CN107451605A - A kind of simple target recognition methods based on channel condition information and SVMs - Google Patents
A kind of simple target recognition methods based on channel condition information and SVMs Download PDFInfo
- Publication number
- CN107451605A CN107451605A CN201710568384.0A CN201710568384A CN107451605A CN 107451605 A CN107451605 A CN 107451605A CN 201710568384 A CN201710568384 A CN 201710568384A CN 107451605 A CN107451605 A CN 107451605A
- Authority
- CN
- China
- Prior art keywords
- csi
- data
- target
- svm
- simple target
- 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
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B17/00—Monitoring; Testing
- H04B17/30—Monitoring; Testing of propagation channels
- H04B17/309—Measuring or estimating channel quality parameters
- H04B17/345—Interference values
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computer Networks & Wireless Communication (AREA)
- Life Sciences & Earth Sciences (AREA)
- Electromagnetism (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Signal Processing (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Quality & Reliability (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
A kind of simple target recognition methods based on channel condition information and SVMs proposed by the present invention, it is not necessary to build special hardware facility, make full use of prior wireless network, the function of simple target identification is reached using common commercial router can.After CSI initial data is obtained, use density-based algorithms DBSCAN to cluster the subcarrier data in channel first with denoising, then the data after denoising are carried out using the moving average algorithm based on weights smooth.After data prediction, the present invention carries out characteristics extraction using Principal Component Analysis Algorithm to data.Data after pretreatment and feature extraction can more accurately reflect the Main change of signal and dimension substantially reduces, and are favorably improved target identification precision and reduce computation complexity.The present invention is by means of the SVM multi-classification algorithms based on one against one strategies, the statistical model of non-linear dependence between acquisition destination object and received signals fingerprint, so as to reach the purpose of simple target identification.
Description
Technical field
The present invention relates to field of target recognition, more particularly to one kind based on channel condition information and to use SVMs skill
Art carries out simple target and knows method for distinguishing.
Background technology
WLAN based on Wi-Fi obtains widespread deployment indoors, while data transport service is provided, may be used also
Simple target identification service is provided.The proportion that water in human body accounts for is 70 percent, and water is to have very to radio frequency signal
Strong absorbability, so people knows from experience produces reflection, scattering, diffraction, decay and other effects to the Wi-Fi signal of surrounding, pass through
The special fingerprint characteristic that monitoring human body can be formed to the interference caused by Wi-Fi signal, whether can be that people carries out letter to target
Single target identification.
Received signal strength indicator (Received Signal Strength can be obtained from Wi-Fi signal
Indicator, RSSI) and channel condition information (Channel State Information, CSI).RSSI is current using most
Extensive energy response, but the accurate perception under its coarseness and variable sexual incompatibility multi-path indoor environment, for target identification
Precision is very poor.CSI is physical layer attributes, the decay factor that description signal is propagated between transmitter and receiver, including scattering,
The information such as environmental attenuation, range attenuation, the interference of the narrow-band signal from frequency range can be resisted, it is steady enough in a static environment
It is fixed, it can be made a response immediately when disturbed, and the signal from mulitpath can be differentiated, multipath effect influences small.With just
Handing over frequency division multiplexing (Orthogonal Frequency Division Multiplexing, OFDM), technology is in a wireless local area network
Application, can more easily obtain CSI at present.CSI includes the amplitude and phase information of each subcarrier, using the teaching of the invention it is possible to provide abundant
Frequency domain information, so as to improve the accuracy of target identification.
On the process nature of simple target identification and a process classified, interference of the CSI signals for human body show
Substantially, there is good identification, it is possible to solve the problem of simple target identifies using the thought of classification.SVMs
(Support Vector Machines, SVM) is a kind of machine learning method based on Statistical Learning Theory, is solving higher-dimension
Many advantages are shown with terms of nonlinear problem, by means of svm classifier, can be obtained non-between identification target and received signals fingerprint
The statistical model of linear dependence.
The content of the invention
The present invention is a kind of simple target identification classified based on channel condition information (CSI) and SVMs (SVM)
Method, comprise the following steps：
Step 1：Environment is disposed, and the simple target identification based on Wi-Fi requires in-door covering Wi-Fi signal, selection signal
Less 5G frequency ranges are disturbed, equipment is two notebook computers, is respectively arranged with Intel link 5300agn wireless business network interface cards；
Step 2：CSI raw data acquisitions, some CSI initial data of different target object are gathered, including：Send day
Line number, reception antenna number, transmission frequency, channel condition information CSI matrixes；
Step 3：CSI data predictions, including：(1) dimension of CSI matrixes first in initial data is removed, will be produced
Two-dimensional matrix be transformed into logarithm (power) space from linear (level) space, by each complex conversion in matrix into value；
(2) there are 30 subcarriers in every a pair channels for sending and receiving antenna composition, using density-based algorithms
Density-Based Spatial Clustering of Applications with Noise (DBSCAN) are believed each
Road is clustered, by deleting outlier denoising；(3) using the moving average algorithm based on weights to the CSI data after denoising
Carry out smooth；
Step 4：CSI characteristics extractions, use principal component analysis (Principal Component Analysis, PCA)
Algorithm carries out dimensionality reduction and characteristics extraction to pretreated CSI data, produces CSI sample fingerprints；
Step 5：SVM model trainings；Step is as follows：(1) CSI sample fingerprints are normalized；(2) CSI fingerprint samples are based on
This, establishes the more disaggregated models of SVM, the corresponding target of each classification；
Step 6：Simple target identifies that step is as follows：(1) CSI raw data acquisitions are carried out according to step 2；(2) according to
Step 3 carries out CSI data predictions；(3) CSI Data Dimensionality Reductions and characteristics extraction are carried out according to step 4, obtains live signal
Fingerprint；(4) determine the CSI fingerprints measured in real time according to the voting results of the more disaggregated models of SVM representated by target, realize letter
Single goal identifies；
Brief description of the drawings
Fig. 1 is the simple target recognition methods flow chart based on CSI and SVM；
Fig. 2 is the environment deployment diagram of the simple target identification based on CSI；
Embodiment
The embodiment of the present invention is described below in conjunction with the accompanying drawings, so as to those skilled in the art preferably
Understand the present invention.Requiring particular attention is that in the following description, may desalination and ignore it is relevant with the present invention
Know the content introduction of function and design.
In the present embodiment, the present invention mainly includes following link to simple target recognition methods：Data acquisition, data
Pretreatment, the feature extraction of data, simple target Classification and Identification, flow chart is as shown in figure 1, specific implementation step is as follows：
Step 1：Environment is disposed, and the simple target identification based on Wi-Fi requires in-door covering Wi-Fi signal, experiment scene
A length of 6 meters of layout, a width of 7 meters, the less 5G frequency ranges of Systematic selection signal interference, equipment is two notebook computers, model
It is the THINKPAD of association T400 and X201i respectively, they are equipped with intel link 5300agn wireless business network interface cards, should
Network interface card has 3 antennas.T400 sets directive sending to broadcast bag data, the equipment as signal is sent, and X201i is set to monitor mould
Formula, the equipment as signal is received, layout type schematic diagram is referring to Fig. 2.
Step 2：CSI raw data acquisitions, the training stage, 4 kinds of scenes will be divided into gather the sample of training data, point
It is not dummy, male laboratory technician and the female laboratory technician that nobody, foam are done.Make target static per second in position, X201i shown in Fig. 2
20 CSI initial data from T400 are gathered, including：Transmission antenna number Ntx, reception antenna number Nrx, packet transmission
Frequency f, channel state information matrix H.Channel condition information H is a Ntx×Nrx× 30 three-dimensional matrice, the third dimension are OFDM
30 subcarriers information h=in channel | h | ejsinθ, | h | it is subcarrier amplitude, θ is sub-carrier phase.
Step 3：CSI data generate, and for the CSI initial data of collection, remove CSI matrix Hs first first ties up
Degree, obtain NtxIndividual Nrx× 30 two-dimensional matrix, two-dimensional matrix is converted into logarithm (power) space from linear (level) space,
And by each complex conversion in matrix into value.
Step 4：CSI data de-noisings, send and receive antenna for every a pair and form a channel, therefore a pair of AP-MP are included
Ntx×NrxBar channel；Every channel includes 30 subcarriers, therefore a pair of AP-MP include Ntx×Nrx× 30 subcarriers.According to
CSI data sets are divided into N by channeltx×NrxIndividual Sub Data Set, each Sub Data Set include 30 subcarriers information.Make subdata
Serial number index (1~30), the amplitude value per subcarriers are concentrated, is applied on each Sub Data Set based on density
Clustering algorithm DBSCAN is clustered.Two parameters in DBSCAN are that field radius e and minimum include points minOpt respectively,
Its sorting procedure includes：(1) it is non-access state, i.e. " unvisited " by all object tags in Sub Data Set；(2) it is random
Selection one does not access object o (index, value), labeled as " visited "；Check whether o neighborhood comprises at least
MinOpt object：If it is not, then mark o is outlier；If it is, create a new cluster C and a Candidate Set for o
N is closed, all objects in o neighborhood are placed in candidate collection N；(3) DBSCAN is iteratively the object that other clusters are not belonging in N
It is added in C, until N is sky, cluster C is completed；(4) go to step (2) and handle next object；(5) outlier will be marked as
Sample data is deleted from training set corresponding to object, reaches the purpose of data de-noising.
Step 5：CSI data smoothings, the CSI data after denoising are put down using the moving average algorithm based on weights
Slide to reduce data fluctuations.Assuming that a CSI subcarriers sequence to moment t is (v1, v2,…,vt), then in moment t CSI
Smooth valueIt is the weighted average of preceding m value, m is sliding window size：
Step 6：CSI characteristics extractions, (1) assume N be presentapAP-MP is combined, each pair AP-MP includes Ntx×NrxBar
Channel, every channel have 30 subcarriers, then the dimension of every CSI data is Nap×Ntx×Nrx× 30, dimension is very high.CSI
Each contribution of the value to classification is different in data, therefore carries out dimensionality reduction simultaneously to CSI data using principal component analysis PCA algorithms
Extract maximally effective feature.PCA target is to find r (r<Nap×Ntx×Nrx× 30) individual new feature, each new feature are original
The linear combination of feature, new feature can reflect the principal character of legacy data, and can compress the scale of legacy data.(2) it is false
If training sample set matrix is X, sample size N.By PCA, transition matrix C can be obtained, sample matrix X is by conversion
Matrix S afterwards, and the feature weight L=(L arranged in descending order1,L2,…,Ln), n is characterized dimension.Calculate feature LrAccumulation
Contribution rate is：
If Lr>95%, i.e., the accumulation contribution rate of preceding r feature is more than predetermined threshold value 95%, then takes L=(L1,L2,…,Ln) in
Preceding r feature F=(L1,L2,…,Lr) as the feature extracted.Preceding r row in matrix S form principal component matrix R.R conducts
The sample data of following model training.
Step 7：CSI sample fingerprints generate and normalization, it is assumed that the CSI sample sets of detection zone are X={ c1,
c2,…,cN, N is number of samples；ci={ ci1,ci2,…,cirThe sample after feature extraction is represented, r is sample dimension, normalizing
Change obtains：
cijRepresent sample ciJ-th of characteristic value, cminRepresent the minimum value of all characteristic values, cmaxRepresent all characteristic values
Maximum；
Step 8：Svm classifier model training, svm classifier model training is carried out to each target.Due to multiple identification mesh be present
Mark, therefore multiple target classification is assorting process more than one, is expanded to SVM from two classification using one-against-one strategies
More classification, finally carry out ballot and obtain final classification result.Assuming that destination number is K, the corresponding class of each target, then
For the quantity of class for K, it is necessary to train K (K-1)/2 grader, each grader carries out two classification to two targets.Assuming that
ωiAnd ωjTarget d is represented respectivelyiWith target djCorresponding class, then for class ωiWith class ωjThe training of SVM classifier be exactly
Using from ωiAnd ωjIn CSI sample fingerprints solve following problem：
Wherein C is constant, and ξ is one group of slack variable, ctIt is a CSI sample fingerprint, N is number of samples.
Step 9：Online target identification, when realistic objective identifies, algorithm passes through training according to the CSI fingerprints gathered in real time
Obtained SVM classifier determines the classification of target.Comprise the following steps：(1) mode of step 2 gathers CSI initial data；(2)
CSI data are pre-processed in the way of step 3 to step 6 and feature extraction；(3) produced in the way of step 7
CSI fingerprints simultaneously normalize；(4) representated by determining the CSI fingerprints measured in real time according to the voting results of multiple svm classifier models
Target.
The beneficial effects of the invention are as follows：Simple target identification solution based on CSI need not build special hardware
Facility, prior wireless network is made full use of, the function of simple target identification is reached using common commercial router can.Obtaining
After taking CSI initial data, the present invention pre-processes to data, first using density-based algorithms DBSCAN to channel
In subcarrier data clustered with denoising, then the data after denoising are carried out using the moving average algorithm based on weights
Smoothly.After data prediction, the present invention carries out characteristics extraction using Principal Component Analysis Algorithm to data.Pretreatment and feature carry
Data after taking can more accurately reflect the Main change of signal and dimension substantially reduces, and are favorably improved target identification
Precision simultaneously reduces computation complexity.The present invention is obtained by means of the SVM multi-classification algorithms based on one-against-one strategies
The statistical model of non-linear dependence between destination object and received signals fingerprint, so as to reach the purpose of simple target identification.This
Invention can reach the precision of more than 98.7% target identification.
Although the illustrative embodiment of the present invention is described above, in order to the technology people of this technology neck
Member understands the present invention, it should be apparent that the invention is not restricted to the scope of embodiment, to the ordinary skill of the art
For personnel, as long as various change, in the spirit and scope of the present invention that appended claim limits and determines, these become
Change is it will be apparent that all utilize the innovation and creation of present inventive concept in the row of protection.
Claims (2)
1. the present invention is a kind of simple target identification side classified based on channel condition information (CSI) and SVMs (SVM)
Method, mainly including herein below：The collection and pretreatment of CSI data, CSI data characteristicses are extracted, to extracting feature off-line training
Multiple SVM models, simple target ONLINE RECOGNITION stage.
Technical scheme is as follows：
Step 1：Simple target identification based on Wi-Fi requires in-door covering Wi-Fi signal, and selection signal disturbs less 5G
Frequency range, equipment are two notebook computers, are respectively arranged with Intel link 5300agn wireless business network interface cards；
Step 2：Some CSI initial data of different target object are gathered, including：Transmission antenna number, reception antenna number,
Transmission frequency, channel condition information CSI matrixes；
Step 3：CSI data predictions, including：(1) dimension of CSI matrixes first in initial data is removed, two by caused by
Dimension matrix is transformed into logarithm (power) space from linear (level) space, by each complex conversion in matrix into value；(2) it is every
There are 30 subcarriers in a pair of channels for sending and receiving antenna composition, using density-based algorithms Density-
Based Spatial Clustering of Applications with Noise (DBSCAN) gather to each channel
Class, by deleting outlier denoising；(3) the CSI data after denoising are carried out using the moving average algorithm based on weights smooth；
Step 4：CSI characteristics extractions, using principal component analysis (PCA) algorithm to pretreated CSI data carry out dimensionality reduction and
Characteristics extraction, produce CSI sample fingerprints；
Step 5：SVM model trainings, step are as follows：(1) CSI sample fingerprints are normalized；(2) CSI sample fingerprints are based on, are built
The more disaggregated models of vertical SVM, the corresponding target of each classification；
Step 6：Simple target identifies that step is as follows：(1) CSI raw data acquisitions are carried out according to step 2；(2) according to step
Three carry out CSI data predictions；(3) CSI Data Dimensionality Reductions and characteristics extraction are carried out according to step 4, obtains live signal and refer to
Line；(4) determine the CSI fingerprints measured in real time according to the voting results of the more disaggregated models of SVM representated by target.
A kind of 2. simple target classified based on channel condition information CSI and support vector machines according to claim 1
Recognition methods, its feature are (1) data prediction, and each channel is entered using density-based algorithms (DBSCAN)
Row cluster, deletes outlier denoising；(2) CSI Data Dimensionality Reductions and characteristics extraction, using principal component analysis (PCA) algorithm to pre-
CSI data after processing carry out dimensionality reduction and obtain CSI sample fingerprints；(3) by means of the SVM based on one-against-one strategies
Multi-classification algorithm, the statistical model of non-linear dependence between target and received signals fingerprint is obtained, reach the purpose of target identification.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710568384.0A CN107451605A (en) | 2017-07-13 | 2017-07-13 | A kind of simple target recognition methods based on channel condition information and SVMs |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710568384.0A CN107451605A (en) | 2017-07-13 | 2017-07-13 | A kind of simple target recognition methods based on channel condition information and SVMs |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107451605A true CN107451605A (en) | 2017-12-08 |
Family
ID=60488564
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710568384.0A Pending CN107451605A (en) | 2017-07-13 | 2017-07-13 | A kind of simple target recognition methods based on channel condition information and SVMs |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107451605A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109061600A (en) * | 2018-09-28 | 2018-12-21 | 上海市刑事科学技术研究院 | A kind of target identification method based on millimetre-wave radar data |
| CN109657572A (en) * | 2018-12-04 | 2019-04-19 | 重庆邮电大学 | Goal behavior recognition methods after a kind of wall based on Wi-Fi |
| CN110730473A (en) * | 2019-09-03 | 2020-01-24 | 中国人民解放军陆军工程大学 | Signal feature extraction method for WiFi activity recognition |
| CN111262637A (en) * | 2020-01-15 | 2020-06-09 | 湖南工商大学 | Human body behavior identification method based on Wi-Fi channel state information CSI |
| CN111836278A (en) * | 2019-08-09 | 2020-10-27 | 维沃移动通信有限公司 | Measurement reporting and processing method, equipment and medium in secondary cell activation |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20120159527A1 (en) * | 2010-12-16 | 2012-06-21 | Microsoft Corporation | Simulated group interaction with multimedia content |
| CN104619014A (en) * | 2015-01-09 | 2015-05-13 | 中山大学 | SVM-KNN (Support Vector Machine-K Nearest Neighbor)-based indoor positioning method |
| CN105721086A (en) * | 2016-03-11 | 2016-06-29 | 重庆科技学院 | Wireless channel scene recognition method based on unscented Kalman filter artificial neural network (UKFNN) |
| CN106131958A (en) * | 2016-08-09 | 2016-11-16 | 电子科技大学 | A kind of based on channel condition information with the indoor Passive Location of support vector machine |
| CN106792560A (en) * | 2016-12-30 | 2017-05-31 | 北京理工大学 | Target identification method based on wireless reception of signals intensity |
-
2017
- 2017-07-13 CN CN201710568384.0A patent/CN107451605A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20120159527A1 (en) * | 2010-12-16 | 2012-06-21 | Microsoft Corporation | Simulated group interaction with multimedia content |
| CN104619014A (en) * | 2015-01-09 | 2015-05-13 | 中山大学 | SVM-KNN (Support Vector Machine-K Nearest Neighbor)-based indoor positioning method |
| CN105721086A (en) * | 2016-03-11 | 2016-06-29 | 重庆科技学院 | Wireless channel scene recognition method based on unscented Kalman filter artificial neural network (UKFNN) |
| CN106131958A (en) * | 2016-08-09 | 2016-11-16 | 电子科技大学 | A kind of based on channel condition information with the indoor Passive Location of support vector machine |
| CN106792560A (en) * | 2016-12-30 | 2017-05-31 | 北京理工大学 | Target identification method based on wireless reception of signals intensity |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109061600A (en) * | 2018-09-28 | 2018-12-21 | 上海市刑事科学技术研究院 | A kind of target identification method based on millimetre-wave radar data |
| CN109657572A (en) * | 2018-12-04 | 2019-04-19 | 重庆邮电大学 | Goal behavior recognition methods after a kind of wall based on Wi-Fi |
| CN111836278A (en) * | 2019-08-09 | 2020-10-27 | 维沃移动通信有限公司 | Measurement reporting and processing method, equipment and medium in secondary cell activation |
| CN111836278B (en) * | 2019-08-09 | 2023-05-26 | 维沃移动通信有限公司 | Measurement reporting and processing method, device and medium in secondary cell activation |
| CN110730473A (en) * | 2019-09-03 | 2020-01-24 | 中国人民解放军陆军工程大学 | Signal feature extraction method for WiFi activity recognition |
| CN111262637A (en) * | 2020-01-15 | 2020-06-09 | 湖南工商大学 | Human body behavior identification method based on Wi-Fi channel state information CSI |
| CN111262637B (en) * | 2020-01-15 | 2021-09-28 | 湖南工商大学 | Human body behavior identification method based on Wi-Fi channel state information CSI |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107480699A (en) | A kind of intrusion detection method based on channel condition information and SVMs | |
| CN106131958A (en) | A kind of based on channel condition information with the indoor Passive Location of support vector machine | |
| Ahmadien et al. | Predicting path loss distribution of an area from satellite images using deep learning | |
| Li et al. | AF-DCGAN: Amplitude feature deep convolutional GAN for fingerprint construction in indoor localization systems | |
| CN107451605A (en) | A kind of simple target recognition methods based on channel condition information and SVMs | |
| CN105101408B (en) | Indoor orientation method based on distributed AP selection strategy | |
| CN107992882A (en) | A kind of occupancy statistical method based on WiFi channel condition informations and support vector machines | |
| Kim et al. | Towards deep learning-aided wireless channel estimation and channel state information feedback for 6G | |
| CN114359738B (en) | Cross-scene robust indoor people number wireless detection method and system | |
| CN112040400B (en) | Single-site indoor fingerprint positioning method based on MIMO-CSI, storage medium and equipment | |
| CN109299698B (en) | Wireless channel scene recognition method based on support vector machine | |
| CN104135327A (en) | Spectrum sensing method based on support vector machine | |
| CN110049549A (en) | More fusion indoor orientation methods and its system based on WiFi fingerprint | |
| Ding et al. | Device-free location-independent human activity recognition using transfer learning based on CNN | |
| CN110933628B (en) | Fingerprint indoor positioning method based on twin network | |
| CN109526021A (en) | A kind of classification method and recognition methods of high-speed railway wireless channel scene | |
| CN113037410B (en) | Channel identification method, device, transmission method, transmission equipment, base station and medium | |
| CN108122310A (en) | A kind of people flow rate statistical method based on WiFi channel state informations and dynamic time warping | |
| Tsai et al. | Refined autoencoder-based CSI hidden feature extraction for indoor spot localization | |
| Zhou et al. | Wiflowcount: Device-free people flow counting by exploiting doppler effect in commodity wifi | |
| CN109255874A (en) | A kind of passage and number detection method based on general commercial WiFi equipment | |
| Gu et al. | Device‐Free Human Activity Recognition Based on Dual‐Channel Transformer Using WiFi Signals | |
| Yuan et al. | Features extraction and analysis for device-free human activity recognition based on channel statement information in b5G wireless communications | |
| CN112867021A (en) | Improved TrAdaBoost-based indoor positioning method for transfer learning | |
| Wasilewska et al. | Artificial intelligence for radio communication context-awareness |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20171208 |
|
| WD01 | Invention patent application deemed withdrawn after publication |