CN107567093A - A kind of fingerprint positioning method based on healthy and strong rarefaction representation cluster - Google Patents
A kind of fingerprint positioning method based on healthy and strong rarefaction representation cluster Download PDFInfo
- Publication number
- CN107567093A CN107567093A CN201710770981.1A CN201710770981A CN107567093A CN 107567093 A CN107567093 A CN 107567093A CN 201710770981 A CN201710770981 A CN 201710770981A CN 107567093 A CN107567093 A CN 107567093A
- Authority
- CN
- China
- Prior art keywords
- fingerprint
- equal
- rsrc
- intensity
- test point
- 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
Landscapes
- Collating Specific Patterns (AREA)
- Position Fixing By Use Of Radio Waves (AREA)
Abstract
A kind of fingerprint positioning method based on healthy and strong rarefaction representation cluster, mainly solves in the prior art that positioning precision is inadequate, and acquisition time is long, the problems such as robustness deficiency.The sparse prior information of the invention for mainly utilizing test point, its core be using RSRC algorithms to the finger print data of test point after cluster sparse table goes out under fingerprint base, so as to realize the purpose being accurately positioned, offline fingerprint base is built first, then fingerprint base is optimized by affine propagation clustering, on-line stage is determined candidate's class, realizes the estimation to test position using RSRC algorithms to test point fingerprint intensity rarefaction representation, minimization amendment residual error.Experiment shows that the present invention has more preferable locating effect compared with currently existing scheme.
Description
Technical Field
The invention belongs to the field of wireless positioning, and particularly relates to a fingerprint positioning method based on robust sparse representation clustering.
Background
With the development of WLAN, a mode of acquiring location information by means of communication devices such as smart phones, PADs, and notebook computers is increasingly popular. Location-based services have wide applications in the fields of medical treatment, emergency treatment, navigation and the like, and an accurate positioning technology is the key point of the location-based services. Therefore, the method has great theoretical value and market application prospect for the research of the positioning technology.
Due to the popularization of wifi equipment, the fingerprint strength of the received signals can be utilized for positioning, other sensors are not needed, the cost is low, the operation is convenient, and a positioning algorithm based on fingerprints becomes mainstream. The existing positioning algorithm based on fingerprints needs a large amount of off-line training, is sensitive to noise data, and simultaneously needs to be further improved in positioning accuracy.
Disclosure of Invention
In order to overcome the defects of the prior art, the invention aims to provide a fingerprint positioning method based on robust sparse representation clustering, based on which, the RSRC algorithm is applied to fingerprint positioning, so that the positioning precision is improved
In order to achieve the purpose, the invention adopts the technical scheme that:
a fingerprint positioning method based on robust sparse representation clustering comprises the following steps:
1) Processing the off-line sampled data to obtain an original database
Wherein R is ij The intensity of the fingerprint received by the jth sensor by the ith access point is represented, i is more than or equal to 1 and less than or equal to M, j is more than or equal to 1 and less than or equal to N, M represents the number of offline sampling points, and N represents the number of the sensors in the sampling area.
2) Obtaining optimized fingerprint library through affine propagation clustering
Wherein S ab Representing the intensity of the fingerprint received by the b sensor at the center of the a-th class, a is more than or equal to 1 and less than or equal to P, b is more than or equal to 1 and less than or equal to N, and P represents the number of classes。
3) And (3) utilizing sparse priori information of the test points, sparsely listing out the fingerprint intensity of the test points through an RSRC algorithm, correcting residual errors to the minimum degree, determining candidate class centers, and estimating the positions of the test points.
Since data can be influenced by noise in the actual test process, the influence of noise needs to be considered if positioning is accurate. Wherein y, y 0 ,e 0 ,t 0 And respectively representing the fingerprint intensity of the actual test point, the fingerprint intensity of the test point after denoising, the noise and the sparse table coefficient of the fingerprint intensity of the pure test point under the fingerprint library phi. Then the
Solving equation y = Φ by RSRC algorithm 1 t 1 To obtain t 0 And (4) minimizing the corrected residual error to determine candidate class centers and estimating the positions of the test points.
RSRC solving:
(1) solving for y = Φ using sparse representation algorithm 1 t 1 To obtain t 0 ,e 0
(2) Calculating the residual error after the thinning table is extracted after correction
③r ξ And (y) arranging in a descending order, and weighting the coordinates of the lower class centers of the first plurality of corresponding catalogs to output the position estimation value of the test point.
The invention has the beneficial effects that:
1) The method estimates the positions of the test points by using the sparse prior information of the test points through the RSRC algorithm, and has higher positioning accuracy compared with the existing scheme.
2) The invention takes the influence of noise into consideration and has stronger robustness compared with the prior scheme.
3) The method is based on fingerprint intensity positioning, and is suitable for RSSI positioning scenes based on WIFI, bluetooth and RFID.
Drawings
Fig. 1 is a schematic diagram of a signal flow structure according to the present invention.
FIG. 2 is a graph showing the cumulative distribution of RSRC, SRC, WKNN, KNN, NN errors in the present invention.
FIG. 3 is a graph of the average error, the maximum error, and the minimum error of RSRC, SRC, WKNN, KNN, NN in the present invention.
Detailed Description
The present invention will be further described with reference to the following examples.
Taking fig. 1 as an example, a robust sparse representation clustering-based fingerprint positioning method includes the following steps:
1) Processing the off-line sampled data to obtain an original database
Wherein R is ij The intensity of the fingerprint received by the jth sensor by the ith access point is represented, i is more than or equal to 1 and less than or equal to M, j is more than or equal to 1 and less than or equal to N, M represents the number of offline sampling points, and N represents the number of sensors in the sampling area.
2) Obtaining optimized fingerprint library through affine propagation clustering
Wherein S ab The fingerprint intensity of the b sensor received by the a-th class center is represented, a is more than or equal to 1 and less than or equal to P, b is more than or equal to 1 and less than or equal to N, and P represents the number of classes.
3) And (3) by using sparse prior information of the test points, the fingerprint intensity of the test points is obtained through a sparse table of an RSRC algorithm, the residual error is corrected to be minimized to determine candidate class centers, and the positions of the test points are estimated.
Since data can be influenced by noise in the actual test process, the influence of the noise needs to be considered if positioning is accurate. Wherein y, y 0 ,e 0 ,t 0 And respectively representing the actual test point fingerprint intensity, the denoised test point fingerprint intensity, the noise and the sparsity of the pure test point fingerprint intensity under a fingerprint library phi to obtain coefficients. Then
Solving the equation y = Φ by RSRC algorithm 1 t 1 To obtain t 0 And (4) minimizing the corrected residual error to determine candidate class centers and estimating the positions of the test points.
RSRC solving:
(1) solving for y = Φ using sparse representation algorithm 1 t 1 To obtain t 0 ,e 0
(2) Calculating the residual error after the sparse table is obtained after correction
③r ξ And (y) arranging in a descending order, and weighting the coordinates of the lower class centers of the first plurality of corresponding catalogs to output the position estimation value of the test point.
FIG. 1 is a schematic diagram of a system structure, in which an offline fingerprint library is obtained by processing offline sampling data as shown in FIG. 1, the fingerprint library is optimized by affine propagation clustering, the fingerprint intensity of a test signal is obtained by sparse table in an online stage, candidate class centers are determined by minimizing a modified residual error, and the position of the test point is estimated
FIG. 2 shows the cumulative distribution of RSRC, SRC, WKNN, KNN, and NN errors, where the cumulative distribution curve of RSRC is closer to the left, which shows that the positioning accuracy of RSRC is better than that of SRC, WKNN, KNN, and NN
Fig. 3 shows the average error, maximum error, and minimum error of RSRC, SRC, WKNN, KNN, NN, where RSRC < SRC < WKNN < KNN < NN, RSRC < WKNN < NN, maximum error RSRC < SRC < WKNN < KNN < NN, and minimum error KNN < RSRC = SRC = WKNN < NN. The combination of the indexes shows that the positioning effect of RSRC is superior to that of SRC, WKNN, KNN and NN.
Claims (1)
1. A fingerprint positioning method based on robust sparse representation clustering is characterized by comprising the following steps:
1) Processing the off-line sampled data to obtain an original database
Wherein R is ij The intensity of the fingerprint received by the jth sensor by the ith access point is represented, i is more than or equal to 1 and less than or equal to M, j is more than or equal to 1 and less than or equal to N, M represents the number of offline sampling points, and N represents the number of sensors in the sampling area;
2) Obtaining an optimized fingerprint library through affine propagation clustering
Wherein S ab Representing the intensity of the fingerprint received by the b-th sensor by the a-th class center, wherein a is more than or equal to 1 and less than or equal to P, b is more than or equal to 1 and less than or equal to N, and P represents the number of classes;
3) Using sparse prior information of the test points, sparsely listing the fingerprint intensity of the test points through an RSRC algorithm, minimizing a correction residual error to determine candidate class centers, and estimating the positions of the test points;
since data can be influenced by noise in the actual test process, the influence of the noise needs to be considered if positioning is accurate, wherein y, y 0 ,e 0 ,t 0 Respectively representing the actual test point fingerprint intensity, the test point fingerprint intensity after denoising, the noise and the sparse of the pure test point fingerprint intensity under the fingerprint library phi, and then
Solving the equation y = Φ by RSRC algorithm 1 t 1 To obtain t 0 The residual error is corrected to be minimized to determine a candidate class center, and the position of a test point is estimated;
RSRC solving:
(1) solving for y = Φ using sparse representation algorithm 1 t 1 To obtain t 0 ,e 0
(2) Calculating the residual error after the thinning table is extracted after correction
③r ξ And (y) arranging in a descending order, and weighting the coordinates of the lower class centers of the first plurality of corresponding catalogs to output the position estimation value of the test point.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710770981.1A CN107567093B (en) | 2017-08-31 | 2017-08-31 | Fingerprint positioning method based on robust sparse representation clustering |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710770981.1A CN107567093B (en) | 2017-08-31 | 2017-08-31 | Fingerprint positioning method based on robust sparse representation clustering |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107567093A true CN107567093A (en) | 2018-01-09 |
CN107567093B CN107567093B (en) | 2020-11-13 |
Family
ID=60978416
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710770981.1A Active CN107567093B (en) | 2017-08-31 | 2017-08-31 | Fingerprint positioning method based on robust sparse representation clustering |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107567093B (en) |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105372628A (en) * | 2015-11-19 | 2016-03-02 | 上海雅丰信息科技有限公司 | Wi-Fi-based indoor positioning navigation method |
-
2017
- 2017-08-31 CN CN201710770981.1A patent/CN107567093B/en active Active
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105372628A (en) * | 2015-11-19 | 2016-03-02 | 上海雅丰信息科技有限公司 | Wi-Fi-based indoor positioning navigation method |
Non-Patent Citations (5)
Title |
---|
CHEN FENG: "compressive sensing based positioning using RSS of WLAN access points", 《IEEE INFOCOM 2010》 * |
CHEN FENG: "Received-Signal-Strength-Based Indoor Positioning Using Compressive Sensing", 《IEEE TRANSACTIONS ON MOBILE COMPUTING》 * |
XIAOLU ZHU: "A Joint Grid Segmentation based Affinity Propagation Clustering Method for Big Data", 《IEEE 2ND INTERNATIONAL CONFERENCE ON DATA SCIENCE AND SYSTEMS》 * |
丁根明: "基于人工智能的室内指纹定位技术研究", 《北京交通大学博士学位论文》 * |
陈宝: "目标数目未知下的压缩感知定位算法", 《南京邮电大学硕士学位论文》 * |
Also Published As
Publication number | Publication date |
---|---|
CN107567093B (en) | 2020-11-13 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109444813B (en) | RFID indoor positioning method based on BP and DNN double neural networks | |
CN110381440A (en) | The fingerprint indoor orientation method of joint RSS and CSI based on deep learning | |
CN105954712B (en) | The direct localization method of the multiple target of associated wireless electric signal complex envelope and carrier phase information | |
CN103369466B (en) | A kind of map match assists indoor orientation method | |
CN107703480A (en) | Mixed kernel function indoor orientation method based on machine learning | |
CN104394588B (en) | Indoor orientation method based on Wi Fi fingerprints and Multidimensional Scaling | |
CN106991705B (en) | Position parameter estimation method based on P3P algorithm | |
CN108495263B (en) | Indoor positioning method based on generalized multi-core function | |
CN108197564A (en) | A kind of assessment system and method for drawing clock experiment | |
CN112966711A (en) | Pointer instrument indicating number identification method and system based on convolutional neural network | |
CN110596688A (en) | High-precision positioning method and device in distributed MIMO radar system | |
Yu et al. | An indoor localization of WiFi based on support vector machines | |
CN110057355B (en) | Indoor positioning method, device and system and computing equipment | |
CN110909804A (en) | Method, device, server and storage medium for detecting abnormal data of base station | |
CN113189541B (en) | Positioning method, device and equipment | |
CN111885703B (en) | Indoor positioning method | |
CN107567093B (en) | Fingerprint positioning method based on robust sparse representation clustering | |
CN113543026A (en) | Multi-floor indoor positioning method based on radial basis function network | |
CN109858517B (en) | Track similarity measurement method taking motion direction as dominant | |
CN112333652B (en) | WLAN indoor positioning method and device and electronic equipment | |
CN111914109A (en) | Depth measurement learning-based garment retrieval technology | |
WO2017049914A1 (en) | Terminal positioning method, apparatus, and system | |
CN113390421B (en) | Unmanned aerial vehicle positioning method and device based on Kalman filtering | |
CN115797616A (en) | Image correction point acquisition method and device, electronic equipment and medium | |
CN114047475A (en) | Iterative positioning method for arrival time difference based on multi-dimensional scaling simplified cost function |
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 |