CN107992882A - A kind of occupancy statistical method based on WiFi channel condition informations and support vector machines - Google Patents

A kind of occupancy statistical method based on WiFi channel condition informations and support vector machines Download PDF

Info

Publication number
CN107992882A
CN107992882A CN201711154554.7A CN201711154554A CN107992882A CN 107992882 A CN107992882 A CN 107992882A CN 201711154554 A CN201711154554 A CN 201711154554A CN 107992882 A CN107992882 A CN 107992882A
Authority
CN
China
Prior art keywords
csi
samples
occupancy
data
svm
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
Application number
CN201711154554.7A
Other languages
Chinese (zh)
Inventor
周瑞
鲁翔
赵浩森
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
University of Electronic Science and Technology of China
Original Assignee
University of Electronic Science and Technology of China
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by University of Electronic Science and Technology of China filed Critical University of Electronic Science and Technology of China
Priority to CN201711154554.7A priority Critical patent/CN107992882A/en
Publication of CN107992882A publication Critical patent/CN107992882A/en
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/23—Clustering techniques
    • G06F18/232—Non-hierarchical techniques
    • G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
    • 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
    • H04B7/00—Radio transmission systems, i.e. using radiation field
    • H04B7/02—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas
    • H04B7/04—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas
    • H04B7/06—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station
    • H04B7/0613—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station using simultaneous transmission
    • H04B7/0615—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station using simultaneous transmission of weighted versions of same signal
    • H04B7/0619—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station using simultaneous transmission of weighted versions of same signal using feedback from receiving side
    • H04B7/0621—Feedback content
    • H04B7/0626—Channel coefficients, e.g. channel state information [CSI]
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B7/00—Radio transmission systems, i.e. using radiation field
    • H04B7/02—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas
    • H04B7/04—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas
    • H04B7/06—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station
    • H04B7/0613—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station using simultaneous transmission
    • H04B7/0615—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station using simultaneous transmission of weighted versions of same signal
    • H04B7/0619—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station using simultaneous transmission of weighted versions of same signal using feedback from receiving side
    • H04B7/0658—Feedback reduction
    • H04B7/0663—Feedback reduction using vector or matrix manipulations

Landscapes

  • Engineering & Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Engineering & Computer Science (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • Artificial Intelligence (AREA)
  • General Physics & Mathematics (AREA)
  • Signal Processing (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Mathematical Physics (AREA)
  • Probability & Statistics with Applications (AREA)
  • Complex Calculations (AREA)
  • Image Analysis (AREA)

Abstract

The present invention proposes a kind of occupancy statistical method returned based on WiFi channel condition informations (CSI) and support vector machines (SVM).Special hardware facility need not be built, only utilizes existing WiFi wireless networks, it becomes possible to realizes that occupancy counts.After CSI data are obtained, the present invention carries out denoising using density-based algorithms DBSCAN to CSI data, then expandable matrix algorithm is used to obtain the non-zero rate of every subcarriers as CSI characteristic fingerprint samples, so as to the influence changed greatly for demographics of enhanced signal amplitude, and reduce the influence of the small change of the signal amplitude caused by ambient noise.The present invention can obtain accurate non-linear dependence model between number and CSI characteristic fingerprint samples, so as to reach the target of accurate statistics occupancy by means of SVM regression algorithms in the case of without considering complex indoor environment.The beneficial effects of the invention are as follows：It can be based on existing WiFi wireless networks and carry out accurate demographics, cost is low, and universality is strong, no privacy concern.

Description

It is a kind of to be counted based on WiFi channel condition informations and the occupancy of support vector machines Method
Technical field
The present invention relates to demographics field, more particularly to it is a kind of based on WiFi channel condition informations and support vector machines into The method of row occupancy statistics.
Background technology
Occupancy is for should much be used to say that critically important information, for example, businessman can be helped to determine into shop customer Quantity and arrange employee accordingly, can monitor public place crowd density and in time start emergency measure to ensure safety, can Adjusted with carrying out automatic air condition according to number in building to save energy etc..Traditional demographic method mainly uses video monitoring side Method or less radio-frequency method.Video frequency monitoring method has obtained extensive use at present, but under low light environment or non line of sight feelings Under condition, camera can not works fine, cause quality monitoring decline or monitoring blind area, also there are larger hidden for video monitoring in addition Private problem, is not suitable for personal air.Monitoring based on less radio-frequency also obtains more concern in recent years, such as uses wireless sensing Device network, infrared ray, ultra wide band etc., but these methods are required for installation and deployment task equipment, and of high cost, universality is poor.
Wireless network based on WiFi has obtained widespread deployment, it can be used for while wireless data transfer services are provided It is monitored and demographics.This method need not increase additional hardware facility, it is not required that personnel carry electronic equipment, only sharp With existing WiFi wireless networks with regard to that can complete demographics, cost is low, and universality is strong, therefore is great market prospects and development The solution of potentiality.Reflection, scattering, diffraction, decay and other effects are produced to the WiFi signal of surrounding since people knows from experience, pass through prison The change for surveying WiFi signal is assured that occupancy.The energy for being used to measure WiFi signal change most popular at present Characteristic is received signal strength indicator (Received Signal Strength Indicator, RSSI), can be used for carrying out Demographics.But due to the complexity of indoor environment, WiFi signal there are multipath effect, i.e., signal can by mulitpath from Transmitting terminal travels to receiving terminal, and has different delays, decay and phase shift per paths, and it is mulitpath letter to cause RSSI Number superposition, so as to cause RSSI unstable, precision is poor during for demographics.
The content of the invention
The purpose of the present invention is for above-mentioned problem, propose that one kind is based on WiFi channel condition informations (Channel State Information, CSI) and support vector machines (Support Vector Machines, SVM) accurate indoor people Number statistical method.
The present invention is for solution above-mentioned technical problem the technical scheme adopted is that a kind of be based on WiFi channel condition informations and support The occupancy statistical method of vector machine, comprises the following steps：
1) occupancy statistical model training step：
The CSI primary data samples under different number scenes 1-1) are gathered, the sample includes transmission antenna number, reception antenna Number, signal transmission frequency, channel condition information CSI matrixes, and record current persons count；
The first dimension of CSI matrixes in primary data sample 1-2) is removed, by the two-dimensional matrix of generation from linear (level) space Logarithm (power) space is transformed into, by each complex conversion in matrix into amplitude；
1-3) apply density-based algorithms (Density-Based Spatial Clustering of Applications with Noise, DBSCAN) to sending and receiving the CSI data in the channel that antenna forms by every a pair Clustered, the CSI data after denoising are obtained by deleting outlier；
Feature extraction 1-4) is carried out to the CSI data after denoising using expandable matrix algorithm, obtains the non-zero rate per subcarriers, As CSI characteristic fingerprint samples.The expandable matrix algorithm comprises the steps of：The amplitude sequence of every subcarriers is converted to one A two-dimensional matrix；Expand the two-dimensional matrix；Calculate the non-zero rate in the two-dimensional matrix after expansion；To all sons in all channels Carrier wave carries out aforesaid operations, obtains non-zero rate vector and is used as CSI characteristic fingerprint samples；
CSI characteristic fingerprint samples 1-5) are based on, are obtained using v-SVR (Support Vector Regression) Algorithm for Training Preliminary SVM regression models, express the dependence between number and CSI characteristic fingerprint samples；
Parameter optimization 1-6) is carried out to preliminary SVM regression models, is attempted using grid search and cross validation method more different Parameter combination, obtains optimal model parameters and optimal SVM regression models.
2) occupancy statistic procedure：
2-1) gather the CSI primary data samples under unknown number scene；
2-2) according to step 1-2), 1-3) and 1-4) CSI primary data samples are handled and feature extraction, obtain current CSI feature samples fingerprints；
2-2) by step 1-5) and SVM regression models 1-6) established, determined according to current CSI feature samples fingerprint current Number in scene.
Due to the multipath effect that interior is abundant so that the demographics based on WiFi RSSI cannot provide enough accuracies.Letter Channel state information CSI is a kind of finer physical layer information, describes the amplitude and phase of each subcarrier in channel, Neng Gouchong Divide and utilize indoor multipath effect, it is sensitiveer to environmental change.Therefore, the method for the invention based on CSI can be more accurate Ground carries out demographics.After CSI data are obtained, the present invention using density-based algorithms DBSCAN to CSI data into Row denoising, then the non-zero rate using the acquisition of expandable matrix algorithm per subcarriers is as CSI characteristic fingerprint samples, so as to strengthen The influence changed greatly for demographics of signal amplitude, and reduce the small change of the signal amplitude caused by ambient noise Influence.The present invention can be obtained number and CSI be special by means of SVM regression algorithms in the case of without considering complex indoor environment Accurate non-linear dependence model between sign sample fingerprint, so as to reach the target of accurate statistics occupancy.
The beneficial effects of the invention are as follows：It can be based on existing WiFi wireless networks and carry out accurate demographics, cost is low, pervasive Strong, the no privacy concern of property, indoor occupant can arbitrarily be walked about, stand or sat.The present invention can reach the essence of mean error number 0.67 Degree, hence it is evident that higher than the demographics precision based on RSSI.
Brief description of the drawings
Fig. 1 is the flow chart of the occupancy statistical method based on CSI and SVM；
Fig. 2 is the environment deployment diagram of the occupancy statistical method based on CSI and SVM.
Embodiment
A kind of occupancy statistical method based on WiFi channel condition informations CSI and support vector machines, flow is as schemed Shown in 1, specific implementation step is as follows：
1) environment is disposed：Demographics based on CSI require in-door covering WiFi signal, and equipment is an access point (Access Point, AP) and a monitoring point (Monitoring Point, MP), configure Intel Wireless Link 5300agn (IWL5300) wireless network card, the network interface card have 3 antennas.AP ends send data, and MP ends receive data, layout type such as Fig. 2 institutes Show.
2) CSI raw data acquisitions：Training stage, W people is set to by indoor scene, W=0, and 1,2,3,4,5 ..., allow him Middle random walk indoors.For the scene of different numbers, receiving terminal MP comes from transmitting terminal with the sample rate collection of 20Hz The CSI initial data of AP, gathers 5 minutes, 1000 conducts is then chosen from the 6000 CSI data samples collected respectively Training sample, to evade because jitter problem caused by environmental disturbances.Sample data includes：Transmission antenna number Ntx, reception antenna number Nrx, data packet transmission frequency f, original CSI matrix Hs.Original CSI matrix Hs are a Ntx×Nrx×Ns Three-dimensional matrice, the third dimension is the N in channelsSubcarriers information：H=| h | ejθ, wherein | h | it is subcarrier amplitude, θ is son Carrier phase.
3) CSI data generate：For the CSI initial data of collection, first dimension in original CSI matrix Hs is removed first, is obtained Obtain NtxA Nrx×NsTwo-dimensional matrix, two-dimensional matrix is then converted into logarithm (power) space from linear (level) space, and By each complex conversion in matrix into amplitude, CSI data are produced.
4) CSI data de-noisings：Due to there are certain noise, i.e., sending and receiving meeting in the data of antenna with a pair in CSI data There are the data that some deviate cluster, and therefore, in order to improve the precision of demographics, the present invention is calculated using density clustering Method DBSCAN removes the noise in CSI data.A channel is formed since every a pair sends and receives antenna, it is a pair of N is included in AP-MPtx×NrxBar channel；Every channel includes NsSubcarriers, therefore include N in a pair of AP-MPtx×Nrx×Ns Subcarriers.CSI data sets are divided into N according to channeltx×NrxA Sub Data Set, each Sub Data Set include Ns subcarriers Information.The serial number index per subcarriers in Sub Data Set is made, amplitude value, is applied on each Sub Data Set DBSCAN algorithms are clustered, and two parameters of DBSCAN are that field radius e and minimum include points minOpt respectively.Denoising walks It is rapid as follows：
A) suitable parameter e and minOpt is rule of thumb chosen；
B) it is non-access state, i.e. " unvisited " by all object tags in Sub Data Set；
C) randomly choose one and do not access object o (undex, value), labeled as " visited "；Whether extremely to check the neighborhood of o MinOpt object is included less：If it is not, then mark o is outlier；If it is, all the points are in mark o neighborhoods " visited ", and an a new cluster R and candidate collection Q is created for o, all objects in the neighborhood of o are then placed on candidate In set Q；
D) iteratively the object that other clusters are not belonging in candidate collection Q is added in cluster R, until Q is sky, cluster R is completed；
E) go to step c) and handle next object；
F) the corresponding sample data of object for being marked as outlier is concentrated from training data and deleted, reach the mesh of data de-noising 's.
5) CSI feature extractions：
A) a subcarrier i is chosen, creates the matrix M of a M × P0And null matrix is initialized as, k is calculated according to the following formula：
Wherein H represents CSI data matrixes, and P represents the number of data packet, and M represents matrix resolution, HijRepresent j-th of data packet I-th of subcarrier amplitude, HmaxAnd HminThe maximum and minimum value of amplitude in CSI data are represented respectively.By matrix M0In The value of row k jth row is set to " 1 ", and such each column only has 1 " 1 ", remaining is all " 0 ".When signal amplitude varies widely When, the column distance where adjacent lines nonzero value will increase；
B) by matrix M0Intermediate value is also set to " 1 " for surrounding's element of the element of " 1 ", and carries out expansive working, generation with expansion rate D Expandable matrix Mc.When signal amplitude varies widely, the overlapping region after expansion will be less so that McIn have it is more " 1 " exists；
C) expandable matrix M is calculatedcIn " 1 " number, then calculate expandable matrix in non-zero rate, be denoted as pi；
D) repeat step a), b), c), iterates to calculate the non-zero rate of all subcarriers, obtains vector Wherein NsFor the quantity of all subcarriers.
E) it is vectorialAs the CSI characteristic fingerprint samples after feature extraction, model training and number are carried out Statistics.
6) SVM regression models are trained：SVM is trained to return using the CSI characteristic fingerprints sample and corresponding number of step 5) acquisition Return model, to establish the dependence between CSI characteristic fingerprints sample and number.Sample format is：
Wherein, tiFor the characteristic fingerprint of sample i, ciFor the number corresponding to sample fingerprint i, m is number of samples, is pijSample i The non-zero rate of sub-carriers j, n are number of subcarriers.Training to SVM regression models be exactly solve the problems, such as it is as follows：
Wherein, w and b represents the direction and position of dependence, v ∈ (0,1] it is a parameter, ∈ represents precision, and C is constant, ξi,It is one group of slack variable, tiRepresent a CSI characteristic fingerprint sample, ciIt is the correspondence number of the fingerprint, m is sample
Quantity.Continuous adjusting parameter carries out regression model training, until obtaining optimized parameter, determines optimum regression model：
Wherein, αiWithFor Lagrange multiplier, K (ti,, t) and it is kernel function, using radial basis function (Radical Basis Function,RBF)。
7) demographics：During actual count number, algorithm passes through denoising and feature extraction according to the CSI initial data gathered in real time Afterwards, the SVM regression models obtained by training determine number, include the following steps：
A) CSI initial data is gathered in the way of step 2)；
B) CSI data are generated in the way of step 3)；
C) denoising is carried out to CSI data in the way of step 4)；
D) feature extraction is carried out to CSI data in the way of step 5)；
E) according to SVM regression models and current CSI characteristic fingerprints sample, real-time number is determined.
The accuracy of identification of embodiment such as following table：

Claims (1)

1. the occupancy statistical method that one kind is returned based on WiFi channel condition informations (CSI) and support vector machines (SVM), bag Include following steps：
1) occupancy statistical model training step：
The CSI primary data samples under different number scenes 1-1) are gathered, the sample includes transmission antenna number, reception antenna Number, signal transmission frequency, channel condition information CSI matrixes, and record current persons count；
The first dimension of CSI matrixes in primary data sample 1-2) is removed, by the two-dimensional matrix of generation from linear (level) space Logarithm (power) space is transformed into, by each complex conversion in matrix into amplitude；
1-3) using density-based algorithms (DBSCAN) to being sent and received by every a pair in the channel that antenna forms CSI data are clustered, and the CSI data after denoising are obtained by deleting outlier；
Feature extraction 1-4) is carried out to the CSI data after denoising using expandable matrix algorithm, obtains the non-zero rate per subcarriers, As CSI characteristic fingerprint samples.The expandable matrix algorithm comprises the steps of：The amplitude sequence of every subcarriers is converted to one A two-dimensional matrix；Expand the two-dimensional matrix；Calculate the non-zero rate in the two-dimensional matrix after expansion；To all sons in all channels Carrier wave carries out aforesaid operations, obtains non-zero rate vector and is used as CSI characteristic fingerprint samples；
1-5) be based on CSI characteristic fingerprint samples, preliminary SVM regression models obtained using v-SVR Algorithm for Training, expression number and Dependence between CSI characteristic fingerprint samples；
Parameter optimization 1-6) is carried out to preliminary SVM regression models, is attempted using grid search and cross validation method more different Parameter combination, obtains optimal model parameters and optimal SVM regression models.
2) occupancy statistic procedure：
2-1) gather the CSI primary data samples under unknown number scene；
2-2) according to step 1-2), 1-3) and 1-4) CSI primary data samples are handled and feature extraction, obtain current CSI feature samples fingerprints；
2-2) by step 1-5) and SVM regression models 1-6) established, determined according to current CSI feature samples fingerprint current Number in scene.
CN201711154554.7A 2017-11-20 2017-11-20 A kind of occupancy statistical method based on WiFi channel condition informations and support vector machines Pending CN107992882A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201711154554.7A CN107992882A (en) 2017-11-20 2017-11-20 A kind of occupancy statistical method based on WiFi channel condition informations and support vector machines

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201711154554.7A CN107992882A (en) 2017-11-20 2017-11-20 A kind of occupancy statistical method based on WiFi channel condition informations and support vector machines

Publications (1)

Publication Number Publication Date
CN107992882A true CN107992882A (en) 2018-05-04

Family

ID=62030372

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201711154554.7A Pending CN107992882A (en) 2017-11-20 2017-11-20 A kind of occupancy statistical method based on WiFi channel condition informations and support vector machines

Country Status (1)

Country Link
CN (1) CN107992882A (en)

Cited By (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108650039A (en) * 2018-05-18 2018-10-12 西北大学 A kind of passive type Population size estimation method based on commercial Wi-Fi
CN109255874A (en) * 2018-09-19 2019-01-22 电子科技大学 A kind of passage and number detection method based on general commercial WiFi equipment
CN109472291A (en) * 2018-10-11 2019-03-15 浙江工业大学 A kind of demographics classification method based on DNN algorithm
CN109600758A (en) * 2018-11-15 2019-04-09 南昌航空大学 A kind of stream of people's quantity monitoring method based on RSS
CN110020677A (en) * 2019-03-19 2019-07-16 电子科技大学 A kind of continuous current number detection method based on WiFi Doppler frequency shift
CN110110689A (en) * 2019-05-15 2019-08-09 东北大学 A kind of pedestrian's recognition methods again
CN111753686A (en) * 2020-06-11 2020-10-09 深圳市三旺通信股份有限公司 CSI-based people number identification method, device, equipment and computer storage medium
CN112633763A (en) * 2020-12-31 2021-04-09 上海三零卫士信息安全有限公司 Artificial neural network ANNs-based grade protection risk study and judgment method
CN112949487A (en) * 2021-03-01 2021-06-11 武汉理工大学 WiFi-based ship cab personnel number detection method and system
CN114923267A (en) * 2022-05-19 2022-08-19 浙江启真医健科技有限公司 Temperature control method and system based on space number
CN115001604A (en) * 2022-05-19 2022-09-02 浙江启真医健科技有限公司 Human body sensing method and system based on WiFi microcontroller

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104239951A (en) * 2014-09-10 2014-12-24 西安交通大学 Unbound people counting method achieved by means of channel state information
WO2016011433A2 (en) * 2014-07-17 2016-01-21 Origin Wireless, Inc. Wireless positioning systems
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
CN107154088A (en) * 2017-03-29 2017-09-12 西安电子科技大学 Activity personnel amount method of estimation based on channel condition information

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2016011433A2 (en) * 2014-07-17 2016-01-21 Origin Wireless, Inc. Wireless positioning systems
CN104239951A (en) * 2014-09-10 2014-12-24 西安交通大学 Unbound people counting method achieved by means of channel state information
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
CN107154088A (en) * 2017-03-29 2017-09-12 西安电子科技大学 Activity personnel amount method of estimation based on channel condition information

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
WEI XI 等: "Electronic frog eye: Counting crowd using WiFi", 《 IEEE INFOCOM 2014 - IEEE CONFERENCE ON COMPUTER COMMUNICATIONS》 *
周言: "基于信道状态信息的室内无线定位技术研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 *

Cited By (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108650039A (en) * 2018-05-18 2018-10-12 西北大学 A kind of passive type Population size estimation method based on commercial Wi-Fi
CN109255874A (en) * 2018-09-19 2019-01-22 电子科技大学 A kind of passage and number detection method based on general commercial WiFi equipment
CN109472291A (en) * 2018-10-11 2019-03-15 浙江工业大学 A kind of demographics classification method based on DNN algorithm
CN109600758B (en) * 2018-11-15 2022-03-29 南昌航空大学 RSS-based people flow monitoring method
CN109600758A (en) * 2018-11-15 2019-04-09 南昌航空大学 A kind of stream of people's quantity monitoring method based on RSS
CN110020677A (en) * 2019-03-19 2019-07-16 电子科技大学 A kind of continuous current number detection method based on WiFi Doppler frequency shift
CN110020677B (en) * 2019-03-19 2021-02-02 电子科技大学 Continuous passerby number detection method based on WiFi Doppler frequency shift
CN110110689A (en) * 2019-05-15 2019-08-09 东北大学 A kind of pedestrian's recognition methods again
CN110110689B (en) * 2019-05-15 2023-05-26 东北大学 Pedestrian re-identification method
CN111753686A (en) * 2020-06-11 2020-10-09 深圳市三旺通信股份有限公司 CSI-based people number identification method, device, equipment and computer storage medium
CN112633763A (en) * 2020-12-31 2021-04-09 上海三零卫士信息安全有限公司 Artificial neural network ANNs-based grade protection risk study and judgment method
CN112633763B (en) * 2020-12-31 2024-04-12 上海三零卫士信息安全有限公司 Grade protection risk studying and judging method based on artificial neural network ANNs
CN112949487A (en) * 2021-03-01 2021-06-11 武汉理工大学 WiFi-based ship cab personnel number detection method and system
CN114923267A (en) * 2022-05-19 2022-08-19 浙江启真医健科技有限公司 Temperature control method and system based on space number
CN115001604A (en) * 2022-05-19 2022-09-02 浙江启真医健科技有限公司 Human body sensing method and system based on WiFi microcontroller
CN115001604B (en) * 2022-05-19 2024-04-12 杭州一炜科技有限公司 Human body sensing method and system based on WiFi microcontroller

Similar Documents

Publication Publication Date Title
CN107992882A (en) A kind of occupancy statistical method based on WiFi channel condition informations and support vector machines
Hsieh et al. Deep learning-based indoor localization using received signal strength and channel state information
Li et al. AF-DCGAN: Amplitude feature deep convolutional GAN for fingerprint construction in indoor localization systems
Ahmadien et al. Predicting path loss distribution of an area from satellite images using deep learning
Klautau et al. 5G MIMO data for machine learning: Application to beam-selection using deep learning
Huang et al. WiDet: Wi-Fi based device-free passive person detection with deep convolutional neural networks
CN107480699A (en) A kind of intrusion detection method based on channel condition information and SVMs
CN112152948A (en) Wireless communication processing method and device
CN106792808A (en) Los path recognition methods under a kind of indoor environment based on channel condition information
CN106131958A (en) A kind of based on channel condition information with the indoor Passive Location of support vector machine
Schmidt et al. SDR-Fi: Deep-learning-based indoor positioning via software-defined radio
CN106717082A (en) Mitigating signal noise for fingerprint-based indoor localization
CN107727095A (en) 3D indoor orientation methods based on spectral clustering and weighting reverse transmittance nerve network
CN109922427A (en) Utilize the intelligent radio positioning system and method for large scale array antenna
CN110933628B (en) Fingerprint indoor positioning method based on twin network
CN108122310A (en) A kind of people flow rate statistical method based on WiFi channel state informations and dynamic time warping
Wang et al. Multi-classification of UWB signal propagation channels based on one-dimensional wavelet packet analysis and CNN
CN112135344B (en) CSI (channel State information) and DCNN (distributed channel neural network) -based equipment-free target positioning method
Zhou et al. Device-free crowd counting with WiFi channel state information and deep neural networks
CN109889975B (en) Terminal fingerprint positioning method based on NB-IoT
Pirzadeh et al. Machine-learning assisted outdoor localization via sector-based fog massive MIMO
Yuan et al. Features extraction and analysis for device-free human activity recognition based on channel statement information in b5G wireless communications
CN108828510A (en) A kind of radio frequency tomography localization method based on gradual change shade weight model
Wang et al. Accurate spectrum map construction using an intelligent frequency-spatial reasoning approach
Tulay et al. Road state inference via channel state information

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
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20180504