CN106793078A - Bluetooth indoor orientation method based on RSSI correction value Dual positionings - Google Patents

Abstract

The invention discloses a kind of bluetooth indoor orientation method based on RSSI correction value Dual positionings, its scheme is：1. gather each iBeacon node to receive the received signal strength value of other iBeacon nodes and correct it, obtain iBeacon node received signal strength value correction matrixs；2. each node to the distance matrix of other nodes is calculated according to the correction matrix；3. the coordinate for estimating each node by distance matrix is worth to error of coordinate；4. gather node to be positioned to receive iBeacon node received signal strength values and correct it, obtain node signal strength correction matrix to be positioned；5. node to be positioned to the distance matrix of other iBeacon nodes is obtained according to the correction matrix, and then obtain the coordinate estimate of node to be positioned；6. the final coordinate of node to be positioned is obtained by the result of step 3 and step 5.Positioning precision of the present invention is high, can be used for home for destitute and intelligence community.

Description

Bluetooth indoor orientation method based on RSSI correction value Dual positionings

Technical field

The invention belongs to communication technical field, more particularly to a kind of bluetooth indoor orientation method can be used for market, endowment Institute, intelligence community and fire-fighting place.

Background technology

With Internet of Things society develop rapidly, sensing intelligent equipment with its low-power consumption, self-organizing, layout facilitate the advantages of Increasingly it is widely used in wisdom society, and indoor positioning has turned into the social important support field of Internet of Things, and base One of focus of indoor positioning will be turned into its unique superiority in the indoor positioning technologies of bluetooth iBeacon.

Being currently based on the indoor positioning algorithms of bluetooth is mainly had the algorithm TOA based on signal transmission time, is passed based on signal The algorithm TDOA of defeated time difference, the algorithm AOA based on direction of arrival degree, algorithm and list based on received signal strength value RSSI Centroid algorithm, wherein：

TOA algorithms, because Indoor is relatively narrow and small, the propagation of signal can be brought propagation by the interference of barrier The time delay of time, and the superposition of time delay can be produced, while the clock essence between this algorithm requirement positioning node and reference mode Plesiochronous, the requirement to hardware is very high, and the complexity of system and the input of cost are larger, cause practicality relatively low.

TDOA algorithms, are the improvement of TOA algorithms, and this algorithm is not the absolute time that reference mode is directly reached using signal Between, but using signal reach two time differences of reference mode and determine the position of positioning node, therefore without necessarily referring to section Clock between point and positioning node is precisely synchronous, it is only necessary to which the clock between reference mode is synchronous, although the algorithm drops The synchronous requirement of low clock, but precision is relatively low.

AOA algorithms, this algorithm needs to obtain angle information in node mounted antennas matrix, but because major part is saved The antenna of point is all omnidirectional, it is impossible to which signal is distinguished in which direction comes from, and needs spy in bluetooth nodes mounted antennas Different hardware device, such as aerial array, cause bluetooth nodes to exceed common sensing section in power consumption, size and price Point, the characteristic with wireless smart devices low cost and low-power consumption is disagreed, so practicality is poor.

RSSI algorithms, the algorithm is by obtaining the relation between received signal strength value RSSI and distance, obtaining distance Model and RSSI between, so as to carry out indoor positioning.But due to radio signal free damping effect, letter under indoor environment Number sink effect, non-line-of-sight propagation effect, multipath effect, shadow effect cause the RSSI value of Bluetooth signal with the change of distance Change doing acutely variation, cause the error range that traditional location algorithm based on received signal strength value RSSI is present compared with Greatly.But because its requirement to clock treatment is low, low cost, complexity is low, so general indoor positioning algorithms are all based on RSSI launches.

Simple substance center algorithm, is the information that all reference modes in its communication range are received by positioning node, and will The geometry barycenter of reference mode is positioned as the estimated location of oneself, but barycenter number is single, from node to be positioned farther out The disturbance degree of the reference mode at place is smaller, it is impossible to which the contribution degree of the reference mode given prominence to the key points from node to be positioned remotely is big Small, the error brought is larger, and positioning precision is not lifted further.

The content of the invention

It is an object of the invention to be directed to above the deficiencies in the prior art, propose a kind of based on RSSI correction value Dual positionings Bluetooth indoor orientation method, to improve the indoor position accuracy of bluetooth iBeacon.

To achieve the above objectives, technical scheme includes as follows：

(1) received signal strength of other iBeacon nodes that each iBeacon node of collection indoor deployment is received Value, constitutes iBeacon node signal strength matrix R, and it is modified using Rod Dixon detection method and Gaussian filter algorithm, The signal intensity correction value of each iBeacon node is obtained, iBeacon node signal strength correction matrixs R' is constituted：

Wherein：

I, j are iBeacon node serial numbers, and n is iBeacon node numbers, i ∈ [1, n], j ∈ [1, n], n>3；

R_ij=[R_ij1 R_ij2 R_ij3 ... R_ij30] it is 1 × 30 dimension matrix, represent that i-th iBeacon node is received j-th 30 groups of signal values that iBeacon nodes send；

R_ij' represent that i-th iBeacon node receives j-th signal intensity correction value of iBeacon nodes.

(2) according to iBeacon node signal strength correction matrix R', using log-distance path loss model model, obtain IBeacon nodes constitute iBeacon nodal distance matrix Ds to the distance value of other iBeacon nodes：

Wherein：D_ijRepresent i-th iBeacon node to j-th distance value of iBeacon nodes, as i=j, D_ij= 0。

(3) according to iBeacon nodal distance matrix Ds, using multiple centroid algorithm, each iBeacon node coordinates are obtained and is estimated Value Q (x_i,y_i), by Q (x_i,y_i) actual coordinate e (v with each iBeacon nodes_i,z_i) compare, obtain error dot P (α, β)；

(4) gather node to be positioned and receive the received signal strength value that iBeacon nodes send, constitute positioning node signal Intensity matrix r, and it is modified using Rod Dixon detection method and Gaussian filter algorithm, the intensity for obtaining node to be positioned is repaiied On the occasion of composition node signal strength correction matrix r' to be positioned：

R=[r₁ r₂ r₃ ... r_i ... r_n], r'=[r₁' r₂' r₃' ... r_i' ... r_n']

Wherein：

r_iIt is 1 × 30 dimension matrix, represents that node to be positioned receives 30 groups of signal values that i-th iBeacon node sends；

r_i' represent that node to be positioned receives i-th signal intensity correction value of iBeacon nodes.

(5) according to node signal strength correction matrix r' to be positioned, using log-distance path loss model model, obtain undetermined Position node constitutes nodal distance matrix d to be positioned to the distance value of other iBeacon nodes：

D=[d₁ d₂ d₃ ... d_i ... d_n]

Wherein：d_iIt is node to be positioned to i-th distance value of iBeacon nodes.

(6) according to nodal distance matrix d to be positioned, using multiple centroid algorithm, the coordinate estimate W of node to be positioned is obtained (x,y)；

(7) according to coordinate estimate W (x, y) and error dot P (α, β), the final coordinate for obtaining node to be positioned is：W(x+ α,y+β)。

Major advantage is the present invention compared with the prior art：

First, compared with existing early stage is pre-processed, the present invention is due to the larger received signal strength of the fluctuation to early stage Value carries out the treatment of Rod Dixon detection algorithm, rejects the violent received signal strength value of shake, and then the data for retaining are used Gaussian filtering process, remains the less received signal strength value of shake to greatest extent.

Second, compared with existing route loss model, the present invention is due to according to lot of experimental data, providing suitable environment system Number term of reference, not only realizes simply, and the coefficient of selection is more applicable for indoor environment.

3rd, compared with existing location algorithm, the present invention calculates the seat of system respectively due to using Dual positioning algorithm The estimated coordinates of mark error and node to be positioned, therefore can calculate and treat according to error of coordinate and the estimated coordinates of node to be positioned The final coordinate of positioning node.

4th, compared with existing location algorithm, the present invention is due to the iBeacon nodes apart from node to be positioned farther out being divided Group, after obtaining its barycenter, barycenter is applied in the iBeacon node groups nearer apart from node to be positioned, then asks for barycenter, is made Obtain positioning precision further to improve, while fitness can be avoided to change with the state change of iBeacon nodes in environment.

Brief description of the drawings

Fig. 1 is of the invention to realize flow chart；

Fig. 2 is the locating effect comparison diagram with existing simple substance center algorithm of the invention.

Specific embodiment

Reference picture 1, the bluetooth indoor orientation method based on RSSI correction value Dual positionings of the invention, it implements step It is rapid as follows：

Step 1, builds iBeacon node signal strength matrixes R.

Indoor environment is divided into some regions with regular hexagon, each apex in region places an iBeacon Node, places n iBeacon node altogether, and each iBeacon node periodically broadcasts itself numbering, own coordinate and reception Signal strength values, gather the received signal strength value of other iBeacon nodes that each iBeacon node is received, and constitute IBeacon node signal strength matrixes R：

Wherein：I, j are iBeacon node serial numbers, and n is iBeacon node numbers, i ∈ [1, n], j ∈ [1, n], n>3；

R_ij=[R_ij1 R_ij2 R_ij3 ... R_ij30] it is 1 × 30 dimension matrix, represent that i-th iBeacon node is received j-th 30 groups of signal values that iBeacon nodes send.

Step 2, is modified to iBeacon node signal strength matrixes R, obtains iBeacon node Rod Dixon matrixes R^*。

2a) sent from other iBeacon nodes using what Rod Dixon detection method rejected that each iBeacon node receives 30 groups of signal values in change violent received signal strength value, i-th iBeacon node is received from j-th 30 groups of received signal strength values of iBeacon nodes are followed successively by R by order arrangement from small to large_ij1,R_ij2,...,R_ij30；

The detection level of rejecting outliers 2b) is determined for a=0.02, determines that Rod Dixon checks critical value M (a, n)；

2c) according to Dixion test statistical formula, exceptional value G and the iBeacon section of iBeacon node most significant ends are calculated Point least significant end exceptional value G'：

2d) by iBeacon nodes most significant end exceptional value G and iBeacon node least significant end exceptional value G' and critical value M (a, N) it is compared：

If G>M (a, n) or G'>M (a, n), then reject the corresponding received signal strength value R of the exceptional value_ijN；

If G≤M (a, n) or G'≤M (a, n), then retain the corresponding received signal strength value R of the exceptional value_ijN, Perform 2e), N is 30 groups of numberings of signal value；

2e) the received signal strength value rearrangement to retaining, repeat step 2a) -2d), until all changes are violent Received signal strength value is removed, and using the final value for retaining as the output of Rod Dixon detection algorithm, constitutes iBeacon nodes Rod Dixon matrix R^*：

Wherein：

R^* _ij=[R^* _ij1 R^* _ij2 R^* _ij3 ... R^* _ijM] it is 1 × M dimension matrixes, M is examined by Rod Dixon in 30 groups of signal values The signal value number retained after method of determining and calculating, R^* _ijRepresent that i-th iBeacon node receives j-th iBeacon node sends 30 Signal value matrix in group signal value by retaining after Rod Dixon detection algorithm；

Step 3, to iBeacon node Rod Dixon matrixes R^*It is modified, obtains iBeacon node signal strength amendment squares Battle array R'.

3a) to iBeacon node Rod Dixon matrixes R^*Gaussian filtering is carried out, iBeacon node Gaussian matrixes R is obtained "；

3b) to iBeacon node Gaussian matrixes R " each element matrix take arithmetic mean of instantaneous value, obtain each iBeacon section The received signal strength correction value of point, constitutes iBeacon node signal strength correction matrixs R'：

Wherein：R_ij' represent that i-th iBeacon node receives j-th signal intensity correction value of iBeacon nodes.

Step 4, builds iBeacon nodal distance matrix Ds.

4a) set up received signal strength correction value to the path loss relational expression of distance between iBeacon nodes：

Wherein：

R_ij' represent that i-th iBeacon node receives j-th signal intensity correction value of iBeacon nodes；

q₀Represent two iBeacon euclidean distance between node pair L₀Reference signal strength correction value at=1 meter；

D_ijRepresent i-th iBeacon node to j-th distance value of iBeacon nodes, as i=j, D_ij=0；

x_εIt is ambient parameter, the influence degree of representation space environment, x with u_εIt is to be received at 1 meter of iBeacon nodes Mean power absolute value, u is path-loss factor, x_εOptimal reference scope is 41-47, and u optimal references scope is 2.15- 4.3；

4b) by step 4a) path loss relational expression be converted into following form：

4c) according to step 4b) draw each iBeacon node to other all iBeacon nodal distances matrix Ds：

Step 5, calculates the coordinate estimate Q (x of each iBeacon nodes_i,y_i)。

5a) by i-th iBeacon node to j-th distance value D of iBeacon nodes_ijArranged by order from small to large Sequence, obtains distance set：D_i1,D_i2,D_i3,...D_ij,...,D_in, wherein D_i1<D_i2<D_i3<...<D_ij<...<D_in；

5b) minimum value D is subtracted with each element in distance set_i1, draw difference set：

0,ΔD_i1,ΔD_i2,ΔD_i3,...,ΔD_ij,...,ΔD_in；

5c) in calculating difference set each element average value

Wherein：ΔD_ij=D_ij-D_i1, represent i-th iBeacon node to the distance between j-th iBeacon node value D_ijWith minimum value D_i1Difference；

5d) iBeacon nodes are respectively classified intoSet A andSet B, be located in set B There is m node, and be divided into one group immediate 3 iBeacon nodes are differed in set B, be divided into m-2 groups, calculate every group Barycenter, its coordinate (x_k,y_k)：

Wherein：K ∈ [1, m-2], (x_k1,y_k1)、(x_k2,y_k2)、(x_k3,y_k3) it is that 3 iBeacon nodes in every group are sat Mark；

5e) by step 5d) barycenter that obtains is considered as the iBeacon nodes of new addition again, then by m-2 center of mass point and collection The n-m iBeacon node closed in A constitutes polygon, and this polygonal barycenter is calculated using centroid algorithm, obtains i-th IBeacon node estimated coordinates Q (x_i,y_i)。

Step 6, calculation error point P (α, β).

According to i-th actual coordinate e (v of iBeacon nodes_i,z_i) and estimated coordinates Q (x_i,y_i), calculation error point P's Abscissa and ordinate：

Obtain error dot

Step 7, builds node signal strength matrix r to be positioned.

Indoor environment is divided into some regions with regular hexagon, each apex in region places an iBeacon Node, places n iBeacon node altogether, and each iBeacon node periodically broadcasts itself numbering, own coordinate and reception Signal strength values, gather node to be positioned and receive the received signal strength value that iBeacon nodes send, and constitute node letter to be positioned Number intensity matrix r：

R=[r₁ r₂ r₃ ... r_i ... r_n]

Wherein：r_iIt is 1 × 30 dimension matrix, represents that node to be positioned receives 30 groups of signals that i-th iBeacon node sends Value；

Step 8, treats positioning node signal intensity matrix r and is modified, and obtains node Rod Dixon matrix r to be positioned^*。

8a) node to be positioned is received sent from other iBeacon nodes 30 are rejected using Rod Dixon detection method Change violent received signal strength value in group signal value, node to be positioned is received from i-th the 30 of iBeacon nodes Group received signal strength value is followed successively by r by order arrangement from small to large_i1,r_i2,...,r_i30；

8b) according to Dixion test statistical formula, node most significant end exceptional value g to be positioned and node to be positioned are calculated most Low side exceptional value g'：

8c) by node most significant end exceptional value g to be positioned and node least significant end exceptional value g' to be positioned and critical value M (a, n) It is compared：

If g>M (a, n) or g'>M (a, n), then reject the corresponding received signal strength value r of the exceptional value_iN；

If g≤M (a, n) or g'≤M (a, n), then retain the corresponding received signal strength value r of the exceptional value_iN, hold Row 8d), N is 30 groups of numberings of signal value；

8d) the received signal strength value rearrangement to retaining, repeat step 8a) -8c), until all changes are violent Received signal strength value is removed, and using the final value for retaining as the output of Rod Dixon detection algorithm, constitutes node to be positioned Rod Dixon matrix r^*：

r^*=[r₁ ^* r₂ ^* r₃ ^* ... r_i ^* ... r_n ^*]

Wherein：r_i ^*=[r^* _i1 r^* _i2 r^* _i3 ... r^* _iM] it is 1 × M dimension matrixes, M is by Rod Dixon in 30 groups of signal values The signal value number retained after detection algorithm, r_i ^*Represent that node to be positioned receives 30 groups of signals that i-th iBeacon node sends Signal value matrix in value by retaining after Rod Dixon detection algorithm.

Step 9, treats positioning node Rod Dixon matrix r^*It is modified, obtains node signal strength correction matrix to be positioned r′。

Positioning node Rod Dixon matrix r 9a) is treated using gaussian filtering^*Processed, obtained node Gaussian Moment to be positioned Battle array r "；

9b) treat positioning node Gaussian matrix r " each element matrix take arithmetic mean of instantaneous value, obtain connecing for node to be positioned Signal intensity correction value is received, node signal strength correction matrix r' to be positioned is constituted：

R'=[r₁' r₂' r₃' ... r_i' ... r_n']

Wherein：r_i' represent that node to be positioned receives i-th signal intensity correction value of iBeacon nodes.

Step 10, builds nodal distance matrix d to be positioned.

Received signal strength correction value 10a) is set up to be damaged to node to be positioned and i-th path of iBeacon nodal distances Consumption relational expression：

Wherein：

r_i' represent that node to be positioned receives i-th signal intensity correction value of iBeacon nodes；

q₁Represent node to be positioned and iBeacon euclidean distance between node pair L₀Reference signal strength correction value at=1 meter；

d_iIt is node to be positioned to i-th distance value of iBeacon nodes.

10b) by step 10a) path loss relational expression be converted into following form：

10c) according to step 10b) draw node to be positioned to other all iBeacon nodal distances matrix d：

D=[d₁ d₂ d₃ ... d_i ... d_n]

Step 11, calculates coordinate estimate W (x, y) of node to be positioned.

11a) by node to be positioned to i-th distance value d of iBeacon nodes_iSorted by order from small to large, obtained Distance set to be positioned：d_ε={ d₁,d₂,d₃,...,d_i,...,d_n, wherein d₁<d₂<d₃<...<d_i<...<d_n；

11b) with distance set d to be positioned_εEach interior element subtracts minimum value d₁, draw difference set d to be positioned_ε'：

d_ε'={ 0, Δ d₁,Δd₂,Δd₃,...,Δd_i,...,Δd_n,

Wherein：Δd_i=d_i-d₁, it represents node to be positioned to i-th distance value d of iBeacon nodes_iWith minimum value d₁ Difference；

11c) calculate difference set d to be positioned_ε' in each element average value

IBeacon nodes 11d) are respectively classified into two set, i.e.,Set C₁WithSet C₂, it is located at set C₂Inside there is z node, set C₂The middle immediate 3 iBeacon nodes of difference are divided into one group, are divided into z- 2 groups, calculate every group of barycenter, its coordinate (x_L,y_L) be：

Wherein：L ∈ [1, z-2], (x_L1,y_L1)、(x_L2,y_L2)、(x_L3,y_L3) it is that 3 iBeacon nodes in every group are sat Mark.

11e) by step 11d) barycenter that obtains is considered as the iBeacon nodes of new addition again, then by z-2 center of mass point with Set C₁N-z interior iBeacon node constitutes polygon, and this polygonal barycenter is calculated using centroid algorithm, is treated Estimated coordinates W (x, y) of positioning node.

Step 12, calculates the final coordinate W (x+ α, y+ β) of node to be positioned.

The horizontal seat of the error dot P that estimated coordinates W (x, y) and step 6 of the node to be positioned obtained according to step 11 are obtained α and ordinate β is marked, the final coordinate for being calculated node to be positioned is：

With reference to emulation experiment, locating effect of the invention is further analyzed.

1. experiment condition

16 iBeacon nodes, 10 nodes to be positioned are arranged in this experiment under 50*50 meters of indoor environment.

2. experiment content

Above-mentioned 10 nodes to be positioned are tested respectively with the inventive method and existing simple substance center algorithm, obtains undetermined The coordinate value of position node, as a result such as Fig. 2.

As it is clear from fig. 2 that the present invention measures the actual coordinate value contrast of the coordinate value and node to be positioned of node to be positioned, its Positioning precision max value of error is 2.413 meters, and minimum value is 0.75 meter, and average value is 1.59 meters.Existing centroid algorithm is measured to be treated The coordinate value of positioning node is contrasted with the actual coordinate value of node to be positioned, and its positioning precision max value of error is 4.089 meters, most Small value is 1.56 meters, and average value is 2.75 meters.

Experiment shows：Positioning precision is improve 25%-35% by the present invention, has larger advantage in positioning indoors.

Claims (7)

1. a kind of bluetooth indoor orientation method based on RSSI correction value Dual positionings, including：

(1) the received signal strength value of other iBeacon nodes that each iBeacon node of collection indoor deployment is received, IBeacon node signal strength matrix R are constituted, and it is modified using Rod Dixon detection method and Gaussian filter algorithm, obtained To the signal intensity correction value of each iBeacon node, iBeacon node signal strength correction matrixs R' is constituted：

$\begin{array}{cc}R=\left[\begin{array}{cccc}0& R_{12}& ...& R_{1n}\\ R_{21}& 0& ...& R_{2n}\\ ...& ...& R_{ij}& ...\\ R_{n1}& R_{n2}& ...& 0\end{array}\right],& R^{\&prime;}=\left[\begin{array}{cccc}0& {R^{\&prime;}}_{12}& ...& {R^{\&prime;}}_{1n}\\ {R^{\&prime;}}_{21}& 0& ...& {R^{\&prime;}}_{2n}\\ ...& ...& {R^{\&prime;}}_{ij}& ...\\ {R^{\&prime;}}_{n1}& {R^{\&prime;}}_{n2}& ...& 0\end{array}\right]\end{array}$

Wherein：

I, j are iBeacon node serial numbers, and n is iBeacon node numbers, i ∈ [1, n], j ∈ [1, n], n>3；

R_ij=[R_ij1 R_ij2 R_ij3 ... R_ij30] it is 1 × 30 dimension matrix, represent that i-th iBeacon node is received j-th 30 groups of signal values that iBeacon nodes send；

R_ij' represent that i-th iBeacon node receives j-th signal intensity correction value of iBeacon nodes.

(2) according to iBeacon node signal strength correction matrix R', using log-distance path loss model model, iBeacon is obtained Node constitutes iBeacon nodal distance matrix Ds to the distance value of other iBeacon nodes：

$D=\left[\begin{array}{cccc}0& D_{12}& ...& D_{1n}\\ D_{21}& 0& ...& D_{2n}\\ ...& ...& D_{ij}& ...\\ D_{n1}& D_{n2}& ...& 0\end{array}\right]$

Wherein：D_ijRepresent i-th iBeacon node to the distance between j-th iBeacon node value, as i=j, D_ij=0.

(3) according to iBeacon nodal distance matrix Ds, using multiple centroid algorithm, the coordinate estimate Q of each iBeacon nodes is obtained (x_i,y_i), by Q (x_i,y_i) actual coordinate e (v with each iBeacon nodes_i,z_i) compare, obtain error dot P (α, β)；

(4) gather node to be positioned and receive the received signal strength value that iBeacon nodes send, constitute positioning node signal intensity Matrix r, and it is modified using Rod Dixon detection method and Gaussian filter algorithm, obtain the intensity amendment of node to be positioned Value, constitutes node signal strength correction matrix r' to be positioned：

R=[r₁ r₂ r₃ ... r_i ... r_n], r'=[r₁' r₂' r₃' ... r_i' ... r_n']

Wherein：

r_iIt is 1 × 30 dimension matrix, represents that node to be positioned receives 30 groups of signal values that i-th iBeacon node sends；

r_i' represent that node to be positioned receives i-th signal intensity correction value of iBeacon nodes.

(5) according to node signal strength correction matrix r' to be positioned, using log-distance path loss model model, section to be positioned is obtained Point constitutes nodal distance matrix d to be positioned to the distance value of other iBeacon nodes：

D=[d₁ d₂ d₃ ... d_i ... d_n]

Wherein：d_iIt is node to be positioned to i-th distance value of iBeacon nodes.

(6) according to nodal distance matrix d to be positioned, using multiple centroid algorithm, obtain node to be positioned coordinate estimate W (x, y)；

(7) according to coordinate estimate W (x, y) and error dot P (α, β), the final coordinate for obtaining node to be positioned is：W(x+α,y+ β)。

2. method according to claim 1, wherein utilizes Rod Dixon detection method and Gaussian filter algorithm pair in step (1) IBeacon node signal strength matrixes R is modified, and carries out in accordance with the following steps：

1.1) indoor environment is divided into some regions with regular hexagon, each apex in region places an iBeacon Node, places n iBeacon node altogether, and each iBeacon node periodically broadcasts itself numbering, own coordinate and reception Signal strength values；

1.2) using Rod Dixon detection method being sent from other iBeacon nodes of rejecting that each iBeacon node receives Change violent received signal strength value in 30 groups of signal values：

1.2a) by i-th iBeacon node receive from j-th the 30 of iBeacon nodes groups of received signal strength value by from It is small to be arranged to big order, it is followed successively by R_ij1,R_ij2,...,R_ij30；

The detection level of rejecting outliers 1.2b) is determined for a=0.02, determines that Rod Dixon checks critical value M (a, n)；

1.2c) according to Dixion test statistical formula, exceptional value G and the iBeacon node of iBeacon node most significant ends is calculated Least significant end exceptional value G'：

$G=\frac{R_{in}-R_{i(n-1)}}{R_{in}-R_{i1}}$

$G^{\&prime;}=\frac{R_{i2}-R_{i1}}{R_{in}-R_{i1}};$

1.2d) by iBeacon nodes most significant end exceptional value G and iBeacon node least significant end exceptional value G' and critical value M (a, n) It is compared：

If G>M (a, n) or G'>M (a, n), then reject the corresponding received signal strength value R of the exceptional value_ijN；

If G≤M (a, n) or G'≤M (a, n), then retain the corresponding received signal strength value R of the exceptional value_ijN, perform 1.2e), N is 30 groups of numberings of signal value；

1.2e) the received signal strength value rearrangement to retaining, repeat step 1.2a) -1.2d), until all changes are violent The signal strength values that receive be removed, and will the final value for retaining as the output of Rod Dixon detection algorithm, constitute iBeacon sections Point Rod Dixon matrix R^*：

$R^{*}=\left[\begin{array}{cccc}0& {R^{*}}_{12}& ...& {R^{*}}_{1n}\\ {R^{*}}_{21}& 0& ...& {R^{*}}_{2n}\\ ...& ...& {R^{*}}_{ij}& ...\\ {R^{*}}_{n1}& {R^{*}}_{n2}& ...& 0\end{array}\right]$

Wherein：R^* _ij=[R^* _ij1 R^* _ij2 R^* _ij3 ... R^* _ijM] it is 1 × M dimension matrixes, M is examined by Rod Dixon in 30 groups of signal values The signal value number retained after method of determining and calculating, R^* _ijRepresent that i-th iBeacon node receives j-th iBeacon node sends 30 Signal value matrix in group signal value by retaining after Rod Dixon detection algorithm.

1.3) to iBeacon node Rod Dixon matrixes R^*Gaussian filtering process is carried out, iBeacon node Gaussian matrixes R is obtained ", so Afterwards to iBeacon node Gaussian matrixes R " each element take arithmetic mean of instantaneous value, the reception signal for obtaining each iBeacon node is strong Degree correction value, constitutes iBeacon node signal strength correction matrixs：

3. method according to claim 1, wherein calculates i-th iBeacon node to j-th iBeacon in step (2) The distance between node value D_ij, carry out in accordance with the following steps：

2.1) received signal strength value to the path loss relational expression of distance is set up：

${R_{ij}}^{\&prime;}=q_0+10u\lg \left(\frac{D_{ij}}{L_0}\right)+x_{\mathrm{\&epsiv;}}$

Wherein：

R_ij' represent that i-th iBeacon node receives j-th signal intensity correction value of iBeacon nodes；

q₀Represent two iBeacon euclidean distance between node pair L₀Reference signal strength correction value at=1 meter；

x_εIt is ambient parameter, the influence degree of representation space environment, x with u_εIt is received at 1 meter of iBeacon nodes flat The absolute value of equal power, u is path-loss factor, x_εOptimal reference scope is 41-47, and u optimal references scope is 2.15-4.3.

2.2) by step 2.1) path loss relational expression be converted into following form：

$D_{ij}={10}^{\frac{q_0+x_{\mathrm{\&epsiv;}}-{R_{ij}}^{\&prime;}}{10u}};$

2.3) according to step 2.2) draw each iBeacon node to other all iBeacon nodal distances matrix Ds：

$D=\left[\begin{array}{cccc}0& D_{12}& ...& D_{1n}\\ D_{21}& 0& ...& D_{2n}\\ ...& ...& D_{ij}& ...\\ D_{n1}& D_{n2}& ...& 0\end{array}\right].$

4. method according to claim 1, wherein calculates error dot P (α, β) in step (3), carries out in accordance with the following steps：

3.1) the coordinate estimate Q (x of each iBeacon nodes are estimated_i,y_i)：

3.1a) by i-th iBeacon node to j-th distance value D of iBeacon nodes_ijSorted by order from small to large, Obtain distance set：D_i1,D_i2,D_i3,...D_ij,...,D_in, wherein D_i1<D_i2<D_i3<...<D_ij<...<D_in；

3.1b) minimum value D is subtracted with each element in distance set_i1, draw difference set：

0,ΔD_i1,ΔD_i2,ΔD_i3,...,ΔD_ij,...,ΔD_in

3.1c) in calculating difference set each element average value

$\mathrm{\&Delta;}\stackrel{\&OverBar;}{D}=\frac{\underset{j=1}{\overset{n}{\&Sigma;}}{\mathrm{\&Delta;D}}_{ij}}{n-1}$

Wherein：ΔD_ij=D_ij-D_i1, represent i-th iBeacon node to the distance between j-th iBeacon node value D_ijWith Minimum value D_i1Difference；

3.1d) iBeacon nodes are respectively classified intoWithTwo set A and B, being located in set B has M node, one group is divided into immediate 3 iBeacon nodes are differed in set B, is divided into m-2 groups, calculates every group of matter The heart, its coordinate (x_k,y_k)：

$(x_k,y_k)=(\frac{x_{k1}+x_{k2}+x_{k3}}{3},\frac{y_{k1}+y_{k2}+y_{k3}}{3}),$

Wherein：K ∈ [1, m-2], (x_k1,y_k1)、(x_k2,y_k2)、(x_k3,y_k3) it is 3 iBeacon node coordinates in every group；

3.1e) by step 3.1d) barycenter that obtains is considered as the iBeacon nodes of new addition again, then by m-2 center of mass point and collection The n-m iBeacon node closed in A constitutes polygon, and this polygonal barycenter is calculated using centroid algorithm, obtains i-th IBeacon node estimated coordinates Q (x_i,y_i)。

3.2) according to i-th actual coordinate e (v of iBeacon nodes_i,z_i) and estimated coordinates Q (x_i,y_i), calculation error point P's Abscissa and ordinate：

$\mathrm{\&alpha;}=\frac{\underset{i=1}{\overset{n-2}{\&Sigma;}}(x_i-v_i)}{n-2},\mathrm{\&beta;}=\frac{\underset{i=1}{\overset{n-2}{\&Sigma;}}(y_i-z_i)}{n-2}.$

5. method according to claim 1, is wherein treated in step (4) using Rod Dixon detection method and Gaussian filter algorithm Positioning node signal intensity matrix r is modified, and carries out in accordance with the following steps：

4.1) indoor environment is divided into some regions with regular hexagon, each apex in region places an iBeacon Node, places n iBeacon node altogether, and each iBeacon node periodically broadcasts itself numbering, own coordinate and reception Signal strength values；

4.2) the 30 groups of letters sent from other iBeacon nodes that node to be positioned is received are rejected using Rod Dixon detection method Change violent received signal strength value in number value：

4.2a) node to be positioned is received from i-th the 30 of iBeacon nodes groups of received signal strength value by from small to large Order arrangement, be followed successively by r_i1,r_i2,...,r_i30；

4.2b) according to Dixion test statistical formula, calculate node most significant end exceptional value g to be positioned and node to be positioned is minimum Hold exceptional value g'：

$g=\frac{r_{in}-r_{i(n-1)}}{r_{in}-r_{i1}},$

$g^{\&prime;}=\frac{r_{i2}-r_{i1}}{r_{in}-r_{i1}};$

4.2c) node most significant end exceptional value g to be positioned and node least significant end exceptional value g' to be positioned and critical value M (a, n) are entered Row compares：

If g>M (a, n) or g'>M (a, n), then reject the corresponding received signal strength value r of the exceptional value_iN；

If g≤M (a, n) or g'≤M (a, n), then retain the corresponding received signal strength value r of the exceptional value_iN, perform 4.2d), N is 30 groups of numberings of signal value；

4.2d) the received signal strength value rearrangement to retaining, repeat step 4.2a) -4.2c), until all changes are violent Received signal strength value be removed, and will the final value for retaining as the output of Rod Dixon detection algorithm, constitute section to be positioned Point Rod Dixon matrix r^*：

r^*=[r₁ ^* r₂ ^* r₃ ^* ... r_i ^* ... r_n ^*]

Wherein：r_i ^*=[r^* _i1 r^* _i2 r^* _i3 ... r^* _iM] it is 1 × M dimension matrixes, M is detected by Rod Dixon in 30 groups of signal values The signal value number retained after algorithm, r_i ^*In representing that node to be positioned receives 30 groups of signal values that i-th iBeacon node sends By the signal value matrix retained after Rod Dixon detection algorithm.

4.3) positioning node Rod Dixon matrix r is treated_i ^*Gaussian filtering process is carried out, node Gaussian matrix r to be positioned is obtained ", so Treat positioning node Gaussian matrix r afterwards " each element matrix take arithmetic mean of instantaneous value, obtain the received signal strength of node to be positioned Correction value, constitutes node signal strength correction matrix to be positioned：R'=[r₁' r₂' r₃' ... r_i' ... r_n']。

6. calculated in method according to claim 1, wherein step (5) node to be positioned to i-th iBeacon node it Between distance value d_i, carry out in accordance with the following steps：

5.1) received signal strength value to the path loss relational expression of distance is set up：

${r_i}^{\&prime;}=q_1+10u\lg \left(\frac{d_i}{L_0}\right)+x_{\mathrm{\&epsiv;}}$

Wherein：

r_i' represent that node to be positioned receives i-th signal intensity correction value of iBeacon nodes；

q₁Represent node to be positioned and iBeacon euclidean distance between node pair L₀Reference signal strength correction value at=1 meter；

x_εIt is ambient parameter, the influence degree of representation space environment, x with u_εIt is received at 1 meter of iBeacon nodes flat The absolute value of equal power, u is path-loss factor, x_εOptimal reference scope is 41-47, and u optimal references scope is 2.15-4.3.

5.2) by step 5.1) path loss relational expression be converted into following form：

$d_i={10}^{\frac{q_1+x_{\mathrm{\&epsiv;}}-{r_i}^{\&prime;}}{10u}};$

5.3) according to step 5.2) draw node to be positioned to other all iBeacon nodal distances matrix d：

D=[d₁ d₂ d₃ ... d_i ... d_n]。

7. method according to claim 1, wherein calculates coordinate estimate W (x, y) of node to be positioned in step (6), Carry out in accordance with the following steps：

6.1) by node to be positioned to i-th distance value d of iBeacon nodes_iSorted by order from small to large, obtain undetermined Position distance set：d_ε={ d₁,d₂,d₃,...,d_i,...,d_n, wherein d₁＜ d₂＜ d₃＜ ... ＜ d_i＜ ... ＜ d_n；

6.2) with distance set d to be positioned_εEach interior element subtracts minimum value d₁, draw difference set d to be positioned_ε'：

d_ε'={ 0, Δ d₁,Δd₂,Δd₃,...,Δd_i,...,Δd_n,

Wherein：Δd_i=d_i-d₁, it represents node to be positioned to the distance between i-th iBeacon node value d_iWith minimum value d₁ Difference；

6.3) difference set d to be positioned is calculated_ε' in each element average value

$\mathrm{\&Delta;}\stackrel{\&OverBar;}{d}=\frac{\underset{i=1}{\overset{n}{\&Sigma;}}{\mathrm{\&Delta;d}}_i}{n-1};$

6.4) iBeacon nodes are respectively classified intoSet C₁WithSet C₂, it is located at set C₂Inside there is z Individual node, set C₂The middle immediate 3 iBeacon nodes of difference are divided into one group, are divided into z-2 groups, calculate every group of matter The heart, its coordinate (x_L,y_L) be：

$(x_L,y_L)=(\frac{x_{L1}+x_{L2}+x_{L3}}{3},\frac{y_{L1}+y_{L2}+y_{L3}}{3}),$

Wherein：L ∈ [1, z-2], (x_L1,y_L1)、(x_L2,y_L2)、(x_L3,y_L3) it is 3 iBeacon node coordinates in every group；

6.5) by step 6.4) barycenter that obtains is considered as the iBeacon nodes of new addition again, then by z-2 center of mass point and set C₁N-z interior iBeacon node constitutes polygon, and this polygonal barycenter is obtained using centroid algorithm, is calculated and treats Estimated coordinates W (x, y) of positioning node.