CN107831468A - A kind of fingerprinting localization algorithm based on affine propagation clustering - Google Patents
A kind of fingerprinting localization algorithm based on affine propagation clustering Download PDFInfo
- Publication number
- CN107831468A CN107831468A CN201710979485.7A CN201710979485A CN107831468A CN 107831468 A CN107831468 A CN 107831468A CN 201710979485 A CN201710979485 A CN 201710979485A CN 107831468 A CN107831468 A CN 107831468A
- Authority
- CN
- China
- Prior art keywords
- fingerprint
- model
- cluster
- similarity
- signal
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- 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
-
- 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
- G01S5/10—Position of receiver fixed by co-ordinating a plurality of position lines defined by path-difference measurements, e.g. omega or decca systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/12—Fingerprints or palmprints
- G06V40/13—Sensors therefor
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/12—Fingerprints or palmprints
- G06V40/1365—Matching; Classification
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Human Computer Interaction (AREA)
- Multimedia (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Collating Specific Patterns (AREA)
Abstract
The present invention relates to a kind of fingerprinting localization algorithm based on affine propagation clustering, including：Finger print data is gathered, including the location of the intensity of the signal of each wireless signal transmitter transmitting received and gathered data.2) similarity between every two fingerprints, including the similarity of signal value and position are defined.3) data of collection are clustered using the clustering method based on affine propagation, the content of each fingerprint is the average, variance and weighted value that a center clustered and the cluster correspond to Gauss model in fingerprint base.4) tuning on-line.
Description
Technical field
The invention belongs to the indoor positioning field based on fingerprint technique, and affine propagation is used to a large amount of finger print datas collected
Clustering algorithm establishes mixed Gauss model by the higher fingerprint cluster of similarity, then to the fingerprint after cluster, uses mixed Gaussian
Model carries out indoor positioning.
Background technology
In recent years, the acquisition of positional information was increasingly taken seriously, and the new business related to location technology continuously emerges.,
Location Based service (LBS) turns into one of focus of research, and has the extensive market space.
The application field of location technology is quite varied, comprising the various aspects such as business, the disaster relief, military, civilian, the world at present
On the most famous alignment system be the U.S. global positioning system (GPS), GPS by satellite can scope covering the whole world, its base
To be positioned using signal time difference, positioning precision can reach within 10 meters present principles, at present on many countries and ground
It is used widely in area.But indoors under environment, because building walls block to electric wave signal, electric wave can only be with refraction
Or the non-line-of-sight propagation mode such as reflection is propagated, and causes GPS location effect to be deteriorated.
It is current main flow to carry out positioning using wireless signal in the research positioned indoors.Indoor positioning technologies mainly divide
For telemetry and non-ranging method, conventional distance-finding method is that arrival time (TOA), reaching time-difference (TDOA), reception signal are strong
Spend (RSS), reach phase difference (PDOA) etc..Conventional non-ranging method is fingerprint location method, the algorithm bag commonly used in fingerprint technique
Include nearest neighbor method (NN), K nearest neighbor methods (KNN), weighting K nearest neighbor methods (WKNN) etc..
Because RSS is readily available and cost is relatively low, therefore the location fingerprint positioning mode based on RSS is current most popular room
One of interior positioning application.RSS methods can be divided into two kinds of thinkings, and one kind is propagation model method, is estimated by acquired signal feature
The propagation model of signal is counted, the method is stronger by indoor multipath and Environmental Noise Influence.Second is fingerprint location method.Fingerprint is determined
Position method is broadly divided into i.e. offline fingerprint collecting stage in two stages and online real-time positioning stage.The offline fingerprint collecting stage passes through
Each emission source manually is measured to the received signal strength RSS of the fingerprint point at each fingerprint point being determined in advance and is built
One position and RSS location fingerprint database.The RSS of online each emission source that positioning stage is received by user in real time
The location fingerprint database progress matching treatment that it is established with off-line phase is simultaneously obtained the position of current target to be positioned by sequence
Put coordinate.The positioning precision of fingerprint technique receives the influence of fingerprint collecting quantity, and fingerprint collecting point is more intensive, and positioning precision will be got over
It is high.But largely collection fingerprint can increase the cost of storage fingerprint.In the On-line matching stage, due to the fingerprint quantity for needing to match
Increase, amount of calculation increases, and the positioning used time increases.
The content of the invention
It is an object of the invention to provide a kind of carrying cost is relatively low, the less fingerprinting localization algorithm of amount of calculation.Technical scheme
It is as follows：
A kind of fingerprinting localization algorithm based on affine propagation clustering, comprises the following steps：
1) finger print data is gathered, includes the intensity and collection number of the signal of each wireless signal transmitter transmitting received
According to location.
2) similarity between every two fingerprints is defined, this similarity is made up of two parts, and Part I is signal value
Similarity, it is defined as the Euclidean distance sum of the signal intensity of all identical signal sources in two fingerprints；Part II is finger
The similarity of line position, it is defined as the Euclidean distance of two fingerprint collecting positions；This two parts is each multiplied by one and pre-set
Coefficient after the sum that is added, as the similarity between two fingerprints.
3) data of collection are clustered using the clustering method based on affine propagation, using EM algorithm to putting
The attraction matrix and ownership matrix for penetrating propagation clustering algorithm are iterated calculating until restraining or reaching certain iterations, complete
The cluster of paired finger print data.
3) after the completion of cluster, the center per a kind of fingerprint is calculated, uses the weight of all fingerprint coordinates in the cluster
Position of centre of gravity of the heart as cluster, then the mixed Gauss model of fingerprint is calculated, per a kind of fingerprint, a corresponding Gauss model, needs to count
The average and variance and weight of this model are calculated, it is each in last fingerprint base using EM algorithm calculating three above value
The content of bar fingerprint is the average, variance and weighted value that a center clustered and the cluster correspond to Gauss model.
4) signal intensity is substituted into each group of Gauss model and calculates one by tuning on-line phase acquisition to one group of signal intensity
Individual probable value, the probability that each model obtains is successively multiplied by with the centre coordinate of the weight of the model and model, by each mould
The value that type obtains is added as last position location.
The present invention is clustered using the clustering method based on affine propagation to the data of collection, to the fingerprint number after cluster
According to its mixed Gauss model is calculated, fingerprint base is stored in using the parameter of mixed Gauss model as new fingerprint, reduces depositing for fingerprint
Reserves.When being positioned to target, the data gathered in real time are substituted into each sub- Gauss model of mixed Gauss model generation, obtained each
The weight of individual sub- Gauss model, then the center that each sub- Gauss model correspond to fingerprint cluster, Suo Youzi are multiplied by respectively with this weight
The result and target location finally to position of model.
Brief description of the drawings
Fig. 1 shows the FB(flow block) of holistic approach of the present invention.
Embodiment
1) finger print data, data include the strong of the signal of each wireless signal transmitter transmitting received in large quantities for collection
The location of degree and gathered data.
2) data of collection are clustered using the clustering method based on affine propagation, using affine propagation clustering algorithm
Similarity of the definition in clustering between each fingerprint is needed, this similarity is made up of two parts, and Part I is the phase of signal value
Like degree, the Euclidean distance sum of the signal intensity of all identical signal sources in two fingerprints is defined as；Part II is fingerprint
The similarity of position, it is defined as the Euclidean distance of two fingerprint collecting positions.This two parts is each multiplied by one and pre-set
The sum being added after coefficient, as the similarity between two fingerprints.Using EM algorithm to radiation propagation clustering algorithm
Attract matrix and ownership matrix to be iterated calculating until restraining or reaching certain iterations, complete to gather finger print data
Class.
3) after the completion of cluster, the center per a kind of fingerprint is calculated, uses the weight of all fingerprint coordinates in the cluster
Position of centre of gravity of the heart as cluster.The mixed Gauss model of fingerprint is calculated again, and per a kind of fingerprint, a corresponding Gauss model, needs to count
The average and variance and weight of this model are calculated, it is each in last fingerprint base using EM algorithm calculating three above value
The content of bar fingerprint is the average, variance and weighted value that a center clustered and the cluster correspond to Gauss model.
4) signal intensity is substituted into each group of Gauss model and calculates one by tuning on-line phase acquisition to one group of signal intensity
Individual probable value, the probability that each model obtains is successively multiplied by with the centre coordinate of the weight of the model and model, by each mould
The value that type obtains is added as last position location.
Claims (1)
1. a kind of fingerprinting localization algorithm based on affine propagation clustering, comprises the following steps：
1) finger print data is gathered, including the intensity of the signal of each wireless signal transmitter transmitting received and gathered data institute
The position at place.
2) similarity between every two fingerprints is defined, this similarity is made up of two parts, and Part I is similar for signal value
Degree, it is defined as the Euclidean distance sum of the signal intensity of all identical signal sources in two fingerprints；Part II is fingerprint bit
The similarity put, it is defined as the Euclidean distance of two fingerprint collecting positions；This two parts be each multiplied by one pre-set be
The sum being added after number, as the similarity between two fingerprints；
3) data of collection are clustered using the clustering method based on affine propagation, radiation passed using EM algorithm
Broadcast the attraction matrix of clustering algorithm and ownership matrix is iterated calculating until restraining or reaching certain iterations, completion pair
The cluster of finger print data；
3) after the completion of cluster, the center per a kind of fingerprint is calculated, is made using the center of gravity of all fingerprint coordinates in the cluster
For the position of centre of gravity of cluster, then the mixed Gauss model of fingerprint is calculated, a corresponding Gauss model, need to calculate this per a kind of fingerprint
The average and variance and weight of model, three above value is calculated using EM algorithm, each refers in last fingerprint base
The content of line is the average, variance and weighted value that a center clustered and the cluster correspond to Gauss model；
4) signal intensity is substituted into each group of Gauss model and calculates one generally by tuning on-line phase acquisition to one group of signal intensity
Multiplied by with the centre coordinate of the weight of the model and model, each model is obtained successively for rate value, the probability that each model obtains
To value be added as last position location.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710979485.7A CN107831468A (en) | 2017-10-19 | 2017-10-19 | A kind of fingerprinting localization algorithm based on affine propagation clustering |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710979485.7A CN107831468A (en) | 2017-10-19 | 2017-10-19 | A kind of fingerprinting localization algorithm based on affine propagation clustering |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107831468A true CN107831468A (en) | 2018-03-23 |
Family
ID=61648498
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710979485.7A Pending CN107831468A (en) | 2017-10-19 | 2017-10-19 | A kind of fingerprinting localization algorithm based on affine propagation clustering |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107831468A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2022126373A1 (en) * | 2020-12-15 | 2022-06-23 | Nokia Technologies Oy | Enhanced fingerprint positioning |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102333372A (en) * | 2011-09-15 | 2012-01-25 | 中国科学院计算技术研究所 | Real-time positioning method and system based on radio frequency fingerprints |
| CN103634902A (en) * | 2013-11-06 | 2014-03-12 | 上海交通大学 | Novel indoor positioning method based on fingerprint cluster |
| 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 |
| CN107077617A (en) * | 2017-01-25 | 2017-08-18 | 深圳市汇顶科技股份有限公司 | fingerprint extraction method and device |
-
2017
- 2017-10-19 CN CN201710979485.7A patent/CN107831468A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102333372A (en) * | 2011-09-15 | 2012-01-25 | 中国科学院计算技术研究所 | Real-time positioning method and system based on radio frequency fingerprints |
| CN103634902A (en) * | 2013-11-06 | 2014-03-12 | 上海交通大学 | Novel indoor positioning method based on fingerprint cluster |
| 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 |
| CN107077617A (en) * | 2017-01-25 | 2017-08-18 | 深圳市汇顶科技股份有限公司 | fingerprint extraction method and device |
Non-Patent Citations (4)
| Title |
|---|
| QIDENG JIANG ET AL.: "A Probabilistic Radio Map Construction Scheme for Crowdsourcing-Based Fingerprinting Localization", 《IEEE SENSORS JOURNAL》 * |
| 张俪文 等: "仿射传播聚类在室内定位指纹库中的应用研究", 《测绘通报》 * |
| 王燕科: "WAVE协议栈下的无线定位研究与实现", 《中国优秀硕士学位论文全文数据库工程科技Ⅱ辑》 * |
| 蔡海尼: "《基于语义的Web服务集成研究及其应用》", 31 March 2015 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2022126373A1 (en) * | 2020-12-15 | 2022-06-23 | Nokia Technologies Oy | Enhanced fingerprint positioning |
| US11914059B2 (en) | 2020-12-15 | 2024-02-27 | Nokia Technologies Oy | Enhanced fingerprint positioning |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108696932B (en) | Outdoor fingerprint positioning method using CSI multipath and machine learning | |
| CN110933599B (en) | Self-adaptive positioning method fusing UWB and WIFI fingerprints | |
| CN105044662B (en) | A kind of fingerprint cluster multi-point joint indoor orientation method based on WIFI signal intensity | |
| CN105242239B (en) | A kind of indoor subregion localization method based on the sub-clustering of mass-rent fingerprint and matching | |
| Yan et al. | Review of range-based positioning algorithms | |
| CN104602342B (en) | A kind of high efficiency indoor localization method based on iBeacon equipment | |
| CN101695152B (en) | Indoor positioning method and system thereof | |
| CN109275095A (en) | A kind of indoor locating system based on bluetooth, positioning device and localization method | |
| Torres-Sospedra et al. | A realistic evaluation of indoor positioning systems based on Wi-Fi fingerprinting: The 2015 EvAAL–ETRI competition | |
| CN106559750B (en) | Passive type localization method in multiple target room based on passive ultra-high frequency RFID | |
| CN102209382A (en) | Wireless sensor network node positioning method based on received signal strength indicator (RSSI) | |
| CN105137390A (en) | Indoor positioning method based on AP with adjustable transmitted power | |
| CN105916201A (en) | KNN fingerprint positioning method based on RSS Euclidean distance fitting | |
| CN109672973A (en) | A kind of indoor positioning blending algorithm based on most strong AP method | |
| CN112748397A (en) | UWB positioning method based on self-adaptive BP neural network under non-line-of-sight condition | |
| CN103997717A (en) | Real-time indoor positioning system and method | |
| CN110082716B (en) | Indoor positioning system and positioning method for complex environment of hospital | |
| CN110536245A (en) | A kind of indoor wireless positioning method and system based on deep learning | |
| AU2020203007B2 (en) | A method of setting-up a range-based tracking system utilising a tracking coordinate system | |
| CN110351660B (en) | Bluetooth indoor positioning method based on double-step fingerprint matching architecture | |
| CN109655790A (en) | Multi-target detection and identification system and method based on indoor LED light source | |
| CN102967848B (en) | Positioning method based on distance relationship library and received signal intensity | |
| CN108732534A (en) | A kind of multi-tag Cooperative Localization Method based on weighting MDS | |
| Yilmaz et al. | Indoor positioning using visual and inertial sensors | |
| CN103698740A (en) | Weighted value selecting-based WIFI (wireless fidelity) indoor positioning method |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180323 |
|
| RJ01 | Rejection of invention patent application after publication |