CN110007269A - A kind of two stages wireless signal fingerprint positioning method based on Gaussian process - Google Patents
A kind of two stages wireless signal fingerprint positioning method based on Gaussian process Download PDFInfo
- Publication number
- CN110007269A CN110007269A CN201910272922.0A CN201910272922A CN110007269A CN 110007269 A CN110007269 A CN 110007269A CN 201910272922 A CN201910272922 A CN 201910272922A CN 110007269 A CN110007269 A CN 110007269A
- Authority
- CN
- China
- Prior art keywords
- signal
- reference point
- signal strength
- variance
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 48
- 238000005259 measurement Methods 0.000 claims description 13
- 238000012549 training Methods 0.000 claims description 12
- 239000011159 matrix material Substances 0.000 claims description 8
- 238000009826 distribution Methods 0.000 claims description 5
- 230000004807 localization Effects 0.000 claims description 4
- 230000008901 benefit Effects 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 5
- 238000012423 maintenance Methods 0.000 description 4
- 241000208340 Araliaceae Species 0.000 description 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 1
- 235000003140 Panax quinquefolius Nutrition 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 235000008434 ginseng Nutrition 0.000 description 1
- 230000005389 magnetism Effects 0.000 description 1
- 238000004321 preservation Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
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
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (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 kind of two stages wireless signal fingerprint positioning method based on Gaussian process, generates highdensity virtual data base based on sparse tranining database using Gaussian process;User's approximate range is determined using tranining database first in position fixing process, is then accurately positioned using highdensity virtual fingerprint database, to improve the positioning accuracy of indoor environment.
Description
Technical field
The invention belongs to cordless communication network technical fields, are related to a kind of wireless signal fingerprint positioning method, and in particular to
A kind of two stages wireless signal fingerprint positioning method based on Gaussian process.
Background technique
As navigator fix service gos deep into each fields such as people's life, military affairs, finance and people to navigator fix demand
Raising, GNSS leads to the further development for seriously constraining navigator fix industry for the deficiency of indoor environment location.People
A large amount of new location technologies are developed, positioning, infrared positioning, acoustic location, WLAN including camera position, RFID is positioned,
Light positioning, Zigbee positioning, UWB positioning, pseudolite positioning, inertial positioning, Magnetic oriented and combination various technologies and sensor
Integrated positioning etc., these location technologies have its scope of application, and there is no a kind of effective low-power consumption, high-precision so far
Degree, towards common hardware, is applicable in the location technology of indoor environment.With popularizing for wireless network, received based on wireless network
The indoor positioning technologies of received signals fingerprint have the advantages such as required hardware is widely present, positioning accuracy is ideal, obtain and close extensively
Note.
Wireless signal fingerprint location mainly includes two steps: off-line training and tuning on-line.The work of off-line training step
Work is creation received signals fingerprint database, that is, the reference point (Reference Point, RP) in different location is collected
Received signals fingerprint information preservation in the database, radio signal source can be Wi-Fi, GSM, FM, DTM or earth magnetism field signal
Deng received signals fingerprint can be signal strength, signal distributions or signal variance etc..The tuning on-line stage currently measures user
To received signals fingerprint and database in received signals fingerprint compare, user should be at the corresponding reference of most matched received signals fingerprint
Point position.With popularizing for WiFi equipment, being based on common smart mobile phone based on Wi-Fi wireless signal fingerprint location can be real
It is existing, have many advantages, such as that deployment cost is low, positioning accuracy is higher, has there is many commercial products, such as Google Map at present
Indoor, WiFiSlam and Rtmap etc..
Indoor position accuracy based on received signals fingerprint depends in received signals fingerprint database with reference to dot density and data
Library timeliness.Identical a piece of area reference dot density is higher, then positions more accurate;More frequent, the positioning accuracy of database update
It is higher.But creation needs higher cost with maintenance high density signal fingerprint database, and is sometimes difficult to.For example,
To create received signals fingerprint database for the room area of 10m*10m, it is 1/square meter with reference to dot density, then needs to acquire altogether
The received signals fingerprint information of 100 reference points.In each reference point, require to be measured for several times to obtain reliable signal and refer to
Line.For bigger region, reference point quantity will be with exponential increase.For some regions, when measurement, is likely difficult to reach, to make
At the missing of received signals fingerprint.Therefore, building high density signal fingerprint database is generally difficult in reality.In addition, in order to keep
The timeliness of database needs periodically to be updated, and maintenance cost is equally very huge, causes many commercial products in reality
In it is unavailable.For example, the received signals fingerprint data acquired more than 100,000 venues are claimed by Google company, but actually using
When, only small part venue can position, and Rtmap actual measurement positioning accuracy is also far from reaching theoretical precision.
Even if having had been built up highdensity database, other problems can be equally brought in position fixing process.User is fixed every time
When position, current received signals fingerprint can be all measured, and send it to location server.Server will be in user fingerprints and database
All data matched, find most like reference point, higher with reference to dot density, database size is huger, finds
Journey time-consuming is longer.When having a large number of users while carrying out position requests, positioning real-time will be affected.
In short, wireless signal fingerprint location has many advantages, such as that deployment is easy, algorithm is simple, but there is also high density fingerprints
Database sharing and the problem that maintenance cost is high, positioning time is long.
Summary of the invention
In order to solve the above-mentioned technical problem, the two stages received signals fingerprint positioning based on Gaussian process that the invention proposes a kind of
Method.
The technical scheme adopted by the invention is that: a kind of two stages wireless signal fingerprint location side based on Gaussian process
Method, which comprises the following steps:
Step 1: the virtual data base building based on Gaussian process;
Highdensity virtual fingerprint database is generated based on sparse tranining database using Gaussian process;
Step 2: two stages wireless signal fingerprint location;
User's approximate range is determined using tranining database first, is then carried out using highdensity virtual fingerprint database
It is accurately positioned, to improve the positioning accuracy of indoor environment.
The present invention has the advantage that
(1) database sharing cost substantially reduces, can be quick from step 1 as can be seen that the present invention is based on Gaussian processes
High density signal fingerprint database is generated, point-to-point measurement is not needed, greatly improves database sharing efficiency.
(2) database sharing maintenance substantially reduces, and from step 1 as can be seen that the present invention is based on Gaussian processes, can be based on
A small amount of test data quickly updates received signals fingerprint database, does not need point-to-point measurement, greatly improve database update efficiency.
(3) positioning time greatly shortens, from step 2 as can be seen that the present invention passes through two steps --- and the first step is based on
Tranining database reduces search range, and second step uses high density virtual data base to be accurately positioned --- obtain family position with
And location estimation variance, it does not need directly high density data library to be used to be positioned, greatly shortens positioning time.
Detailed description of the invention
Fig. 1 is the flow chart of the embodiment of the present invention;
Fig. 2 is to create range S locating for user's possibility based on TS in the embodiment of the present inventiontSchematic diagram.
Specific embodiment
Understand for the ease of those of ordinary skill in the art and implement the present invention, with reference to the accompanying drawings and embodiments to this hair
It is bright to be described in further detail, it should be understood that implementation example described herein is merely to illustrate and explain the present invention, not
For limiting the present invention.
Referring to Fig.1, a kind of two stages wireless signal fingerprint positioning method based on Gaussian process provided by the invention, including
Following steps:
Step 1: the virtual data base building based on Gaussian process;
Highdensity virtual fingerprint database is generated based on sparse tranining database using Gaussian process;
Assuming that reference point signal strength similar in position has correlation, it may be assumed that
Wherein, xi,xjRespectively indicate i, the coordinate of j point, wherein xi=(ai,bi),xj=(aj,bj);yi,yjIt respectively indicates
Signal strength at i, j point;It is signal strength measurement variance, δijFor Kronecker delta function, it is defined as follows:
k(xi,xj) it is kernel function, most popular kernel function is gaussian kernel function:
In formula (3)It is signal variance and scale coefficient respectively with l, the two parameters determine position approximated reference point
The degree of correlation of signal strength;
Give one group of training data (X, Y)={ (xi,yi) | i=1,2 ..., n }, based in Gaussian process estimation space
Anticipate a point x signal strength y probability distribution:
It is μ that signal strength y, which obeys mean value,x, variance beNormal distribution, wherein mean μxAre as follows:
Wherein, K is n × n matrix,It is signal strength measurement variance, I is unit matrix, kxFor the vector of n × 1, vector
In element be respectively all the points in unknown point x and training data related coefficient:
kx(i, 0)=k (xi,x),xi∈X (6)
The variance of signal strength yAre as follows:
Wherein, K is n × n matrix, correlation of the element between training data in matrix:
K (i, j)=k (xi,xj),xi,xj∈X (8)
Assuming that localization region area is S, by artificial method, the signal strength of each reference point of point-to-point measurement will join
Examination point coordinate is associated with signal strength, forms tranining database DB(tr), need to construct high density virtual data base DB(v), reference
Dot density is ρ(v), ρ(v)It is any to choose;
DB(tr)In training data be (X, Y)={ (xi,yi) | i=1,2 ..., n }, first with density p in the S of region(v)
Uniform design reference point, subsequently for all reference points, using formula (5) estimated signal strength, using formula (7) estimation side
Difference, in calculating process, parameterEstimated with l using conventional hyper parameter method;So far, just it is based on Gauss mistake
Journey, in tranining database DB(tr)On the basis of, it is established that high density virtual data base DB(v)。
Step 2: two stages wireless signal fingerprint location;
User's approximate range is determined using tranining database first, is then carried out using highdensity virtual fingerprint database
It is accurately positioned, to improve the positioning accuracy of indoor environment.
Assuming that the received signals fingerprint that user's t moment measures is RSSt, RSSt={ RSSt,i, i=1,2 ..., b }, wherein
RSSt,iIndicate the signal strength for i-th of signal source that t is measured, b indicates signal source sum;
In the first stage, tranining database DB is used first(tr)Coarse localization is carried out, RSS is calculatedtWith DB(tr)In all ginsengs
The signal distance of examination point:
Wherein, q is coefficient, used here as manhatton distance, i.e. q=1,For j-th of reference in tranining database
The signal strength for i-th of signal source that point measures;
The signal and tranining database DB of user's measurement are calculated using formula (9)(tr)In all reference points signal away from
From selecting wherein apart from the smallest k reference point (k is the integer greater than 1) composing training collection TS;
Enter second stage below, see Fig. 2, based on range S locating for TS creation user's possibilityt:
In position fixing process, to high density virtual data base DB(v)Middle StAll reference points in range are calculated and are measured with user
The distance of signal:
Wherein, δj,iIndicate high density virtual data base DB(v)In j-th of reference point signal estimate variance, by formula (7)
It calculates;For high density virtual data base DB(v)In the signal strength of i-th of signal source that measures of j-th of reference point,
It is calculated using formula (5);It is the estimation side for increasing reference point that distance in formula (11), which is calculated with tradition KWNN difference,
Difference: reference point estimate variance is bigger, indicates that the reference point signal strength error is larger, then being selected probability accordingly reduces;
It selects wherein apart from the smallest k reference point (k is the integer greater than 1), user location is calculated using following formula:
Wherein, For the weight of reference point j, calculate as follows:
dt,j pIndicate dt,jP power, p is adjustable parameter, pass through adjust p obtain best estimate;
The variance of location estimation are as follows:
Family position and location estimation variance are obtained, position fixing process is completed.
It should be understood that the part that this specification does not elaborate belongs to the prior art.
It should be understood that the above-mentioned description for preferred embodiment is more detailed, can not therefore be considered to this
The limitation of invention patent protection range, those skilled in the art under the inspiration of the present invention, are not departing from power of the present invention
Benefit requires to make replacement or deformation under protected ambit, fall within the scope of protection of the present invention, this hair
It is bright range is claimed to be determined by the appended claims.
Claims (3)
1. a kind of two stages wireless signal fingerprint positioning method based on Gaussian process, which comprises the following steps:
Step 1: the virtual data base building based on Gaussian process;
Highdensity virtual fingerprint database is generated based on sparse tranining database using Gaussian process;
Step 2: two stages wireless signal fingerprint location;
User's approximate range is determined using tranining database first, is then carried out using highdensity virtual fingerprint database accurate
Positioning, to improve the positioning accuracy of indoor environment.
2. the two stages wireless signal fingerprint positioning method according to claim 1 based on Gaussian process, which is characterized in that
The specific implementation process of step 1 is:
Assuming that reference point signal strength similar in position has correlation, it may be assumed that
Wherein, xi,xjRespectively indicate i, the coordinate of j point, wherein xi=(ai,bi),xj=(aj,bj);yi,yjRespectively indicate i, j point
Locate signal strength;It is signal strength measurement variance, δijFor Kronecker delta function, it is defined as follows:
k(xi,xj) it is kernel function, most popular kernel function is gaussian kernel function:
In formula (3)It is signal variance and scale coefficient respectively with l, the two parameters determine position approximated reference point signal
The degree of correlation of intensity;
Give one group of training data (X, Y)={ (xi,yi) | i=1,2 ..., n }, based on any one in Gaussian process estimation space
The probability distribution of the signal strength y of point x:
It is μ that signal strength y, which obeys mean value,x, variance beNormal distribution, wherein mean μxAre as follows:
Wherein, K is n × n matrix,It is signal strength measurement variance, I is unit matrix, kxFor the vector of n × 1, in vector
Element is the related coefficient of all the points in unknown point x and training data respectively:
kx(i, 0)=k (xi,x),xi∈X (6)
The variance of signal strength yAre as follows:
Wherein, K is n × n matrix, correlation of the element between training data in matrix:
K (i, j)=k (xi,xj),xi,xj∈X (8)
Assuming that localization region area is S, and by artificial method, the signal strength of each reference point of point-to-point measurement, by reference point
Coordinate is associated with signal strength, forms tranining database DB(tr), need to construct high density virtual data base DB(v), reference point is close
Degree is ρ(v), ρ(v)It is any to choose;
DB(tr)In training data be (X, Y)={ (xi,yi) | i=1,2 ..., n }, first with density p in the S of region(v)Uniformly
Reference point is selected, subsequently for all reference points, using formula (5) estimated signal strength, using formula (7) estimate variance,
In calculating process, parameterEstimated with l using conventional hyper parameter method;So far, it is just based on Gaussian process,
Tranining database DB(tr)On the basis of, it is established that high density virtual data base DB(v)。
3. the two stages wireless signal fingerprint positioning method according to claim 2 based on Gaussian process, which is characterized in that
The specific implementation process of step 2 is:
Assuming that the received signals fingerprint that user's t moment measures is RSSt, RSSt={ RSSt,i, i=1,2 ..., b }, wherein RSSt,iTable
Show the signal strength for i-th of signal source that t is measured, b indicates signal source sum;
In the first stage, tranining database DB is used first(tr)Coarse localization is carried out, RSS is calculatedtWith DB(tr)In all reference points
Signal distance:
Wherein, q is coefficient, used here as manhatton distance, i.e. q=1,For j-th of reference point measurement in tranining database
The signal strength of i-th of the signal source arrived;
The signal and tranining database DB of user's measurement are calculated using formula (9)(tr)In all reference points signal distance, choosing
Selecting wherein is the integer greater than 1 apart from the smallest k reference point composing training collection TS, k;
Enter second stage below, based on range S locating for TS creation user's possibilityt:
In position fixing process, to high density virtual data base DB(v)Middle StAll reference points in range calculate and user's measuring signal
Distance:
Wherein, δj,iIndicate high density virtual data base DB(v)In j-th of reference point signal estimate variance, by formula (7) count
It calculates;For high density virtual data base DB(v)In the signal strength of i-th of signal source that measures of j-th of reference point, adopt
It is calculated with formula (5);It is the estimate variance for increasing reference point that distance in formula (11), which is calculated with tradition KWNN difference:
Reference point estimate variance is bigger, indicates that the reference point signal strength error is larger, then being selected probability accordingly reduces;
Wherein apart from the smallest k reference point, user location is calculated using following formula for selection:
Wherein, For the weight of reference point j, calculate as follows:
dt,j pIndicate dt,jP power, p is adjustable parameter, pass through adjust p obtain best estimate;
The variance of location estimation are as follows:
Family position and location estimation variance are obtained, position fixing process is completed.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910272922.0A CN110007269A (en) | 2019-04-04 | 2019-04-04 | A kind of two stages wireless signal fingerprint positioning method based on Gaussian process |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910272922.0A CN110007269A (en) | 2019-04-04 | 2019-04-04 | A kind of two stages wireless signal fingerprint positioning method based on Gaussian process |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110007269A true CN110007269A (en) | 2019-07-12 |
Family
ID=67170078
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910272922.0A Pending CN110007269A (en) | 2019-04-04 | 2019-04-04 | A kind of two stages wireless signal fingerprint positioning method based on Gaussian process |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110007269A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110557829A (en) * | 2019-09-17 | 2019-12-10 | 北京东方国信科技股份有限公司 | Positioning method and positioning device for fusing fingerprint database |
CN111212474A (en) * | 2020-01-09 | 2020-05-29 | 安徽理工大学 | Visible light indoor positioning method for regenerated fingerprint |
CN112015743A (en) * | 2020-05-28 | 2020-12-01 | 广州杰赛科技股份有限公司 | Indoor positioning system fingerprint database construction method and device |
CN117872269A (en) * | 2024-03-13 | 2024-04-12 | 电子科技大学 | High-precision positioning method for self-adaptive data processing |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105636201A (en) * | 2016-03-14 | 2016-06-01 | 中国人民解放军国防科学技术大学 | Indoor positioning method based on sparse signal fingerprint database |
-
2019
- 2019-04-04 CN CN201910272922.0A patent/CN110007269A/en active Pending
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105636201A (en) * | 2016-03-14 | 2016-06-01 | 中国人民解放军国防科学技术大学 | Indoor positioning method based on sparse signal fingerprint database |
Non-Patent Citations (1)
Title |
---|
王松 等: "《物联网工程综合实验实训教程》", 东南大学出版社 * |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110557829A (en) * | 2019-09-17 | 2019-12-10 | 北京东方国信科技股份有限公司 | Positioning method and positioning device for fusing fingerprint database |
CN111212474A (en) * | 2020-01-09 | 2020-05-29 | 安徽理工大学 | Visible light indoor positioning method for regenerated fingerprint |
CN112015743A (en) * | 2020-05-28 | 2020-12-01 | 广州杰赛科技股份有限公司 | Indoor positioning system fingerprint database construction method and device |
CN117872269A (en) * | 2024-03-13 | 2024-04-12 | 电子科技大学 | High-precision positioning method for self-adaptive data processing |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110007269A (en) | A kind of two stages wireless signal fingerprint positioning method based on Gaussian process | |
Sadowski et al. | Memoryless techniques and wireless technologies for indoor localization with the internet of things | |
CN103402258B (en) | Wi-Fi (Wireless Fidelity)-based indoor positioning system and method | |
CN105704652B (en) | Fingerprint base acquisition and optimization method in a kind of positioning of WLAN/ bluetooth | |
CN105717483B (en) | A kind of location determining method and device based on multi-source positioning method | |
He et al. | INTRI: Contour-based trilateration for indoor fingerprint-based localization | |
TW201329486A (en) | Positioning method | |
CN111722180B (en) | Kalman filtering-based indoor pedestrian positioning method, device and system | |
CN109444814A (en) | A kind of indoor orientation method based on bluetooth and RFID fusion positioning | |
Pasku et al. | Analysis of nonideal effects and performance in magnetic positioning systems | |
CN108307301A (en) | Indoor orientation method based on RSSI rangings and track similitude | |
CN105792349B (en) | signal intensity distribution establishing method and wireless positioning system | |
CN109379711B (en) | positioning method | |
CN105531599A (en) | Method and apparatus for time of flight fingerprint and geo-location | |
KR101749098B1 (en) | System for assuming position of base station and method for assuming position of base station thereof | |
Gu et al. | Trajectory estimation and crowdsourced radio map establishment from foot-mounted imus, wi-fi fingerprints, and gps positions | |
Astafiev et al. | Development of indoor positioning algorithm based on Bluetooth Low Energy beacons for building RTLS-systems | |
CN103888979B (en) | Indoor positioning method based on wireless local area network | |
Li et al. | Location estimation in large indoor multi-floor buildings using hybrid networks | |
CN105044659B (en) | Indoor positioning device and method based on ambient light spectrum fingerprint | |
Pu et al. | Fingerprint-based localization performance analysis: From the perspectives of signal measurement and positioning algorithm | |
Ndzukula et al. | A bluetooth low energy based system for personnel tracking | |
Dong et al. | Neural network based radio fingerprint similarity measure | |
Pendao et al. | FastGraph enhanced: High accuracy automatic indoor navigation and mapping | |
Zhuang et al. | Autonomous WLAN heading and position for smartphones |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190712 |