CN103796305B - Indoor positioning method based on Wi-Fi position fingerprint - Google Patents

Abstract

The invention discloses an indoor positioning method based on Wi-Fi position fingerprints. The indoor positioning method includes the step of setting up a position fingerprint library, a selection method of Wi-Fi accessing points used for positioning calculation, and an indoor positioning algorithm based on the fuzzy logic. When the position fingerprint library is constructed, strength data of received signals from all the Wi-Fi accessing points are needed to be detected and collected in the indoor environment for positioning and the strength of the received signals and position coordinates are correlated so as to finish setting up the position fingerprint library. When the Wi-Fi accessing points used for positioning calculation are selected, the distinction degrees of all the Wi-Fi accessing points and the strength of the received signals of the Wi-Fi accessing points from the positioning stage are comprehensively considered, and part of the Wi-Fi accessing points are selected from all the Wi-Fi accessing points and used for carrying positioning calculation. In the time of positioning calculation, firstly, the fuzzy logic is used for calculating the level of similarity among the position fingerprints. The European distance among the position fingerprints is comprehensively considered on the basis of the level of similarity to finish position estimation.

Description

A kind of indoor orientation method based on Wi-Fi location fingerprints

Technical field

The present invention relates to a kind of indoor orientation method, and in particular to a kind of indoor positioning side based on Wi-Fi location fingerprints Method.

Background technology

As national economy is developed rapidly with scientific and technical, people are in daily life to the demand based on location-based service It is continuously increased, and how to accurately determine customer location is then to realize the basic and key based on location-based service.At present, Ren Mensheng Living and work most of the time all concentrates on the indoor environments such as building, market, dining room, and high-precision indoor positioning technologies can Operating efficiency and quality of life are effectively improved, the demand of indoor positioning service is growing.In the last few years, with science and technology Continue to develop, WLAN is popularized all the more, increasing indoor environment has arranged WAP, wireless network Overlay area constantly expand, particularly Wi-Fi network, these access points are providing the basis of abundance for indoor positioning technologies While facility, the cost required for realizing positioning is also reduced.In numerous indoor positioning technologies, the room based on location fingerprint Interior location technology can obtain ideal positioning precision on the premise of low cost.Therefore, research is referred to based on Wi-Fi positions The indoor positioning technologies of line are imperative.

Knowable to the general principle of indoor positioning technologies of the analysis based on location fingerprint, the interior based on Wi-Fi location fingerprints Location technology is divided into sample phase and positioning stage.Sample phase is first in the setting that the indoor environment for needing to be positioned is artificial A series of reference point known to positional informations, then places wireless signal receiving device at each reference point, detects and gathers Received signal strength data from each AP；Finally the received signal strength data for collecting is processed, and will be every The information such as the received signal strength value after position coordinates, the treatment of individual reference point are stored in location fingerprint storehouse.Positioning stage exists first The received signal strength information from each AP is detected and gathered at tested point, and tested point and position are compared followed by location algorithm The proximity of each reference point in fingerprint base is put, so as to realize treating the estimation of point position coordinate.

Traditional indoor positioning technologies based on Wi-Fi location fingerprints, only deposit when sample phase builds location fingerprint storehouse The average value of received signal strength data is stored up, all AP that can be detected in positioning stage use are used for location Calculation or use Relatively simple standard carries out AP selections, the Euclidean distance between only considering location fingerprint in location Calculation it is as a reference point with The criterion of proximity between tested point.This will cause certain influence to positioning precision.

The content of the invention

In view of the drawbacks described above of prior art, the technical problems to be solved by the invention are to provide a kind of based on Wi-Fi The indoor orientation method of location fingerprint, the Euclidean distance between in location Calculation not only considering location fingerprint, it is also contemplated that position Degree of correlation between fingerprint, improves the precision of positioning precision；Reception signal is not only stored when location fingerprint storehouse is built strong The average value of degrees of data, also stores the standard deviation of received signal strength data, further increases the precision of positioning precision.

The present invention provides a kind of indoor orientation method based on Wi-Fi location fingerprints, it is characterised in that method includes following Step：

(1) indoor Wi-Fi location fingerprints storehouse is built；

(2) received signal strength data from all Wi-Fi access points is gathered at tested point；

(3) the Wi-Fi access points for location Calculation are selected；

(4) Euclidean distance and phase between the Wi-Fi location fingerprints and the Wi-Fi location fingerprints of reference point of tested point are calculated Like degree, the position coordinates of tested point is estimated.

Euclidean distance between not only considering location fingerprint in location Calculation, it is also contemplated that the related journey between location fingerprint Degree, improves the precision of positioning precision.

Further, step (1) builds indoor Wi-Fi location fingerprints storehouse, comprises the following steps：

(11) multiple reference points are set indoors；

(12) at each reference point, the received signal strength data from all Wi-Fi access points is detected；

(13) relevant information of the received signal strength data of all Wi-Fi access points of each reference point, constitutes reference The location fingerprint of point；

(14) by the position coordinates of the location fingerprint associated reference point of each reference point, recorded Wi-Fi location fingerprints storehouse.

Further, step (11) sets multiple reference points indoors, comprises the following steps：

(111) according to indoor area and positioning precision, interior is divided into multiple size identical grids；

(112) each net center of a lattice is as a reference point.

Further, the related of the received signal strength data of all Wi-Fi access points of reference point is believed in step (13) Breath, is included at reference point the average value of the received signal strength data from all Wi-Fi access points, standard deviation and all The title of Wi-Fi access points.

The average value of received signal strength data is not only stored when location fingerprint storehouse is built, received signal strength is also stored The standard deviation of data, further increases the precision of positioning precision.

Further, when selecting the Wi-Fi access points for location Calculation in step (3), according to the area of Wi-Fi access points The received signal strength value size evaluation standard W from different Wi-Fi access points gathered at indexing and tested point, selection L larger Wi-Fi access point of standard on data W values is used for location Calculation.

When position coordinates to tested point is estimated, selected part Wi-Fi access points are used for location Calculation, can carry Computational complexity is effectively reduced while high position precision.

Further, evaluation standard W is comprised the following steps：

(31) discrimination of indoor each Wi-Fi access point is calculated, the computational methods of discrimination are：

Wherein, D (AP_j) j-th discrimination of Wi-Fi access points is represented, m represents the number of reference point,Table Show from j-th average value of the received signal strength data of Wi-Fi access points at i-th reference point,Represent From j-th standard deviation of the received signal strength data of Wi-Fi access points at i-th reference point；

(32) discrimination of indoor all Wi-Fi access points is designated as D (AP), is then had：

D (AP)=[D (AP₁),D(AP₂),…,D(AP_n)]

Wherein, n represents the Wi-Fi number of access points of indoor arrangement；

(33) the received signal strength value size from different Wi-Fi access points gathered at tested point is then considered. Positioning stage, R is designated as by the received signal strength from each Wi-Fi access points gathered at tested point, then R is represented by：

R=[RSS₁,RSS₂,…,RSS_j,…,RSS_n]

Wherein, RSS_jRepresent tested point at gather from j-th received signal strength of Wi-Fi access points；

(34) calculation for selecting the standard on data W for location Calculation Wi-Fi access points is：

W=[w₁,w₂,…,w_n]

Wherein, norRSS_jRepresent RSS_jResult after normalization, normalized computational methods are：

Wherein, RSS_maxRepresent the maximum in R, RSS_minRepresent the minimum value in R.

After being calculated the standard on data W of Wi-Fi access points, the element in W is arranged according to descending, before taking afterwards The corresponding Wi-Fi access points of l element are used for follow-up location Calculation.

Further, the localization method based on Wi-Fi location fingerprints as claimed in claim 1, it is characterised in that step (4) similarity degree between the Wi-Fi location fingerprints of tested point described in and the Wi-Fi location fingerprints of the reference point uses One fuzzy logic is calculated.

Further, step (4) is including step：

(41) by between the Wi-Fi location fingerprints of the Wi-Fi location fingerprints of the tested point and all reference points Euclidean distance EdRSS is normalized calculating；

(42) by between the Wi-Fi location fingerprints of the Wi-Fi location fingerprints of the tested point and all reference points Similarity degree SimRSS is normalized calculating；

(43) using normalized result as the second fuzzy logic input, calculate the tested point and the reference point it Between close on degree.

Further, first fuzzy logic or second fuzzy logic membership function selection triangle or It is trapezoidal.

Further, the method for the position coordinates of estimation tested point is in step (4)：Choose and close on degree most with tested point K reference point high is used for the estimation of tested point position coordinates, and computational methods are：

Wherein,It is the estimate of tested point position coordinates, the output of the second fuzzy logic is designated as O=[o₁,o₂,… o_m], then o_iRepresent and close on degree between tested point and i-th reference point, (x_i,y_i) represent the position coordinates for corresponding to reference point.

Compared with prior art, the indoor orientation method based on Wi-Fi location fingerprints that the present invention is provided has and following has Beneficial effect：

(1) not only consider location fingerprint in location Calculation between Euclidean distance, it is also contemplated that the phase between location fingerprint Pass degree, improves the precision of positioning precision；

(2) average value of received signal strength data is not only stored when location fingerprint storehouse is built, also storage receives signal The standard deviation of intensity data, further increases the precision of positioning precision；

(3) when the position coordinates to tested point is estimated, selected part Wi-Fi access points are used for location Calculation, can be with Computational complexity is effectively reduced while positioning precision is improved.

The technique effect of design of the invention, concrete structure and generation is described further below with reference to accompanying drawing, with It is fully understood from the purpose of the present invention, feature and effect.

Brief description of the drawings

Fig. 1 is the structural representation of the indoor orientation method based on Wi-Fi location fingerprints；

Fig. 2 is the schematic diagram that location fingerprint storehouse builds；

Fig. 3 is schematic diagram of the selection for the Wi-Fi access points of location Calculation；

Fig. 4 is the schematic diagram that location Calculation is carried out using fuzzy logic.

Specific embodiment

As shown in figure 1, the indoor orientation method based on Wi-Fi location fingerprints is including location fingerprint storehouse structure, for positioning The Wi-Fi of calculating accesses three parts of point selection and location Calculation.

Fig. 2 is the schematic diagram that location fingerprint storehouse builds.

When building location fingerprint storehouse, the requirement of the size and positioning precision of indoor environment is considered first, by interior Environment is divided into the essentially identical grid of many sizes, and each net center of a lattice is as a reference point.Then in each reference point Received signal strength data of place's collection from all Wi-Fi access points, and location fingerprint is stored.If using F_iRepresent the Location fingerprint at i reference point, location fingerprint can be expressed as follows：

In formula,J-th name information of Wi-Fi access points of the reference point is represented, Wi-Fi access points can be stored Name or MAC Address.Represent the reference point from j-th received signal strength data of Wi-Fi access points Average value,Represent the reference point from j-th standard deviation of the received signal strength data of Wi-Fi access points.n The number of the Wi-Fi access points that expression can be detected at the reference point, in actual applications, different reference points are detected The number of Wi-Fi access points may be different.

The relevant information storage that location fingerprint is included is corresponded after finishing, it is necessary to be set up with the positional information of reference point Relation.If using L_iRepresent location fingerprint F_iThe positional information of correspondence reference point, positional information can be expressed as follows：

L_i=(x_i,y_i) (2)

In formula, x_iAnd y_iFor representing the position coordinates of reference point.The indoor positioning technologies of this paper primary studies two dimension, if Carry out the research of three-dimensional indoor positioning technologies, the positional information L of reference point_iIn should include three-dimensional coordinate information.Assuming that needing The number that reference point is disposed in the indoor environment for being positioned is m, and positional information is designated as L=(L₁,L₂,…,L_m), m position refers to Line is designated as F=(F₁,F₂,…,F_m).After location fingerprint and positional information are established into one-to-one relation, position is just completed The structure of fingerprint base.

Fig. 3 is schematic diagram of the selection for the Wi-Fi access points of location Calculation.

When selecting the Wi-Fi access points for location Calculation, the received signal strength number for being obtained according to sample phase first According to the discrimination of each Wi-Fi access point in indoor environment is judged, the computational methods of discrimination are as follows：

In formula, D (AP_j) j-th discrimination of Wi-Fi access points is represented, m represents the number of reference point.Can by formula (1) Know,The reference point is represented from j-th average value of the received signal strength data of Wi-Fi access points, Represent the reference point from j-th standard deviation of the received signal strength data of Wi-Fi access points.

If the quantity that Wi-Fi access points are arranged in indoor environment is n, by the area of all Wi-Fi access points in indoor environment Indexing is designated as D (AP), then have：

D (AP)=[D (AP₁),D(AP₂),…,D(AP_n)] (4)

In formula, n represents the Wi-Fi number of access points arranged in indoor environment, RSS_jRepresent that what is gathered at tested point comes from J-th received signal strength of Wi-Fi access points.

In positioning stage, the received signal strength from each Wi-Fi access points gathered at tested point is designated as R, then R can It is expressed as follows：

R=[RSS₁,RSS₂,…,RSS_n] (5)

The standard on data that Wi-Fi accesses point selection is designated as W, the calculation of W is as follows：

W=[w₁,w₂,…,w_n]

In formula, norRSS_jRepresent RSS_jResult after normalization, normalized computational methods are as follows：

In formula, RSS_maxRepresent the maximum in R, RSS_minRepresent the minimum value in R.

After being calculated the standard on data W of Wi-Fi access points, the element in W is arranged according to descending, before taking afterwards The corresponding Wi-Fi access points of l element are used for follow-up location Calculation.

Generally choosing 4 Wi-Fi access points can obtain preferable positioning precision for location Calculation.

Fig. 4 is the schematic diagram that location Calculation is carried out using fuzzy logic.

When carrying out location Calculation, only chosen from the Wi- for location Calculation in the location fingerprint of tested point and reference point The relevant information of the received signal strength data of Fi access points.

First, the Euclidean distance between tested point location fingerprint and reference point locations fingerprint is calculated, computing formula is as follows：

Then, the location fingerprint and the location fingerprint of all reference points to tested point are compared, and comparative approach is as follows：

In formula, l represents the quantity of the Wi-Fi access points for location Calculation, and m represents location fingerprint in location fingerprint storehouse Quantity namely sets the quantity of reference point, diff (R, F)_ijThat is the value stored in the i-th row jth row in diff (R, F) is to treat The absolute value of measuring point and i-th reference point difference of received signal strength on j-th Wi-Fi access point.

Afterwards, the similarity degree between tested point location fingerprint and reference point locations fingerprint is calculated using fuzzy logic.Mould The input variable number of fuzzy logic be l, i.e., for location Calculation Wi-Fi access points quantity, jth, (1≤j≤l) individual input Variable is diff (R, F)_ij, it is output as the similarity degree sim of tested point and i-th reference point_i,sim_i∈[0,1].Due to each For being had differences between the Wi-Fi access points of location Calculation, the position of arrangement is also different, and this is allowed for from each Wi-Fi The excursion of the received signal strength of access point has differences, it is difficult to determine diff (R, F)_ijSpan.Therefore need By diff (R, F)_ijIt is normalized according to formula (7) after calculating and is further used as the input of fuzzy logic.Degree of membership in fuzzy logic Function can select the membership function such as conventional triangle, trapezoidal.The general principle that fuzzy rule should be followed is：diff(R,F)_ij It is smaller, illustrate tested point and reference point on corresponding Wi-Fi access points closer to, tested point and the close Wi-Fi of reference point Number of access point is more, then illustrate that similarity degree between the two is higher.

Euclidean distance between tested point location fingerprint and all reference point locations fingerprints is designated as EdRSS, then EdRSS can It is expressed as：

EdRSS=[dist (R, F₁),dist(R,F₂),…,dist(R,F_m)] (10)

By the similarity degree between tested point location fingerprint and all reference point locations fingerprints, SimRSS is designated as, then SimRSS is represented by：

SimRSS=[sim₁,sim₂,…,sim_m] (11)

Afterwards, EdRSS and SimRSS are normalized calculating according to formula (7) respectively, the result for obtaining is designated as respectively NorEdRSS and norSimRSS.The input that norEdRSS and norSimRSS is used as second layer fuzzy logic is carried out into positioning meter Calculate.The membership function of fuzzy logic can select the membership function such as conventional triangle, trapezoidal.What fuzzy rule should be followed General principle is：Euclidean distance between location fingerprint is smaller, and similarity degree is bigger, then facing between tested point and corresponding reference point Short range degree is higher.

The output of second layer fuzzy logic is designated as O=[o₁,o₂,…o_m], then o_iRepresent tested point and i-th reference point Between close on degree, o_iBetween bigger explanation tested point and corresponding reference point to close on degree higher.Selection is faced with tested point K reference point of short range degree highest is used for the estimation of tested point position coordinates.Method of estimation is as follows：

In formula,Represent the estimate of tested point position coordinates, (x_i,y_i) represent the position coordinates for corresponding to reference point, K Represent the reference point quantity chosen for location estimation.

The indoor orientation method based on Wi-Fi location fingerprints that the present invention is provided, position is not only considered in location Calculation Euclidean distance between fingerprint, it is also contemplated that the degree of correlation between location fingerprint, improves the precision of positioning precision；Building position The average value of received signal strength data is not only stored when putting fingerprint base, the standard deviation of received signal strength data is also stored, entered One step improves the precision of positioning precision；When position coordinates to tested point is estimated, selected part Wi-Fi access points are used for Location Calculation, can be effectively reduced computational complexity while positioning precision is improved.

Preferred embodiment of the invention described in detail above.It should be appreciated that one of ordinary skill in the art without Need creative work just can make many modifications and variations with design of the invention.Therefore, the technology of all the art Personnel are available by logical analysis, reasoning, or a limited experiment on the basis of existing technology under this invention's idea Technical scheme, all should be in the protection domain being defined in the patent claims.

Claims (8)

1. a kind of indoor orientation method based on Wi-Fi location fingerprints, it is characterised in that the described method comprises the following steps：

(1) indoor Wi-Fi location fingerprints storehouse is built；

(2) received signal strength data from all Wi-Fi access points is gathered at tested point；

(3) the Wi-Fi access points for location Calculation are selected；

(4) Euclidean distance and phase between the Wi-Fi location fingerprints and the Wi-Fi location fingerprints of reference point of the tested point are calculated Like degree, the position coordinates of the tested point is estimated；

Wherein, when selecting the Wi-Fi access points for location Calculation in step (3), discrimination according to Wi-Fi access points and treat The received signal strength value size evaluation standard W from different Wi-Fi access points gathered at measuring point, selects the numerical value L larger Wi-Fi access point of standard W values is used for location Calculation；

Wherein, the standard on data W is calculated to comprise the following steps：

(31) discrimination of indoor each Wi-Fi access point is calculated, the computational methods of the discrimination are：

$D\left({\mathrm{AP}}_j\right)=\frac{\sqrt{\underset{i=1}{\overset{m}{\Σ}}{({\mathrm{RSSmean}}_i^j-\frac{1}{m}\underset{i=1}{\overset{m}{\Σ}}{\mathrm{RSSmean}}_i^j)}^2}}{\underset{i=1}{\overset{m}{\Σ}}{\mathrm{RSSstd}}_i^j}$

Wherein, D (AP_j) j-th discrimination of Wi-Fi access points is represented, m represents the number of reference point,Represent From j-th average value of the received signal strength data of Wi-Fi access points at i-th reference point,Represent i-th From j-th standard deviation of the received signal strength data of Wi-Fi access points at individual reference point；

(32) discrimination of indoor all Wi-Fi access points is designated as D (AP), is then had：

D (AP)=[D (AP₁),D(AP₂),…,D(AP_n)]

Wherein, n represents the Wi-Fi number of access points of the indoor arrangement；

(33) the received signal strength value size from different Wi-Fi access points gathered at tested point is then considered；In positioning In the stage, the received signal strength from each Wi-Fi access points gathered at tested point is designated as R, then R is represented by：

R=[RSS₁,RSS₂,…,RSS_j,…,RSS_n]

Wherein, RSS_jRepresent tested point at gather from j-th received signal strength of Wi-Fi access points；

(34) calculation for selecting the standard on data W for location Calculation Wi-Fi access points is：

W=[w₁,w₂,…,w_n]

$w_j=\frac{D\left({\mathrm{AP}}_j\right)}{\underset{j=1}{\overset{m}{\Σ}}D\left({\mathrm{AP}}_j\right)}\·{\mathrm{norRSS}}_j$

Wherein, norRSS_jRepresent RSS_jResult after normalization, normalized computational methods are：

${\mathrm{norRSS}}_j=\frac{{\mathrm{RSS}}_{max}-{\mathrm{RSS}}_j}{{\mathrm{RSS}}_{max}-{\mathrm{RSS}}_{min}}$

Wherein, RSS_maxRepresent the maximum in R, RSS_minRepresent the minimum value in R.

2. the indoor orientation method of Wi-Fi location fingerprints is based on as claimed in claim 1, it is characterised in that step (1) builds The indoor Wi-Fi location fingerprints storehouse that needs are positioned, comprises the following steps：

(11) in the multiple reference points of the indoor setting；

(12) at each described reference point, the received signal strength data from all Wi-Fi access points is detected；

(13) relevant information of the received signal strength data of all described Wi-Fi access points of each reference point, is constituted The location fingerprint of the reference point；

(14) location fingerprint of each reference point is associated the position coordinates of the reference point, is recorded described Wi-Fi Put fingerprint base.

3. the indoor orientation method of Wi-Fi location fingerprints is based on as claimed in claim 2, it is characterised in that step (11) exists The multiple reference points of indoor setting, comprise the following steps：

(111) according to the indoor area and positioning precision, the interior is divided into multiple size identical grids；

(112) each described net center of a lattice is used as the reference point.

4. the indoor orientation method of Wi-Fi location fingerprints is based on as claimed in claim 2, it is characterised in that in step (13) The relevant information of the received signal strength data of all described Wi-Fi access points of the reference point, is included in the reference point Place, the average value of the received signal strength data from all Wi-Fi access points, standard deviation and all Wi-Fi The title of access point.

5. the indoor orientation method of Wi-Fi location fingerprints is based on as claimed in claim 1, it is characterised in that institute in step (4) The similarity degree stated between the Wi-Fi location fingerprints of tested point and the Wi-Fi location fingerprints of the reference point is obscured using first Logical calculated.

6. the indoor orientation method of Wi-Fi location fingerprints is based on as claimed in claim 5, it is characterised in that step (4) includes Step：

(41) by the Euclidean between the Wi-Fi location fingerprints of the Wi-Fi location fingerprints of the tested point and all reference points Calculating is normalized apart from EdRSS；

(42) will be similar between the Wi-Fi location fingerprints of the tested point and the Wi-Fi location fingerprints of all reference points Degree SimRSS is normalized calculating；

(43) using normalized result as the input of the second fuzzy logic, calculate between the tested point and the reference point Close on degree.

7. the indoor orientation method based on Wi-Fi location fingerprints as described in claim 5 or 6, it is characterised in that described The membership function selection triangle or trapezoidal of one fuzzy logic or second fuzzy logic.

8. the indoor orientation method of Wi-Fi location fingerprints is based on as claimed in claim 6, it is characterised in that step is estimated in (4) The method for counting the position coordinates of the tested point is：Choose and close on K reference point of degree highest for institute with the tested point The estimation of tested point position coordinates is stated, computational methods are：

$(\hat{x},\hat{y})=\underset{i=1}{\overset{K}{\Σ}}\frac{o_i}{\underset{i=1}{\overset{K}{\Σ}}{o}_i}\×(x_i,y_i)$

Wherein,It is the estimate of the tested point position coordinates, the output of the second fuzzy logic is designated as O=[o₁,o₂,… o_m], then o_iRepresent and close on degree between the tested point and i-th reference point, (x_i,y_i) represent the position for corresponding to reference point Coordinate.