2021-06-29 | 武汉科技大学 | Channel modeling method and system based on Bayesian optimization |
| WO2022156469A1 (en) * | 2021-01-22 | 2022-07-28 | 中兴通讯股份有限公司 | Channel scene recognition method, network device, and storage medium |
| CN115189791A (en) * | 2022-09-08 | 2022-10-14 | 浙江万胜智能科技股份有限公司 | Method and system for improving data accuracy of communication module of intelligent electric meter |
| WO2023005746A1 (en) * | 2021-07-26 | 2023-02-02 | 中兴通讯股份有限公司 | Channel modeling method and apparatus, storage medium, and electronic apparatus |
| WO2023020203A1 (en) * | 2021-08-17 | 2023-02-23 | 中兴通讯股份有限公司 | Wireless channel modeling method and apparatus, electronic device, and storage medium |
| WO2023051528A1 (en) * | 2021-09-30 | 2023-04-06 | 中兴通讯股份有限公司 | Channel scenario recognition method and apparatus, electronic device, and storage medium |
| WO2023061303A1 (en) * | 2021-10-13 | 2023-04-20 | 中兴通讯股份有限公司 | Large-scale fading modeling and estimation method, system, and device, and storage medium |
| WO2024001326A1 (en) * | 2022-06-28 | 2024-01-04 | 中兴通讯股份有限公司 | Physical channel simulation method, device, channel simulation system, and storage medium |
| CN117471575A (en) * | 2023-12-28 | 2024-01-30 | 河海大学 | Typhoon wave height forecasting method based on BO-LSTM neural network model |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6947891B2 (en) * | 2000-06-26 | 2005-09-20 | Korea Advanced Institute Of Science & Technology | Efficient speech recognition system bases on an auditory model |
| EP2800401A1 (en) * | 2013-04-29 | 2014-11-05 | Thomson Licensing | Method and Apparatus for compressing and decompressing a Higher Order Ambisonics representation |
| CN108512621A (en) * | 2018-03-02 | 2018-09-07 | 东南大学 | A kind of Wireless Channel Modeling method based on neural network |
| CN108566257A (en) * | 2018-04-27 | 2018-09-21 | 电子科技大学 | A kind of signal recovery method based on reverse transmittance nerve network |
| CN108599809A (en) * | 2018-03-14 | 2018-09-28 | 中国信息通信研究院 | Full duplex self-interference signal number removing method and device |
| CN108763848A (en) * | 2018-02-10 | 2018-11-06 | 江西航天经纬化工有限公司 | A kind of mechanical properties of propellant prediction technique based on BP artificial neural networks |
| CN109450830A (en) * | 2018-12-26 | 2019-03-08 | 重庆大学 | Channel estimation methods based on deep learning under a kind of high-speed mobile environment |
| CN109660287A (en) * | 2018-12-10 | 2019-04-19 | 深圳大学 | A kind of antenna selecting method based on deep learning |
-
2019
- 2019-04-26 CN CN201910343878.8A patent/CN110113119A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6947891B2 (en) * | 2000-06-26 | 2005-09-20 | Korea Advanced Institute Of Science & Technology | Efficient speech recognition system bases on an auditory model |
| EP2800401A1 (en) * | 2013-04-29 | 2014-11-05 | Thomson Licensing | Method and Apparatus for compressing and decompressing a Higher Order Ambisonics representation |
| CN108763848A (en) * | 2018-02-10 | 2018-11-06 | 江西航天经纬化工有限公司 | A kind of mechanical properties of propellant prediction technique based on BP artificial neural networks |
| CN108512621A (en) * | 2018-03-02 | 2018-09-07 | 东南大学 | A kind of Wireless Channel Modeling method based on neural network |
| CN108599809A (en) * | 2018-03-14 | 2018-09-28 | 中国信息通信研究院 | Full duplex self-interference signal number removing method and device |
| CN108566257A (en) * | 2018-04-27 | 2018-09-21 | 电子科技大学 | A kind of signal recovery method based on reverse transmittance nerve network |
| CN109660287A (en) * | 2018-12-10 | 2019-04-19 | 深圳大学 | A kind of antenna selecting method based on deep learning |
| CN109450830A (en) * | 2018-12-26 | 2019-03-08 | 重庆大学 | Channel estimation methods based on deep learning under a kind of high-speed mobile environment |
Non-Patent Citations (2)
| Title |
|---|
| JINGWEN TIAN 等: "Research of Wireless Fading Channel Modeling Based on Radial Basis Function Network", 《2009 5TH INTERNATIONAL CONFERENCE ON WIRELESS COMMUNICATIONS, NETWORKING AND MOBILE COMPUTING》 * |
| 李树: "第五代移动通信参数化信道建模与仿真技术研究", 《中国博士学位论文全文库 信息科技辑》 * |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111313987A (en) * | 2020-02-14 | 2020-06-19 | 南京工程学院 | Neural network-based path loss prediction method suitable for smart home |
| CN111313987B (en) * | 2020-02-14 | 2021-11-02 | 南京工程学院 | Neural network-based path loss prediction method suitable for smart home |
| CN111212003A (en) * | 2020-02-14 | 2020-05-29 | 南京工程学院 | Channel root mean square delay spread prediction method suitable for intelligent medical treatment |
| WO2022156469A1 (en) * | 2021-01-22 | 2022-07-28 | 中兴通讯股份有限公司 | Channel scene recognition method, network device, and storage medium |
| CN113055111A (en) * | 2021-03-04 | 2021-06-29 | 武汉科技大学 | Channel modeling method and system based on Bayesian optimization |
| CN113037668A (en) * | 2021-05-20 | 2021-06-25 | 武汉科技大学 | Millimeter wave point-to-point communication channel equalization method |
| WO2023005746A1 (en) * | 2021-07-26 | 2023-02-02 | 中兴通讯股份有限公司 | Channel modeling method and apparatus, storage medium, and electronic apparatus |
| WO2023020203A1 (en) * | 2021-08-17 | 2023-02-23 | 中兴通讯股份有限公司 | Wireless channel modeling method and apparatus, electronic device, and storage medium |
| WO2023051528A1 (en) * | 2021-09-30 | 2023-04-06 | 中兴通讯股份有限公司 | Channel scenario recognition method and apparatus, electronic device, and storage medium |
| WO2023061303A1 (en) * | 2021-10-13 | 2023-04-20 | 中兴通讯股份有限公司 | Large-scale fading modeling and estimation method, system, and device, and storage medium |
| WO2024001326A1 (en) * | 2022-06-28 | 2024-01-04 | 中兴通讯股份有限公司 | Physical channel simulation method, device, channel simulation system, and storage medium |
| CN115189791B (en) * | 2022-09-08 | 2023-01-13 | 浙江万胜智能科技股份有限公司 | Method and system for improving data accuracy of communication module of intelligent electric meter |
| CN115189791A (en) * | 2022-09-08 | 2022-10-14 | 浙江万胜智能科技股份有限公司 | Method and system for improving data accuracy of communication module of intelligent electric meter |
| CN117471575A (en) * | 2023-12-28 | 2024-01-30 | 河海大学 | Typhoon wave height forecasting method based on BO-LSTM neural network model |
| CN117471575B (en) * | 2023-12-28 | 2024-03-08 | 河海大学 | Typhoon wave height forecasting method based on BO-LSTM neural network model |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110113119A (en) | A kind of Wireless Channel Modeling method based on intelligent algorithm | |
| Wang et al. | 6G wireless channel measurements and models: Trends and challenges | |
| CN110418354A (en) | It is a kind of that propagation model wireless network planning method is exempted from based on machine learning | |
| CN108696932A (en) | It is a kind of using CSI multipaths and the outdoor fingerprint positioning method of machine learning | |
| Cheng et al. | CNN-based mmWave path loss modeling for fixed wireless access in suburban scenarios | |
| CN104618045B (en) | Collected data-based wireless channel transmission model establishing method and system | |
| CN111062466B (en) | Method for predicting field intensity distribution of cell after antenna adjustment based on parameters and neural network | |
| Zhao et al. | Neural network and GBSM based time-varying and stochastic channel modeling for 5G millimeter wave communications | |
| CN112051576B (en) | Intelligent multi-frequency microwave rainfall monitoring method | |
| CN105871486B (en) | The channel model construction method and emulation mode of wireless sensor network | |
| Lee et al. | Path loss exponent prediction for outdoor millimeter wave channels through deep learning | |
| CN110247719A (en) | The playback of 5G time varying channel and emulation mode based on machine learning | |
| CN108540987B (en) | LTE network coverage state evaluation method and device | |
| CN109068267A (en) | A kind of indoor orientation method based on LoRa SX1280 | |
| CN104039008B (en) | A kind of hybrid locating method | |
| CN114630348A (en) | Base station antenna parameter adjusting method and device, electronic equipment and storage medium | |
| Kim et al. | mmWave path loss modeling for urban scenarios based on 3D-convolutional neural networks | |
| Féral et al. | Large-scale modeling of rain fields from a rain cell deterministic model | |
| Fernández Anitzine et al. | Influence of training set selection in artificial neural network-based propagation path loss predictions | |
| CN112114386B (en) | High-time-space resolution microwave rainfall monitoring method | |
| CN111817806B (en) | Automobile air interface communication performance test method and system | |
| Yi et al. | Full-wave simulation and scattering modeling for terahertz communications | |
| CN107205237A (en) | The optimization method and device of a kind of network structure | |
| CN103391562A (en) | Method and device for achieving network testing based on uplink interference signals | |
| Song et al. | Air-to-ground large-scale channel characterization by ray tracing |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190809 |