CN101651951A - Establishing method and positioning method of indoor positioning network of support vector machine based on WLAN - Google Patents
Establishing method and positioning method of indoor positioning network of support vector machine based on WLAN Download PDFInfo
- Publication number
- CN101651951A CN101651951A CN200910072893A CN200910072893A CN101651951A CN 101651951 A CN101651951 A CN 101651951A CN 200910072893 A CN200910072893 A CN 200910072893A CN 200910072893 A CN200910072893 A CN 200910072893A CN 101651951 A CN101651951 A CN 101651951A
- Authority
- CN
- China
- Prior art keywords
- signal
- svm classifier
- positioning
- subsignal
- rss
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Abstract
The invention provides an establishing method and a positioning method of an indoor positioning network of a support vector machine based on WLAN, relating to an establishing method and a positioningmethod of a positioning network in the field of pattern recognition, being used for indoor positioning and solving the problems that the prior method is unable to effectively learn and adapt to the nonlinearity and the nongaussian statistical property of RSS signals and have overlarge searching and matching space and high calculation complexity. The establishing method of the indoor positioning network comprises the following steps: establishing a WLAN network; measuring and recording RSS signals of a reference point to establish an integral signal covering graph; using an SVM classifier to divide a plurality of subsignal covering graphs; converting a multiclass SVM classifier into a two-class SVM classifier; and training the two-class SVM classifier to establish a plurality of independentsubsignal covering graphs. The positioning method comprises the following steps: measuring the RSS signals in a pre-positioning point; pre-positioning the area of the subsignal covering graphs; and positioning the position of the pre-positioning point by a regression function which corresponds to the subsignal covering graphs.
Description
Technical field
The present invention relates to fixer network construction method and localization method in a kind of area of pattern recognition, be specifically related to WLAN indoor positioning network establishing method and localization method.
Background technology
Along with wlan network worldwide by various mechanisms and individual under varying environment (as airport, high-grade office building, research institution, hotel, dining room, campus and family) widespread deployment, and portable terminal universal day by day that disposes wireless network card, the WLAN indoor locating system more and more is subjected to people's attention.The traditional GPS location and the framing signal of Cellular Networks location technology can not effectively cover indoor environment, and are subjected to the influence of indoor complex environment and multipath effect, can not reach required positioning accuracy.Indoor positioning technical research based on WLAN grows up under this application background.At present, indoor positioning technology based on WLAN need not added the location survey specialized hardware in existing WLAN (wireless local area network) facility, but read received signal intensity by the existing wireless network card and the corresponding software of portable terminal, set up the signal coverage diagram, determine mobile subscriber's position by method for mode matching.Existing main pattern matching localization method is weighting k nearest neighbor method and probabilistic method.
Euclidean distance between the RSS sample of signal that weighting k nearest neighbor method is calculated test point and the corresponding RSS sample of signal average of different reference points, draw the reference point and the relevant position coordinate thereof of K Euclidean distance minimum, at last with the inverse of corresponding Euclidean distance as weights, the position of each reference point of linear combination, thus estimate the physical location of test point.Weighting k nearest neighbor method is calculated simple, and the memory space that needs is few, and is single at environment, orientation range hour, positioning accuracy is higher.But, under the general complicated indoor positioning environment, because the influence that multipath effect and personnel walk about etc., the probability distribution of RSS signal on certain fixing reference point presents non-Gauss, non-linear, multi-modal characteristic, the reference point RSS sample of signal average statistical that weighting k nearest neighbor method is utilized can not be represented the distribution character of whole RSS signal, and positioning accuracy is not high.
Probabilistic method is utilized the existing training sample on the reference point, and the RSS signal probability that draws on each reference point distributes.The general Gaussian function that adopts carries out the probability distribution match, draws the average and the bandwidth of the gaussian probability distribution of each reference point.Probabilistic method has made full use of the statistical nature of signal distributions, and positioning accuracy generally wants high than the weighting nearest neighbor method.But in actual applications, for large-scale indoor positioning, space match search scope is bigger, the computation complexity height, and memory space requires bigger.And the probability distribution of RSS signal on certain fixing reference point present non-Gauss, non-linear, multi-modal characteristic, makes the probability-distribution function and the actual probability distribution that simulate differ bigger, thus matching error greatly when causing locating.
Summary of the invention
The purpose of this invention is to provide a kind of SVMs indoor positioning network establishing method and localization method,, can't effectively learn and adapt to non-linear, the non-Gaussian statistics characteristic of RSS signal to solve in the existing WLAN indoor orientation method based on WLAN; And large-scale indoor positioning, the search package space is excessive, the problem that computation complexity is high.The present invention realizes by following proposal:
SVMs indoor positioning network establishing method based on WLAN is realized by following step: one, the room area of desire location is arranged AP, make up thereby finish wlan network; Choose and write down the respective coordinates of reference point and test point at the room area of desire location, measure and the RSS signal of all AP that the record reference point receives, make up the overall signal coverage diagram; Two, whole signal coverage diagram is divided into experimental process signal coverage diagram, partition principle is to make the RSS signal difference opposite sex at partitioning boundary place bigger as far as possible, to reduce boundary error in classification rate; Three, set up a plurality of independently sub-svm classifier devices, the differentiation between two adjacent subsignal coverage diagrams is corresponding to a sub-svm classifier device, and described sub-svm classifier device all is two class svm classifier devices; Four, train the SVMs regression function of each subsignal coverage diagram, promptly draw the Nonlinear Mapping relation of the RSS signal and the physical location of each subsignal coverage diagram.
SVMs indoor positioning network locating method based on WLAN is realized by following step: one, desiring anchor point measure R SS signal; Two, by prior information the RSS signal is judged in advance, drawn this RSS signal and should send into which sub-svm classifier device, go out to desire the regional location of the subsignal coverage diagram at anchor point place by this sub-svm classifier device pre-determined bit; Three, orient the position of desiring anchor point by the pairing regression function of this subsignal coverage diagram.
The present invention proposes a kind of high accuracy, low complex degree based on SVMs (support vectormachines, SVM) network struction of WLAN indoor positioning and localization method.SVMs has deep Statistical Learning Theory basis, and non-linear relation, especially the svm classifier device of energy adaptive learning pattern have high nicety of grading and lower complexity.Localization method at first is positioned to subsignal coverage diagram region by the svm classifier device; Train the mapping function of RSS signal and position coordinates then by the SVM regression algorithm, thereby estimate customer location.This method can adaptive learning RSS signal non-linear, non-Gaussian statistics characteristic, improve positioning accuracy; Simultaneously can effectively reduce the match search space, reduce computation complexity.
The architecture of the main position-based fingerprint recognition of existing WLAN indoor locating system, position fixing process are divided into off-line and online two stages.In off-line phase, realize the sample of signal collection at WLAN internal home network layout, reference point locations mark and selected reference point place, set up the signal coverage diagram of WLAN location; In the online stage, the sample of signal of gathering according to the test point place mates calculating with sample of signal in the signal coverage diagram, draws the location estimation of user terminal.
The present invention also is the architecture that adopts the location fingerprint recognition methods, but has adopted based on svm classifier device pre-determined bit the pinpoint multistep WLAN indoor highly effective of SVM regression function localization method.Solved in the existing WLAN indoor orientation method, can't effectively learn and adapt to non-linear, the non-Gaussian statistics characteristic of RSS signal; And large-scale indoor positioning, the search package space is excessive, the problem that computation complexity is high.
Description of drawings
Fig. 1 is the flow chart based on the SVMs indoor positioning network establishing method of WLAN of embodiment one; Fig. 2 is the flow chart of the localization method of embodiment two, and Fig. 3 is execution mode two positioning experiment environment and the subregion schematic diagram that is marked.The corridor of whole floor (shown in dotted line) all is a locating area, and AP places shown in red-label, and two ellipses are divided into three sub regions with whole locating area.Utilization is received from the RSS sum of AP7, AP8, AP9 can distinguish subregion 1 and subregion 3, judges in advance thereby finish prior information.
Embodiment
Embodiment one: specify embodiments of the present invention below in conjunction with Fig. 1.SVMs indoor positioning network establishing method based on WLAN is realized by following step: one, the room area of desire location is arranged AP, make up thereby finish wlan network; Selecting of AP placement location at first will be satisfied the requirement of WLAN communication, guarantees the even seamless covering of WLAN signal.On this basis, make each anchor point as far as possible, indoor verandas especially, the AP number that the personnel location can receive surpasses three, and is The more the better.Choose and write down the respective coordinates of reference point and test point at the room area of desire location, evenly choose reference point, density be every one of 2m, and the intensive or big zone of signal jitter of personnel is increase reference point density suitably, as electing one of 1m as.The RSS signal of all AP that measurement and record reference point receive makes up the overall signal coverage diagram.Measure and write down the RSS signal and the position coordinates of test point simultaneously.The data of reference point are used for online location and relevant parameters training, are retained in the database; Number of test points is not retained in the database according to being used for the test position fix performance and adjusting relevant parameter.
Two, whole signal coverage diagram is divided into experimental process signal coverage diagram; To reduce the search volume of Matching Location, improve positioning accuracy.The foundation of dividing subsignal coverage diagram boundary is to be selected in the RSS signals such as place that corridor corner, big barrier separate out to change more violent place as far as possible, with the signal difference opposite sex on the boundary both sides that belong to different subsignal coverage diagrams respectively standard more greatly.Help improving nicety of grading like this, reduce the error probability of boundary.
Three, set up a plurality of independently sub-svm classifier devices, the differentiation between two adjacent subsignal coverage diagrams is corresponding to a sub-svm classifier device, and described sub-svm classifier device all is two class svm classifier devices.This step is converted into simple a plurality of independently sub-svm classifier devices with a multiclass svm classifier device of complexity, judgement by prior information, subsignal coverage diagram non-conterminous, apart from each other is made a distinction, the multiclass svm classifier device training of complexity is changed into the training of simple two class svm classifier devices.When orientation range was big, the signal coverage diagram often can be divided into experimental process signal coverage diagram.Simple prior information judgement method in advance is, adopt some AP signal strength signal intensity and, just can judge to exclude the subsignal coverage diagram far away with the physical location of RSS signal, can train a plurality of independently SVM two class graders thus.
Four, train the SVMs regression function of each independent subsignal coverage diagram, promptly draw the Nonlinear Mapping relation of the RSS signal and the physical location of each subsignal coverage diagram.
Utilize ε-insensitive SVMs regression algorithm to construct the Nonlinear Mapping relation of RSS signal and physical location.Given training data (x
i, y
i), i=1 ... l, x
i∈ R
d, x
iBe i learning sample, promptly import the RSS vector, dimension d is the AP number that receives, y
iOutput physics position coordinates for correspondence.Selecting radially, base is a kernel function.Utilize the input and output mapping relations of the position coordinates and the corresponding RSS signal thereof of known reference point, train the parameter of SVMs regression function, draw the RSS signal of each subsignal coverage diagram and the mapping function of position coordinates thus respectively.
Embodiment two: specify present embodiment below in conjunction with Fig. 2.SVMs indoor positioning network locating method based on WLAN is realized by following step: one, desiring anchor point measure R SS signal; By prior information the RSS signal is judged in advance that two, draw this RSS signal and should send into which sub-svm classifier device, the high accuracy of two class svm classifier devices and low complex degree characteristic have guaranteed the accuracy and the real-time of pre-determined bit; Go out the regional location of the subsignal coverage diagram at user place by this svm classifier device pre-determined bit; Three, orient the position of desiring anchor point by the pairing regression function of this subsignal coverage diagram.
Embodiment three: present embodiment specifies the training process of svm classifier device in the execution mode one and utilizes the pre-process of judging to sub-svm classifier device of prior information.
Given training data (x
i, y
i), i=1 ... l, x
i∈ R
d, y ∈ 1,1}
lBe i learning sample, promptly import the RSS vector, dimension d is the AP number that receives, y
i∈ 1,1}
lBe corresponding class.According to structural risk minimization, the original optimization problem form that it will solve is:
ξ
i≥0
B is biasing in the formula (1), and w is a weight coefficient.
Embody empiric risk, C is a balance parameters to empiric risk and VC dimension.By optimization the problems referred to above, can draw the classification hyperplane of maximization class interval.
The indoor positioning zone is divided into plurality of sub-regions, need in theory a multiclass svm classifier device with the consumer premise position to subregion.Because multicategory classification device computation complexity and classification performance all do not have two class grader height, and multiclass svm classifier device is converted into a plurality of independence two class svm classifier devices.Utilize the bigger characteristic of the RSS signal difference opposite sex between the non-conterminous subregion, can utilize simple RSS sum to judge in advance, obtain the RSS signal earlier and belong to which two class svm classifier device.
Problem experimental situation with us is that example is introduced detailed process below.As shown in Figure 3, whole locating area is divided into three sub regions, subregion 1 and subregion 2, subregion 2 and subregion 3 constitute totally two class svm classifier device A and B respectively.The RSS sum that utilization is received from AP7, AP8, AP9 can be judged and distinguishes subregion 1 and subregion 3.If RSS7, RSS8, RSS9 are respectively the RSS that is received from AP7, AP8, AP9, SUM=RSS7+RSS8+RSS9, RSS unit is dBw, gives up unit during calculating, only carries out numerical computations, then:
It is relevant with concrete experimental situation that θ sets, because non-conterminous subregion SUM gap is bigger, sets than being easier to, and this experimental situation is set at-190.
Embodiment four: present embodiment specifies the training process of the step 4 SVMs regression function in the implementation method one, promptly describes how to obtain that the Nonlinear Mapping of RSS signal and physical location concerns in each independent subsignal coverage diagram.
In WLAN indoor positioning algorithm, utilize ε-insensitive SVMs regression algorithm to construct the Nonlinear Mapping relation of RSS signal and physical location.Given training data (x
i, y
i), i=1 ... l, x
i∈ R
d, be i learning sample, promptly import the RSS vector, dimension d is the AP number that receives, y
iOutput physics position coordinates for correspondence.By a Nonlinear Mapping
With data map to a high-dimensional feature space of input sample space, return estimation function at the high-dimensional feature space structure then:
B is biasing in the formula (2), and w is a weight coefficient.
According to structural risk minimization, the original optimization problem form that it will solve is:
In the formula (3) || w||
2The size of control VC dimension, ∑ (ξ
i+ ξ
i *), i=1 ... l embodies empiric risk, and C is a balance parameters to empiric risk and VC dimension.According to the principle of duality and KKT condition, w can be converted into the linear expression of the support vector of lesser amt,
If sv is expressed as the support vector set, then:
In the formula (4),
Can explicitly calculate, draw by the kernel function of calculating a correspondence:
Last corresponding SVMs regression function:
Given parameter ε, C, γ carries out protruding double optimization to formula (3), and the global optimum that can draw w is separated.The output of formula (6) has bidimensional for the physical location coordinate, by respective sample training and parameter search, obtains two SVMs regression functions of independently exporting bidimensional physical location coordinate respectively respectively.
Claims (3)
1, based on the SVMs indoor positioning network establishing method of WLAN, it is characterized in that it realizes by following step: one, the room area of desire location is arranged AP, make up thereby finish wlan network; Choose and write down the respective coordinates of reference point and test point at the room area of desire location, measure and the RSS signal of all AP that the record reference point receives, make up the overall signal coverage diagram; Two, whole signal coverage diagram is divided into experimental process signal coverage diagram, partition principle is to make the RSS signal difference opposite sex at partitioning boundary place bigger as far as possible, to reduce boundary error in classification rate; Three, set up a plurality of independently sub-svm classifier devices, the differentiation between two adjacent subsignal coverage diagrams is corresponding to a sub-svm classifier device, and described sub-svm classifier device all is two class svm classifier devices; Four, train the SVMs regression function of each subsignal coverage diagram, promptly draw the Nonlinear Mapping relation of the RSS signal and the physical location of each subsignal coverage diagram.
2, the SVMs indoor positioning network establishing method based on WLAN according to claim 1 is characterized in that in the step 4, utilizes ε-insensitive SVMs regression algorithm to construct the Nonlinear Mapping relation of RSS signal and physical location; Detailed process is as follows: given training data (x
i, y
i), i=1 ... l, x
i∈ R
d, x
iBe i learning sample, promptly import the RSS vector, dimension d is the AP number that receives, y
iOutput physics position coordinates for correspondence; Selecting radially, base is a kernel function; Utilize the input and output mapping relations of the position coordinates and the corresponding RSS signal thereof of known reference point, train the parameter of SVMs regression function, draw the RSS signal of each subsignal coverage diagram and the mapping function of position coordinates respectively; The RSS signal is imported corresponding SVM regression function, get final product accurate consumer positioning;
According to structural risk minimization, the original optimization problem form that it will solve is:
‖ w ‖ in the formula (1)
2The size of control VC dimension, ∑ (ξ
i+ ξ
i *), i=1 ... l embodies empiric risk, and C is a balance parameters to empiric risk and VC dimension.According to the principle of duality and KKT condition, w can be converted into the linear expression of the support vector of lesser amt,
If sv is expressed as the support vector set, then:
In the formula (2),
Can explicitly calculate, draw by the kernel function of calculating a correspondence:
Last corresponding SVMs regression function:
Given parameter ε, C, γ carries out protruding double optimization to formula (1), and the global optimum that can draw w is separated.The output of formula (4) has bidimensional for the physical location coordinate, by respective sample training and parameter search, obtains two SVMs regression functions of independently exporting bidimensional physical location coordinate respectively respectively.
3, based on the SVMs indoor positioning network locating method of WLAN, it is characterized in that it realizes by following step: one, desiring anchor point measure R SS signal; Two, by prior information the RSS signal is judged in advance, drawn this RSS signal and should send into which sub-svm classifier device, go out to desire the regional location of the subsignal coverage diagram at anchor point place by this sub-svm classifier device pre-determined bit; Three, orient the position of desiring anchor point by the pairing regression function of this subsignal coverage diagram.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN200910072893XA CN101651951B (en) | 2009-09-15 | 2009-09-15 | Establishing method and positioning method of indoor positioning network of support vector machine based on WLAN |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN200910072893XA CN101651951B (en) | 2009-09-15 | 2009-09-15 | Establishing method and positioning method of indoor positioning network of support vector machine based on WLAN |
Publications (2)
Publication Number | Publication Date |
---|---|
CN101651951A true CN101651951A (en) | 2010-02-17 |
CN101651951B CN101651951B (en) | 2011-12-07 |
Family
ID=41674010
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN200910072893XA Expired - Fee Related CN101651951B (en) | 2009-09-15 | 2009-09-15 | Establishing method and positioning method of indoor positioning network of support vector machine based on WLAN |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN101651951B (en) |
Cited By (27)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102395194A (en) * | 2011-08-25 | 2012-03-28 | 哈尔滨工业大学 | ANFIS (Adaptive Neural Fuzzy Inference System) indoor positioning method based on improved GA(Genetic Algorithm) optimization in WLAN (Wireless Local Area Network) environment |
CN102610000A (en) * | 2012-03-14 | 2012-07-25 | 江苏钱旺网络科技有限公司 | Employee attendance location method based on Wi-Fi (wireless fidelity) technology |
CN101695152B (en) * | 2009-10-12 | 2013-01-30 | 中国科学院计算技术研究所 | Indoor positioning method and system thereof |
CN103298107A (en) * | 2013-05-10 | 2013-09-11 | 朱旭东 | Indoor wireless positioning AP (access point) rapid deployment method based on weighted undirected graph |
CN103533648A (en) * | 2013-10-24 | 2014-01-22 | 哈尔滨工业大学 | Position prediction method of mobile users in wireless heterogeneous network based on support vector machine |
CN104007417A (en) * | 2014-05-16 | 2014-08-27 | 上海交通大学 | Indoor positioning method based on block division |
CN104023388A (en) * | 2014-05-09 | 2014-09-03 | 卢彬 | Indoor localization method |
CN104054360A (en) * | 2011-11-08 | 2014-09-17 | 优科无线新加坡(私人)有限公司 | A method and apparatus for determining location information of a position in a multi-storey building |
CN104427612A (en) * | 2013-08-29 | 2015-03-18 | 北京优锐科技有限公司 | WiFi positioning system based on probability and support vector machine computation and method thereof |
CN104469932A (en) * | 2014-11-21 | 2015-03-25 | 北京拓明科技有限公司 | Position fingerprint positioning method based on support vector machine |
CN104507095A (en) * | 2014-12-18 | 2015-04-08 | 哈尔滨工业大学 | Indoor AP (Access Point) deployment method based on user location distribution |
CN104619014A (en) * | 2015-01-09 | 2015-05-13 | 中山大学 | SVM-KNN (Support Vector Machine-K Nearest Neighbor)-based indoor positioning method |
CN104703143A (en) * | 2015-03-18 | 2015-06-10 | 北京理工大学 | Indoor positioning method based on WIFI signal strength |
CN105050052A (en) * | 2015-06-04 | 2015-11-11 | 大连理工大学 | Chi-square measure and sensitivity rule based wireless local area network indoor positioning method |
CN105472733A (en) * | 2015-11-17 | 2016-04-06 | 华南理工大学 | Indoor positioning method based on AP selection |
CN105547299A (en) * | 2015-12-29 | 2016-05-04 | 哈尔滨工业大学 | WLAN (Wireless Local Area Network) indoor positioning method with automatic generation of radio map corner matrix based on track matching |
CN103702416B (en) * | 2013-12-24 | 2017-01-11 | 北京交通大学 | Semi-supervised learning indoor positioning method based on support vector machine |
CN106612495A (en) * | 2015-10-23 | 2017-05-03 | 北斗导航位置服务(北京)有限公司 | Indoor positioning method and system based on propagation loss learning |
CN107144810A (en) * | 2016-12-20 | 2017-09-08 | 金陵科技学院 | Wireless multi-hop range-free localization method based on structural risk minimization |
CN109143164A (en) * | 2018-10-11 | 2019-01-04 | 哈尔滨工业大学 | The radio signal source localization method returned based on Gaussian process |
CN109275096A (en) * | 2018-11-16 | 2019-01-25 | 电子科技大学 | A kind of indoor orientation method based on multilayer converged network Dynamic Matching |
CN110012428A (en) * | 2019-05-22 | 2019-07-12 | 合肥工业大学 | A kind of indoor orientation method based on WiFi |
CN110366108A (en) * | 2019-07-09 | 2019-10-22 | 南京邮电大学 | Indoor orientation method based on channel state information and received signal strength |
CN110831030A (en) * | 2018-08-13 | 2020-02-21 | 华为技术有限公司 | Method for acquiring signal coverage effect diagram and network equipment |
CN110933596A (en) * | 2019-12-04 | 2020-03-27 | 哈尔滨工业大学 | Fingerprint positioning method based on metric learning |
CN111885700A (en) * | 2020-06-08 | 2020-11-03 | 广州杰赛科技股份有限公司 | Mobile terminal positioning method and device combined with support vector machine |
CN114530025A (en) * | 2021-12-31 | 2022-05-24 | 武汉烽理光电技术有限公司 | Tunnel fire alarm method and device based on array grating and electronic equipment |
-
2009
- 2009-09-15 CN CN200910072893XA patent/CN101651951B/en not_active Expired - Fee Related
Cited By (43)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101695152B (en) * | 2009-10-12 | 2013-01-30 | 中国科学院计算技术研究所 | Indoor positioning method and system thereof |
CN102395194B (en) * | 2011-08-25 | 2014-01-08 | 哈尔滨工业大学 | ANFIS (Adaptive Neural Fuzzy Inference System) indoor positioning method based on improved GA(Genetic Algorithm) optimization in WLAN (Wireless Local Area Network) environment |
CN102395194A (en) * | 2011-08-25 | 2012-03-28 | 哈尔滨工业大学 | ANFIS (Adaptive Neural Fuzzy Inference System) indoor positioning method based on improved GA(Genetic Algorithm) optimization in WLAN (Wireless Local Area Network) environment |
CN104054360B (en) * | 2011-11-08 | 2019-02-15 | 优科无线有限公司 | Method and apparatus for determining the location information of position in multi-story structure |
CN104054360A (en) * | 2011-11-08 | 2014-09-17 | 优科无线新加坡(私人)有限公司 | A method and apparatus for determining location information of a position in a multi-storey building |
CN102610000A (en) * | 2012-03-14 | 2012-07-25 | 江苏钱旺网络科技有限公司 | Employee attendance location method based on Wi-Fi (wireless fidelity) technology |
CN103298107A (en) * | 2013-05-10 | 2013-09-11 | 朱旭东 | Indoor wireless positioning AP (access point) rapid deployment method based on weighted undirected graph |
CN103298107B (en) * | 2013-05-10 | 2016-06-22 | 创业软件股份有限公司 | A kind of indoor wireless based on weighted undirected graph positions AP quick deployment method |
CN104427612A (en) * | 2013-08-29 | 2015-03-18 | 北京优锐科技有限公司 | WiFi positioning system based on probability and support vector machine computation and method thereof |
CN103533648A (en) * | 2013-10-24 | 2014-01-22 | 哈尔滨工业大学 | Position prediction method of mobile users in wireless heterogeneous network based on support vector machine |
CN103533648B (en) * | 2013-10-24 | 2016-06-01 | 哈尔滨工业大学 | Based on the wireless isomer network location of mobile users Forecasting Methodology of SVMs |
CN103702416B (en) * | 2013-12-24 | 2017-01-11 | 北京交通大学 | Semi-supervised learning indoor positioning method based on support vector machine |
CN104023388B (en) * | 2014-05-09 | 2019-01-15 | 芜湖泰庆电子科技有限公司 | A kind of indoor orientation method |
CN104023388A (en) * | 2014-05-09 | 2014-09-03 | 卢彬 | Indoor localization method |
CN104007417A (en) * | 2014-05-16 | 2014-08-27 | 上海交通大学 | Indoor positioning method based on block division |
CN104469932A (en) * | 2014-11-21 | 2015-03-25 | 北京拓明科技有限公司 | Position fingerprint positioning method based on support vector machine |
CN104469932B (en) * | 2014-11-21 | 2018-07-06 | 北京拓明科技有限公司 | A kind of location fingerprint localization method based on support vector machines |
CN104507095A (en) * | 2014-12-18 | 2015-04-08 | 哈尔滨工业大学 | Indoor AP (Access Point) deployment method based on user location distribution |
CN104507095B (en) * | 2014-12-18 | 2018-03-13 | 哈尔滨工业大学 | Indoor AP dispositions methods based on user location distribution |
CN104619014A (en) * | 2015-01-09 | 2015-05-13 | 中山大学 | SVM-KNN (Support Vector Machine-K Nearest Neighbor)-based indoor positioning method |
CN104703143A (en) * | 2015-03-18 | 2015-06-10 | 北京理工大学 | Indoor positioning method based on WIFI signal strength |
CN104703143B (en) * | 2015-03-18 | 2018-03-27 | 北京理工大学 | A kind of indoor orientation method based on WIFI signal intensity |
CN105050052A (en) * | 2015-06-04 | 2015-11-11 | 大连理工大学 | Chi-square measure and sensitivity rule based wireless local area network indoor positioning method |
CN106612495A (en) * | 2015-10-23 | 2017-05-03 | 北斗导航位置服务(北京)有限公司 | Indoor positioning method and system based on propagation loss learning |
CN106612495B (en) * | 2015-10-23 | 2019-10-25 | 北斗导航位置服务(北京)有限公司 | A kind of indoor orientation method and system based on propagation loss study |
CN105472733A (en) * | 2015-11-17 | 2016-04-06 | 华南理工大学 | Indoor positioning method based on AP selection |
CN105472733B (en) * | 2015-11-17 | 2019-10-18 | 华南理工大学 | It is a kind of applied in indoor positioning based on AP selective positioning method |
CN105547299A (en) * | 2015-12-29 | 2016-05-04 | 哈尔滨工业大学 | WLAN (Wireless Local Area Network) indoor positioning method with automatic generation of radio map corner matrix based on track matching |
CN107144810B (en) * | 2016-12-20 | 2020-10-13 | 金陵科技学院 | Wireless multi-hop non-ranging positioning method based on structure risk minimization |
CN107144810A (en) * | 2016-12-20 | 2017-09-08 | 金陵科技学院 | Wireless multi-hop range-free localization method based on structural risk minimization |
CN110831030A (en) * | 2018-08-13 | 2020-02-21 | 华为技术有限公司 | Method for acquiring signal coverage effect diagram and network equipment |
CN110831030B (en) * | 2018-08-13 | 2021-09-14 | 华为技术有限公司 | Method for acquiring signal coverage effect diagram and network equipment |
CN109143164A (en) * | 2018-10-11 | 2019-01-04 | 哈尔滨工业大学 | The radio signal source localization method returned based on Gaussian process |
CN109275096A (en) * | 2018-11-16 | 2019-01-25 | 电子科技大学 | A kind of indoor orientation method based on multilayer converged network Dynamic Matching |
CN110012428B (en) * | 2019-05-22 | 2020-12-29 | 合肥工业大学 | Indoor positioning method based on WiFi |
CN110012428A (en) * | 2019-05-22 | 2019-07-12 | 合肥工业大学 | A kind of indoor orientation method based on WiFi |
CN110366108A (en) * | 2019-07-09 | 2019-10-22 | 南京邮电大学 | Indoor orientation method based on channel state information and received signal strength |
CN110933596B (en) * | 2019-12-04 | 2021-04-13 | 哈尔滨工业大学 | Fingerprint positioning method based on metric learning |
CN110933596A (en) * | 2019-12-04 | 2020-03-27 | 哈尔滨工业大学 | Fingerprint positioning method based on metric learning |
CN111885700A (en) * | 2020-06-08 | 2020-11-03 | 广州杰赛科技股份有限公司 | Mobile terminal positioning method and device combined with support vector machine |
CN111885700B (en) * | 2020-06-08 | 2022-04-12 | 广州杰赛科技股份有限公司 | Mobile terminal positioning method and device combined with support vector machine |
CN114530025B (en) * | 2021-12-31 | 2024-03-08 | 武汉烽理光电技术有限公司 | Tunnel fire alarming method and device based on array grating and electronic equipment |
CN114530025A (en) * | 2021-12-31 | 2022-05-24 | 武汉烽理光电技术有限公司 | Tunnel fire alarm method and device based on array grating and electronic equipment |
Also Published As
Publication number | Publication date |
---|---|
CN101651951B (en) | 2011-12-07 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN101651951B (en) | Establishing method and positioning method of indoor positioning network of support vector machine based on WLAN | |
US10884112B2 (en) | Fingerprint positioning method and system in smart classroom | |
Zhang et al. | DeepPositioning: Intelligent fusion of pervasive magnetic field and WiFi fingerprinting for smartphone indoor localization via deep learning | |
CN101778399B (en) | Method for optimizing WLAN (Wireless Local Area Network) indoor ANN (Artificial Neural Network) positioning based on FCM (fuzzy C-mean) and least-squares curve surface fitting methods | |
Elbakly et al. | A robust zero-calibration RF-based localization system for realistic environments | |
Battiti et al. | Statistical learning theory for location fingerprinting in wireless LANs | |
Huang et al. | A fine-grained indoor fingerprinting localization based on magnetic field strength and channel state information | |
CN101873607B (en) | WLAN (Wireless Local Area Network) indoor step-type RD-ANFIS (Region Division-Adaptive Network-based Fuzzy Inference System) positioning method | |
CN108696932A (en) | It is a kind of using CSI multipaths and the outdoor fingerprint positioning method of machine learning | |
CN104540221B (en) | WLAN indoor orientation methods based on semi-supervised SDE algorithms | |
CN101883424A (en) | WLAN (Wireless Local Area Network) indoor KNN (K-Nearest Neighbor) positioning method based on near-neighbor point number optimization | |
Tian et al. | Optimizing AP and Beacon Placement in WiFi and BLE hybrid localization | |
Ning et al. | Outdoor location estimation using received signal strength-based fingerprinting | |
CN108882363A (en) | A kind of multi-direction acquisition combines the WiFi fingerprint indoor orientation method of cluster | |
CN104581945A (en) | WLAN indoor positioning method for distance constraint based semi-supervised APC clustering algorithm | |
Kwak et al. | Magnetic field based indoor localization system: A crowdsourcing approach | |
CN109541537B (en) | Universal indoor positioning method based on ranging | |
Huang et al. | QSFDEW: a fingerprint positioning method based on quadtree search and fractal direction entropy weighting | |
Petric et al. | Indoor localization using multi-operator public land mobile networks and support vector machine learning algorithms | |
Au | Rss-based wlan indoor positioning and tracking system using compressive sensing and its implementation on mobile devices | |
Jiang et al. | Competitive agglomeration based KNN in indoor WLAN localization environment | |
Wei et al. | Zero-cost and map-free shop-level localization algorithm based on crowdsourcing fingerprints | |
Zhang et al. | Fingerprint-based indoor localization using weighted K-nearest neighbor and weighted signal intensity | |
Lin et al. | Two-stage clustering for improve indoor positioning accuracy | |
CN107087259A (en) | Region Wi-Fi hotspot position finding technology based on mobile phone |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant | ||
C17 | Cessation of patent right | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20111207 Termination date: 20120915 |