CN108696818B - Distance measurement method and system for Wi-Fi fingerprint positioning - Google Patents
Distance measurement method and system for Wi-Fi fingerprint positioning Download PDFInfo
- Publication number
- CN108696818B CN108696818B CN201810427419.3A CN201810427419A CN108696818B CN 108696818 B CN108696818 B CN 108696818B CN 201810427419 A CN201810427419 A CN 201810427419A CN 108696818 B CN108696818 B CN 108696818B
- Authority
- CN
- China
- Prior art keywords
- data
- rss
- distance measurement
- training
- model
- 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.)
- Active
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
- H04W4/02—Services making use of location information
- H04W4/025—Services making use of location information using location based information parameters
- H04W4/026—Services making use of location information using location based information parameters using orientation information, e.g. compass
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S5/00—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations
- G01S5/02—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations using radio waves
- G01S5/0252—Radio frequency fingerprinting
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W64/00—Locating users or terminals or network equipment for network management purposes, e.g. mobility management
- H04W64/006—Locating users or terminals or network equipment for network management purposes, e.g. mobility management with additional information processing, e.g. for direction or speed determination
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Position Fixing By Use Of Radio Waves (AREA)
Abstract
The invention discloses a distance measuring method and system for Wi-Fi fingerprint positioning. The distance measurement method comprises the following steps: s1Acquiring RSS signal data and GPS data of sampling points; the GPS data includes a longitude and latitude of the sample point; s2Converting the GPS data and the RSS signal data into first training data and second training data by utilizing a first characteristic conversion function and a second characteristic conversion function; s3Establishing a distance measurement model; s4And training the distance measurement model by using the first training data and the second training data to obtain a final distance measurement model and the weight coefficient vector. The invention uses the weighted Euclidean distance to lead the distance of the signal space to be better mapped to the position space, obtains the weight through the learning algorithm, leads the robustness of the mapping to be better, and greatly improves the positioning accuracy.
Description
Technical Field
The invention relates to the technical field of Wi-Fi (wireless fidelity) fingerprint positioning, in particular to a distance measurement method and a distance measurement system for Wi-Fi fingerprint positioning.
Background
Among outdoor Positioning, GPS (Global Positioning System) Positioning is the most common Positioning method. However, in an indoor environment, GPS positioning often cannot achieve a good positioning effect due to severe attenuation and multipath effects of signals. In this case, many positioning methods, such as base station positioning, ZigBee (ZigBee protocol) positioning, RFID (radio frequency identification) positioning, Wi-Fi fingerprint positioning technology, and the like, have appeared.
Relatively speaking, the method based on Wi-Fi fingerprint positioning is easy to realize and high in precision. However, the current fingerprint positioning method has the following defects:
(1) the signal space is directly regarded as Euclidean space, the fingerprint closest to the Euclidean distance (or other distances such as Manhattan distance) of an RSS (received signal strength) observation value in the signal space is found, and then the position coordinate corresponding to the fingerprint is used as the position of the mobile equipment. However, since the signal space is different from the location space, this approach of directly mapping the distance of the signal space to the location space may cause errors in some cases, such as at points where the signal space is closer, and at points where the location space is farther;
(2) not all APs (wireless access points) can be detected at all positions, the number of APs may be increased or decreased when the APs are detected at different time points, and the unreliable RSS signals of the APs make accurate positioning difficult;
(3) the spread of radio is easily influenced by the environment, the orientation of people has a significant influence on the RSS measured by the mobile device, and the way in which network cards from different suppliers calculate the RSS is somewhat different, which also causes inconsistency in the RSS measurement. The ruggedness of the RSS distribution results in the possibility that location fingerprints that are further away from the true location will be matched when the observation vector is measured, especially when the matching algorithm is more dependent on certain dimensions.
Disclosure of Invention
The invention aims to overcome the defects of large error and low accuracy of a Wi-Fi fingerprint positioning method in the prior art, and provides a distance measuring method and a distance measuring system for Wi-Fi fingerprint positioning.
The invention solves the technical problems through the following technical scheme:
a distance measurement method of Wi-Fi fingerprinting, the distance measurement method comprising the steps of:
S1acquiring RSS signal data and GPS data of sampling points; the GPS data includes a longitude and latitude of the sample point;
S2converting the GPS data and the RSS signal data into first training data and second training data by utilizing a first characteristic conversion function and a second characteristic conversion function;
the first characteristic transfer function is:
the second characteristic transfer function is:
Fr(x,y)=[(x.rssi-y.rssi)2];
wherein x.lat and y.lat represent the latitude of any two sampling points; lon and y.lon characterize the longitude of any two sample points; rsiAnd y.rssiRepresenting the RSS values of the APs of any two sampling points, wherein i is less than or equal to n; n represents the number of the APs;
S3establishing a distance measurement model;
the distance metric model is:
wherein θ represents a weight coefficient vector;
S4and training the distance measurement model by using the first training data and the second training data to obtain a final distance measurement model and the weight coefficient vector.
Preferably, the RSS signal data includes the signal strength of the AP collected at each sampling point;
step S2Previously, the distance metric method further comprises:
deleting data of the AP with frequency smaller than a frequency threshold appearing in the RSS signal data;
and/or deleting the data of the AP with the signal intensity smaller than the intensity threshold value in the RSS signal data;
and/or deleting the data of the AP, of which the average value of the signal intensity collected at different sampling time points in the RSS signal data is not in a preset range.
Preferably, step S4The method specifically comprises the following steps:
establishing a loss function of the distance metric model;
the loss function is:
wherein, alpha, C1And C2Is a hyperparameter of the loss function;
and training the distance measurement model by using the first training data and the second training data, and respectively determining the distance measurement model and the weight coefficient vector when the loss function is the minimum value as a final distance measurement model and a final weight coefficient vector.
Preferably, the distance measurement method further comprises:
acquiring RSS signal data to be measured of a measuring point and cleaning the data;
and inputting the cleaned RSS signal data to be measured into the distance measurement model to calculate the GPS data of the sampling point closest to the measuring point.
A distance-metric system for Wi-Fi fingerprinting, the distance-metric system comprising:
the data acquisition module is used for acquiring RSS signal data and GPS data of sampling points;
the GPS data includes a longitude and latitude of the sample point;
a data conversion module for converting the GPS data and the RSS signal data into first training data and second training data using a first feature conversion function and a second feature conversion function;
the first characteristic transfer function is:
the second characteristic transfer function is:
Fr(x,y)=[(x.rssi-y.rssi)2];
wherein x.lat and y.lat represent the latitude of any two sampling points; lon and y.lon characterize the longitude of any two sample points; rsiAnd y.rssiRepresenting the RSS values of the APs of any two sampling points, wherein i is less than or equal to n; n represents the number of the APs;
the model establishing module is used for establishing a distance measurement model;
the distance metric model is:
wherein θ represents a weight coefficient vector;
and the model training module is used for training the distance measurement model by utilizing the first training data and the second training data to obtain a final distance measurement model and the weight coefficient vector.
Preferably, the RSS signal data includes the signal strength of the AP collected at each sampling point;
the distance metric system further comprises:
the data cleaning module is used for deleting the data of the AP with the frequency smaller than the frequency threshold value in the RSS signal data; and/or deleting the data of the AP with the signal intensity smaller than the intensity threshold value in the RSS signal data; and/or deleting the data of the AP, of which the average value of the signal intensity collected at different sampling time points in the RSS signal data is not in a preset range.
Preferably, the model training module specifically includes:
a function establishing unit, configured to establish a loss function of the distance metric model;
the loss function is:
wherein, alpha, C1And C2Is a hyperparameter of the loss function;
and the model training unit is used for training the distance measurement model by utilizing the first training data and the second training data, and respectively determining the distance measurement model and the weight coefficient vector when the loss function is the minimum value as a final distance measurement model and a final weight coefficient vector.
Preferably, the distance measuring system further comprises:
and the positioning module is used for acquiring the RSS signal data to be measured of the measuring point, cleaning the data and inputting the cleaned RSS signal data to be measured into the distance measurement model so as to calculate the GPS data of the sampling point closest to the measuring point.
The positive progress effects of the invention are as follows: the invention uses the weighted Euclidean distance to lead the distance of the signal space to be better mapped to the position space, obtains the weight through the learning algorithm, leads the robustness of the mapping to be better, and can be suitable for fingerprint points collected under different conditions. Meanwhile, the weight of certain dimensionality is made to be 0 through automatic adjustment, some useless dimensionalities can be automatically removed from a plurality of dimensionalities, the effect of automatically selecting the number of the dimensionalities is achieved, the weight of each dimensionality is limited, each dimensionality can play a role, the dimensionalities do not depend on a certain dimensionality, and therefore even if the certain dimensionality changes during detection, a result can be well matched, and accuracy is greatly improved.
Drawings
Fig. 1 is a flowchart of a distance measurement method for Wi-Fi fingerprint positioning according to embodiment 1 of the present invention.
Fig. 2 is a schematic block diagram of a Wi-Fi fingerprint location distance measurement system according to embodiment 2 of the present invention.
Detailed Description
The invention is further illustrated by the following examples, which are not intended to limit the scope of the invention.
Example 1
As shown in fig. 1, the distance measurement method for Wi-Fi fingerprint positioning of the present embodiment includes the following steps:
The distance measurement method of the present embodiment includes two stages: an offline phase and an online phase. In step 101, that is, in the off-line acquisition phase, the location space is divided into a plurality of different areas, and the area of the area and the number of APs need to be considered comprehensively, so that the total number of the areas and the APs in the location space is appropriate. And collecting RSS signal data and corresponding GPS data of a position space at sampling points of each area. The RSS signal data includes MAC (message authentication code) address identification of different APs and signal strength of each AP collected by each sampling point. The GPS data includes the longitude and latitude of the sample point.
In this embodiment, the RSS signal data is cleaned, that is, the RSS signals are screened, and a relatively stable signal is selected. For example, data of an AP whose frequency appearing in RSS signal data is less than a frequency threshold is deleted; and/or deleting the data of the AP with the signal intensity smaller than the intensity threshold value in the RSS signal data; and/or deleting data of the AP whose average of the signal intensities acquired at different sampling time points is not within a preset range (e.g., 2 times standard deviation).
After data cleansing, the RSS signal measured in one area is represented by the matrix R as:
the corresponding GPS is represented by a location matrix G as:
wherein n is the number of APs in an area, and m is the number of sampling points.
In this embodiment, the first characteristic conversion function is:
wherein x.lat and y.lat represent the latitude of any two sampling points; lon and y.lon characterize the longitude of any two sample points. That is, the GPS coordinate of the sample point x is expressed as x (x.lat, x.lon), and the GPS coordinate of the sample point y is expressed as y (y.lat, y.lon). The coordinates of x and y are both expressed in radians.
The second characteristic transfer function is:
Fr(x,y)=[(x.rssi-y.rssi)2]=[(x.rss1-y.rss1)2 (x.rss2-y.rss2)2 … (x.rssn-y.rssn)2];
fr (x, y) denotes an n-dimensional vector, x.rssiAnd y.rssiRepresenting the RSS values of the APs of any two sampling points, wherein i is less than or equal to n.
After conversion, matrices X (first training data) and Y (second training data) for training are obtained:
And step 104, establishing a distance measurement model.
In this embodiment, the distance metric model is;
wherein θ represents a weight coefficient vector, θ ═ θ1 θ2 … θn]。θ1、θ2And thetanIs a weight coefficient for weighting the euclidean distance.
And 105, training a distance measurement model by using the first training data and the second training data to obtain a final distance measurement model and a final weight coefficient vector.
Step 105 completes the establishment of the fingerprint database of each area in the off-line stage, and the content of the fingerprint database comprises the GPS coordinates, the AP (MAC address and RSS) value and the model parameter theta.
In this embodiment, step 105 specifically includes:
and 105-1, establishing a loss function of the distance measurement model.
Wherein the loss function is:
wherein, alpha, C1And C2Is the hyperparameter of the loss function.
‖θ‖1=∑|θi|;
And 105-2, training a distance measurement model by using the first training data and the second training data, and respectively determining the distance measurement model and the weight coefficient vector when the loss function is the minimum value as a final distance measurement model and a final weight coefficient vector.
Step 105-2, namely solving argminθL (θ), the importance of each AP is controlled by the weight θ. By hyperparametric C1Control L1The size of the regularization term can realize the sparsity of theta, so as to eliminate some less important and even bad-influence APs. Hyperparameter C2Control L2The value of each weight of theta is kept in a small range by the size of the regular term, so that the result is not greatly influenced when the RSS value of an individual AP is abnormal, and the stability of the model is improved. After the training is finished, the pairθ is normalized so that sum (θ) becomes 1.
Model training is to train X, Y matrix input distance measurement model, and update parameter θ by using gradient descent method, where the formula is as follows:
the initial value θ may be set to a small random number and the iterative method may use a fixed size training sample for each training iteration.
In the online stage, that is, using the trained metric model, the measurement point is located, so that the distance metric method further includes:
and 106, acquiring RSS signal data to be measured of the measuring point and cleaning the data.
Data cleansing also filters out APs that are not present in the fingerprint library. Before step 106, online matching may be performed, an area where a measurement point (fingerprint) is located is determined by AP matching, the area may be determined by directly searching a matched MAC address, may be determined by a Kmeans method, or may be directly measured by using the method of this embodiment.
And step 107, inputting the cleaned RSS signal data to be measured into a distance measurement model to calculate the GPS data of the sampling point closest to the measuring point.
And determining the GPS data of the nearest sampling point as the longitude and latitude of the measuring point to realize the positioning of the measuring point. According to the method, the weighted Euclidean distance is used, so that the distance of a signal space can be better mapped to a position space, the weight is obtained through a learning algorithm, the robustness of the mapping is better, and the method can be suitable for fingerprint points acquired under different conditions. Meanwhile, the weight of certain dimensionality is made to be 0 through automatic adjustment, some useless dimensionalities can be automatically removed from a plurality of dimensionalities, the effect of automatically selecting the number of the dimensionalities is achieved, the weight of each dimensionality is limited, each dimensionality can play a role, the dimensionalities do not depend on a certain dimensionality, and therefore even if the certain dimensionality changes during detection, a result can be well matched, and accuracy is greatly improved.
Example 2
As shown in fig. 2, the distance measurement system for Wi-Fi fingerprint location of the present embodiment includes: the system comprises a data acquisition module 1, a data conversion module 2, a model establishment module 3, a model training module 4, a data cleaning module 5 and a positioning module 6. The model training module comprises: the device comprises a function building unit and a model training unit.
The data acquisition module 1 is used for acquiring RSS signal data and GPS data of a sampling point. The RSS signal data includes MAC (message authentication code) address identifiers of different APs and signal strength of each AP, which are collected by each sampling point. The GPS data includes the longitude and latitude of the sample point.
The data cleaning module 5 is used for cleaning the RSS signal data. For example, data of an AP whose frequency appearing in RSS signal data is less than a frequency threshold is deleted; and/or deleting the data of the AP with the signal intensity smaller than the intensity threshold value in the RSS signal data; and/or deleting the data of the AP, of which the average value of the signal intensity collected at different sampling time points in the RSS signal data is not within a preset range.
After data cleansing, the RSS signal measured in one area is represented by the matrix R as:
the corresponding GPS is represented by a location matrix G as:
wherein n is the number of APs in an area, and m is the number of sampling points.
The data conversion module 2 is configured to convert the GPS data and the RSS signal data into first training data and second training data using a first feature conversion function and a second feature conversion function.
The first characteristic transfer function is:
the second characteristic transfer function is:
Fr(x,y)=[(x.rssi-y.rssi)2];
wherein x.lat and y.lat represent the latitude of any two sampling points; lon and y.lon characterize the longitude of any two sample points; rsiAnd y.rssiRepresenting the RSS values of the APs of any two sampling points, wherein i is less than or equal to n.
After conversion, matrices X (first training data) and Y (second training data) for training are obtained:
The model building module 3 is used for building a distance measurement model.
The distance metric model is:
where θ represents the weight coefficient vector.
The model training module 4 is configured to train the distance metric model using the first training data and the second training data to obtain a final distance metric model and a weight coefficient vector.
Specifically, the function establishing unit of the model training module is used for establishing a loss function of the distance metric model. The model training unit is used for training a distance measurement model by utilizing the first training data and the second training data, and respectively determining the distance measurement model and the weight coefficient vector when the loss function is the minimum value as a final distance measurement model and a final weight coefficient vector.
The loss function is:
wherein, alpha, C1And C2Is the hyperparameter of the loss function.
The positioning module 6 is used for acquiring the RSS signal data to be measured of the measuring point, cleaning the data, and inputting the cleaned RSS signal data to be measured into the distance measurement model to calculate the GPS data of the sampling point closest to the measuring point. The GPS data of the nearest sampling point is the longitude and latitude of the measuring point, thereby realizing the positioning of the measuring point.
While specific embodiments of the invention have been described above, it will be appreciated by those skilled in the art that this is by way of example only, and that the scope of the invention is defined by the appended claims. Various changes and modifications to these embodiments may be made by those skilled in the art without departing from the spirit and scope of the invention, and these changes and modifications are within the scope of the invention.
Claims (8)
1. A distance measurement method for Wi-Fi fingerprint positioning is characterized by comprising the following steps:
S1acquiring RSS signal data and GPS data of sampling points; the GPS data includes a longitude and latitude of the sample point;
S2converting the RSS signal data into first training data by using a second characteristic conversion function, and converting the GPS data into first training data by using a first characteristic conversion functionSecond training data;
the first characteristic transfer function is:
the second characteristic transfer function is:
Fr(x,y)=[(x.rss1-y.rss1)2,(x.rss2-y.rss2)2,...,(x.rssi-y.rssi)2,...,(x.rssn-y.rssn)2];
the first training data is a matrix X:
the second training data is a matrix Y:
wherein x.lat and y.lat represent the latitude of any two sampling points; x.lon and y.lon characterize the longitude of the any two sample points; rsiAnd y.rssiRepresenting the RSS value of the ith AP of any two sampling points, wherein i is less than or equal to n; n represents the number of the APs, and m is the number of sampling points;
S3establishing a distance measurement model;
the distance metric model is:
wherein θ represents a weight coefficient vector, θ ═ θ1,θ2,...,θn]The importance of each AP is controlled by the weight theta;
S4training the distance measurement model by using the first training data and the second training data to obtain a final distance measurement model and the weight coefficient vector;
step S4The method specifically comprises the following steps:
and establishing a loss function of the distance measurement model, training the distance measurement model by using the first training data and the second training data, and respectively determining the distance measurement model and the weight coefficient vector when the loss function is the minimum value as a final distance measurement model and a final weight coefficient vector.
2. The Wi-Fi fingerprinting method of claim 1, wherein the RSS signal data includes a signal strength of the AP acquired at each sample point;
step S2Previously, the distance metric method further comprises:
deleting data of the AP with frequency smaller than a frequency threshold appearing in the RSS signal data;
and/or deleting the data of the AP with the signal intensity smaller than the intensity threshold value in the RSS signal data;
and/or deleting the data of the AP, of which the average value of the signal intensity collected at different sampling time points in the RSS signal data is not in a preset range.
4. The Wi-Fi fingerprinting method of claim 1, wherein the distance metric method further comprises:
acquiring RSS signal data to be measured of a measuring point and cleaning the data;
and inputting the cleaned RSS signal data to be measured into the distance measurement model to calculate the GPS data of the sampling point closest to the measuring point.
5. A distance-metric system for Wi-Fi fingerprinting, the distance-metric system comprising:
the data acquisition module is used for acquiring RSS signal data and GPS data of sampling points;
the GPS data includes a longitude and latitude of the sample point;
the data conversion module is used for converting the RSS signal data into first training data by using a second characteristic conversion function and converting the GPS data into second training data by using the first characteristic conversion function;
the first characteristic transfer function is:
the second characteristic transfer function is:
Fr(x,y)=[(x.rss1-y.rss1)2,(x.rss2-y.rss2)2,...,(x.rssi-y.rssi)2,...,(x.rssn-y.rssn)2];
the first training data is a matrix X:
the second training data is a matrix Y:
wherein x.lat and y.lat represent the latitude of any two sampling points; x.lon and y.lon characterize the longitude of the any two sample points; rsiAnd y.rssiRepresenting the RSS value of the ith AP of any two sampling points, wherein i is less than or equal to n; n represents the number of the APs, and m is the number of sampling points;
the model establishing module is used for establishing a distance measurement model;
the distance metric model is:
wherein θ represents a weight coefficient vector, θ ═ θ1,θ2,...,θn]The importance of each AP is controlled by the weight theta;
the model training module is used for training the distance measurement model by utilizing the first training data and the second training data to obtain a final distance measurement model and the weight coefficient vector;
the model training module specifically comprises:
a function establishing unit, configured to establish a loss function of the distance metric model;
and the model training unit is used for training the distance measurement model by utilizing the first training data and the second training data, and respectively determining the distance measurement model and the weight coefficient vector when the loss function is the minimum value as a final distance measurement model and a final weight coefficient vector.
6. The Wi-Fi fingerprinting distance metric system of claim 5, wherein the RSS signal data includes a signal strength of the AP acquired at each sampling point;
the distance metric system further comprises:
the data cleaning module is used for deleting the data of the AP with the frequency smaller than the frequency threshold value in the RSS signal data; and/or deleting the data of the AP with the signal intensity smaller than the intensity threshold value in the RSS signal data; and/or deleting the data of the AP, of which the average value of the signal intensity collected at different sampling time points in the RSS signal data is not in a preset range.
8. The Wi-Fi fingerprinting distance measurement system of claim 5, wherein the distance measurement system further comprises:
and the positioning module is used for acquiring the RSS signal data to be measured of the measuring point, cleaning the data and inputting the cleaned RSS signal data to be measured into the distance measurement model so as to calculate the GPS data of the sampling point closest to the measuring point.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810427419.3A CN108696818B (en) | 2018-05-07 | 2018-05-07 | Distance measurement method and system for Wi-Fi fingerprint positioning |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810427419.3A CN108696818B (en) | 2018-05-07 | 2018-05-07 | Distance measurement method and system for Wi-Fi fingerprint positioning |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108696818A CN108696818A (en) | 2018-10-23 |
CN108696818B true CN108696818B (en) | 2020-12-01 |
Family
ID=63846146
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810427419.3A Active CN108696818B (en) | 2018-05-07 | 2018-05-07 | Distance measurement method and system for Wi-Fi fingerprint positioning |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108696818B (en) |
Families Citing this family (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114449435B (en) * | 2020-10-19 | 2023-12-15 | ***通信集团重庆有限公司 | Fingerprint positioning method based on measurement learning of large-interval neighbor |
CN114501364B (en) * | 2022-02-22 | 2023-12-22 | 成都市联洲国际技术有限公司 | Indoor positioning method and device based on WIFI signals and electronic equipment |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101247650A (en) * | 2008-03-20 | 2008-08-20 | 中科院嘉兴中心微***所分中心 | Node indoor locating method based on wireless signal strength in wireless sensor network |
CN101754235A (en) * | 2008-12-11 | 2010-06-23 | 财团法人工业技术研究院 | Method for collecting, constructing and positioning training position data of positioning system |
CN101986757A (en) * | 2010-11-05 | 2011-03-16 | 中兴通讯股份有限公司 | Positioning method and device for mobile terminal |
CN106376080A (en) * | 2016-11-09 | 2017-02-01 | 北京邮电大学 | AP filtering method and device |
CN107333276A (en) * | 2017-05-27 | 2017-11-07 | 南京邮电大学 | WiFi location fingerprint localization methods based on weighted chi-square distance |
-
2018
- 2018-05-07 CN CN201810427419.3A patent/CN108696818B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101247650A (en) * | 2008-03-20 | 2008-08-20 | 中科院嘉兴中心微***所分中心 | Node indoor locating method based on wireless signal strength in wireless sensor network |
CN101754235A (en) * | 2008-12-11 | 2010-06-23 | 财团法人工业技术研究院 | Method for collecting, constructing and positioning training position data of positioning system |
CN101986757A (en) * | 2010-11-05 | 2011-03-16 | 中兴通讯股份有限公司 | Positioning method and device for mobile terminal |
CN106376080A (en) * | 2016-11-09 | 2017-02-01 | 北京邮电大学 | AP filtering method and device |
CN107333276A (en) * | 2017-05-27 | 2017-11-07 | 南京邮电大学 | WiFi location fingerprint localization methods based on weighted chi-square distance |
Non-Patent Citations (2)
Title |
---|
Accuracy Indicator for Fingerprinting Localization Systems;Vahideh Moghtadaiee;《Proc. Of IEEE/ION Position,Location and Navigation Symposium》;20120712;第2-4节 * |
机器学习中损失函数问题研究;苏美红;《中国优秀硕士学位论文全文数据库 信息科技辑》;20151215;全文 * |
Also Published As
Publication number | Publication date |
---|---|
CN108696818A (en) | 2018-10-23 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106851573B (en) | Log path loss model-based joint weighting K nearest neighbor indoor positioning method | |
CN110166991B (en) | Method, device, apparatus and storage medium for locating electronic device | |
CN106646338B (en) | A kind of quickly accurate indoor orientation method | |
CN110557716B (en) | Indoor positioning method based on lognormal model | |
CN112147573A (en) | Passive positioning method based on amplitude and phase information of CSI (channel State information) | |
CN105657823B (en) | WIFI indoor weighted K nearest neighbor positioning algorithm based on kernel function main feature extraction | |
CN110602651B (en) | Positioning method based on WIFI position fingerprint and positioning system of robot | |
CN110536257B (en) | Indoor positioning method based on depth adaptive network | |
CN112040397B (en) | CSI indoor fingerprint positioning method based on adaptive Kalman filtering | |
CN109640262B (en) | Positioning method, system, equipment and storage medium based on mixed fingerprints | |
CN108696818B (en) | Distance measurement method and system for Wi-Fi fingerprint positioning | |
CN111148030A (en) | Fingerprint database updating method and device, server and storage medium | |
CN111757464A (en) | Region contour extraction method and device | |
CN113079466B (en) | Fingerprint database construction method, device, equipment and computer storage medium | |
CN108225332B (en) | Indoor positioning fingerprint map dimension reduction method based on supervision | |
CN107290714B (en) | Positioning method based on multi-identification fingerprint positioning | |
CN114584230B (en) | Predictive channel modeling method based on countermeasure network and long-term and short-term memory network | |
CN111465093A (en) | Fingerprint positioning method and device, electronic equipment and computer storage medium | |
CN112714493A (en) | Object position positioning method and positioning system | |
CN110611952B (en) | Fingerprint matching and positioning method | |
CN110290466B (en) | Floor distinguishing method, device, equipment and computer storage medium | |
Abdullah et al. | Machine learning algorithm for wireless indoor localization | |
Anuwatkun et al. | A diff-based indoor positioning system using fingerprinting technique and k-means clustering algorithm | |
Fang et al. | Compensating for orientation mismatch in robust Wi-Fi localization using histogram equalization | |
Wei et al. | Handling device heterogeneity in Wi-Fi based indoor positioning systems |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |