CN107295538A - Position the computational methods and the localization method and position indicator using confidence level of confidence level - Google Patents
Position the computational methods and the localization method and position indicator using confidence level of confidence level Download PDFInfo
- Publication number
- CN107295538A CN107295538A CN201610191108.2A CN201610191108A CN107295538A CN 107295538 A CN107295538 A CN 107295538A CN 201610191108 A CN201610191108 A CN 201610191108A CN 107295538 A CN107295538 A CN 107295538A
- Authority
- CN
- China
- Prior art keywords
- positioning result
- confidence level
- characteristic vector
- grid
- target
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W24/00—Supervisory, monitoring or testing arrangements
- H04W24/02—Arrangements for optimising operational condition
-
- 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
- G01S11/00—Systems for determining distance or velocity not using reflection or reradiation
- G01S11/02—Systems for determining distance or velocity not using reflection or reradiation using radio waves
- G01S11/06—Systems for determining distance or velocity not using reflection or reradiation using radio waves using intensity measurements
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W24/00—Supervisory, monitoring or testing arrangements
- H04W24/08—Testing, supervising or monitoring using real traffic
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W64/00—Locating users or terminals or network equipment for network management purposes, e.g. mobility management
- H04W64/003—Locating users or terminals or network equipment for network management purposes, e.g. mobility management locating network equipment
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Mathematical Physics (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Optimization (AREA)
- Mathematical Analysis (AREA)
- Computational Mathematics (AREA)
- Evolutionary Biology (AREA)
- Remote Sensing (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Operations Research (AREA)
- Probability & Statistics with Applications (AREA)
- Life Sciences & Earth Sciences (AREA)
- Algebra (AREA)
- Bioinformatics & Computational Biology (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Collating Specific Patterns (AREA)
- Position Fixing By Use Of Radio Waves (AREA)
- Length Measuring Devices With Unspecified Measuring Means (AREA)
Abstract
Present disclose provides the computational methods to the positioning confidence level in fingerprint location, and use the fingerprint positioning method and fingerprint location instrument of positioning confidence level raising positional accuracy.The computational methods of positioning confidence level according to embodiments of the present invention include:Obtain the actual measurement characteristic vector of target to be positioned;According to the positioning result of target to be positioned, the set of eigenvectors of the peripheral position of the positioning result is obtained;Calculate the Euclidean distance between the actual measurement characteristic vector of target to be positioned and the set of eigenvectors of the peripheral position;And, the confidence level of the positioning result is calculated according to the Euclidean distance calculated.The confidence level of fingerprint location result can be effectively judged by the embodiment of the present disclosure.
Description
Technical field
The present invention relates to the location fingerprint positioning field in any wide-area deployment, more particularly, to
The determination method of positioning confidence level in fingerprint location and the fingerprint location side using positioning confidence level
Method and location equipment.
Background technology
Fingerprint location is a kind of emerging location technology.The principle of fingerprint location is undetermined by calculating
The similitude of the signal strength values vector of all sample points of signal strength values vector sum in site is obtained
Final positioning result.
Fingerprint positioning method is generally divided into two steps:Off-line training modelling phase and on-line prediction are fixed
The position stage.The off-line training modelling phase refers in advance using professional measuring apparatus to target area by one
Fixed space density carries out the collection of finger print information, then according to the fingerprint of collection and the ground of collection point
Manage the fingerprint database that coordinate data generates whole region;On-line prediction positioning stage, wireless terminal
Finger print information measurement is carried out in target point, then the fingerprint and fingerprint database are compared again
Match somebody with somebody, estimate the geographical position of the target point.
In order to make it easy to understand, some abbreviation terms that will be used herein are briefly described below:
RSSI:The signal intensity of reception indicates (Received Signal Strength
Indication)。
SVM:Support vector machines (Support Vector Machine).In machine learning field,
SVM, which is one, the learning model of supervision, commonly used to carry out pattern-recognition, classification and return
Return analysis.
kNN:Being used in fingerprint location determines that one kind of active user position commonly uses resolving side
Method, detailed process is:K nearest reference point of distance test point Euclidean distance is found first, its
Coordinate can with Pm=(Xm, Ym, Zm) represent, wherein m value be 1~k between,
Then positioning result P is the average of the k reference point locations, i.e. P=∑s (Xm, Ym, Zm)/k.
In the prior art, the alignment system of comparative maturity such as GNSS system can be in the same of positioning
When there is provided the judge criterion of the system positioning performance, for example pass through geometric dilution of precision GDOP equivalent
Estimate come the precision to positioning result.The judge value of the positioning performance is for data fusion, connection
That closes the technologies such as positioning realizes important in inhibiting.But current fingerprint location only exports positioning for user
As a result, the estimation of positioning performance judge value can not be provided as above-mentioned GNSS system, this into
Important technical bottleneck when fingerprint location technology is merged with other location technologies.
The content of the invention
In view of the above mentioned problem of existing fingerprint location technology, the present invention proposes a kind of fixed to fingerprint
The computational methods of positioning confidence level that positional accuracy in position is judged, and also proposed and make
The method and position indicator of the positional accuracy of fingerprint location are improved with positioning confidence level.
The disclosure is mainly proposed:In fingerprint location, by obtaining the characteristic vector of geographic grid,
And the region to be checked amplified centered on the positioning result of target to be positioned, by target to be positioned
The set of eigenvectors of the vectorial grids included with region to be checked of actual measurement RSSI carry out Euclidean one by one
Distance is calculated, and resulting Euclidean distance summation is asked down again, that is, obtains the credible of the positioning result
The estimate of degree, the value is more big then to indicate that confidence level is higher.The method provided by the disclosure can have
Judge the confidence level of fingerprint location result in effect ground.
According to the first aspect of the disclosure, there is provided a kind of calculating of the positioning confidence level of fingerprint location
Method, including:Obtain the actual measurement characteristic vector of target to be positioned;According to the positioning of target to be positioned
As a result, the set of eigenvectors of the peripheral position of the positioning result is obtained;Calculate target to be positioned
Survey the Euclidean distance between characteristic vector and the set of eigenvectors of the peripheral position;And, root
The confidence level of the positioning result is calculated according to the Euclidean distance calculated.
In certain embodiments, the signal intensity that the characteristic vector includes receiving indicates that RSSI is special
Levy vector.
In certain embodiments, the feature of the peripheral position of position indicated by the positioning result is obtained
Vector set includes:Centered on the positioning result of target to be positioned, to around amplifying predetermined model
Enclose, regard the region after amplification as region to be checked;And, obtain what the region to be checked was included
The characteristic vector of each grid, to generate the set of eigenvectors of the peripheral position.
In certain embodiments, the actual measurement characteristic vector and the peripheral position of target to be positioned are calculated
Set of eigenvectors between Euclidean distance include:Calculate respectively the actual measurement feature of target to be positioned to
Amount and the peripheral position characteristic vector concentrate each grid characteristic vector between Euclidean away from
From.
In certain embodiments, the credible of the positioning result is calculated according to the Euclidean distance calculated
Degree includes:Euclidean distance summation to calculating respectively;And, with the Euclidean distance sum calculated
The confidence level as the positioning result reciprocal.
In certain embodiments, the method for calculating confidence level also includes off-line training step, institute
Stating off-line training step includes:Area to be targeted is divided into several grids according to physical space;
For each grid, the training set for the grid is generated by off-line training, and by right
RSSI value in all each dimensions of characteristic vector sample in the training set is averaging to obtain
The characteristic vector of the grid.
According to the second aspect of the disclosure there is provided a kind of fingerprint positioning method, including:According to preceding
State the positioning confidence level for the current positioning result of method calculating for calculating confidence level;Judge to work as prelocalization knot
Whether the confidence level of fruit is more than thresholding;If the confidence level of current positioning result is more than or equal to thresholding,
Then export current positioning result;If the confidence level of current positioning result is less than thresholding, obtain described
The grid minimum with the Euclidean distance of current positioning result in peripheral position, and by acquired net
The corresponding position of lattice as the target to be positioned after renewal current positioning result;Repeat the above,
Until output confidence level is more than the positioning result of thresholding.
According to the third aspect of the disclosure there is provided a kind of fingerprint location instrument, including:
Receiver, is configured to:Receive signal;
Processing unit, is configured to:According to signal is received, determine the actual measurement feature of target to be positioned to
Amount;According to the positioning result of target to be positioned, the feature of the peripheral position of the positioning result is obtained
Vector set;Calculate the actual measurement characteristic vector of target to be positioned and the set of eigenvectors of the peripheral position
Between Euclidean distance;And, according to the Euclidean distance that is calculated calculate the positioning result can
Reliability.
In certain embodiments, the fingerprint location instrument can also include:Memory, is configured to deposit
Store up the grid search-engine vector of area to be targeted.The grid search-engine vector of the area to be targeted can lead to
Cross following manner determination:Area to be targeted is divided into several grids according to physical space;For
Each grid, the training set for the grid is generated by off-line training, and by described
RSSI value in all each dimensions of characteristic vector sample in training set is averaging described to obtain
The characteristic vector of grid.
The position indicator can also include:Output unit, is configured to export positioning result and institute together
State the confidence level of positioning result.
In certain embodiments, the processing unit in the fingerprint location instrument is additionally configured to:Judge to work as
Whether the confidence level of prelocalization result is more than thresholding;If the confidence level of current positioning result is more than etc.
In thresholding, then current positioning result is exported;If the confidence level of current positioning result is less than thresholding,
Obtain the grid minimum with the Euclidean distance of current positioning result in the peripheral position, and by institute
The corresponding position of grid of acquisition as the target to be positioned after renewal current positioning result;Repeat
Above-mentioned processing, until output confidence level is more than the positioning result of thresholding.
Present disclose provides the confidence level computational methods that can effectively judge fingerprint location result.According to
The fingerprint location instrument of the embodiment of the present disclosure can export confidence level valuation together with positioning result, overcome
Technical bottleneck when fingerprint location technology is merged with other location technologies.
Brief description of the drawings
Illustrate preferred embodiment of the present disclosure below in conjunction with the accompanying drawings, the above and other of the present invention will be made
Objects, features and advantages are clearer, wherein:
Fig. 1 shows the positioning confidence level of the calculating fingerprint location according to one embodiment of the disclosure
Method flow chart;
The method that Fig. 2 shows the positioning confidence level of the calculating fingerprint location according to the embodiment of the present disclosure
The schematic diagram that implements;
Fig. 3 A and Fig. 3 A show the schematic diagram of two examples in region to be checked;
Fig. 4 A and Fig. 4 A show two examples in the region to be checked for being labelled with grid search-engine vector
Schematic diagram;
Fig. 5 shows the flow chart of the fingerprint positioning method of one embodiment according to the disclosure;
Fig. 6 show according to one of the fingerprint positioning method of the embodiment of the present disclosure implement show
It is intended to;And
Fig. 7 shows the block diagram of the fingerprint location instrument according to one embodiment of the disclosure.
In all accompanying drawings of the disclosure, same or analogous reference mark represents identical or phase
As key element.
Embodiment
The principle and essence of the disclosure are described with reference to some exemplary embodiments below in conjunction with accompanying drawing
God.It should be appreciated that providing these embodiments merely to enabling those skilled in the art preferably
Understand and then realize the disclosure, and not limit the scope of the present disclosure in any way.In addition, being
For the sake of easy, the detailed description pair with known technology of the present invention without direct correlation is eliminated,
To prevent the understanding of the present invention from causing to obscure.
Terms used herein is only used for describing exemplary embodiment, and is not intended to the exemplary reality of limitation
Apply example.As used herein, unless explicitly pointed out in context, otherwise singulative is not excluded for
Plural form can be included.It should also be understood that ought be in this manual in use, "and/or" includes
Correlation lists one or more any and all combinations of item.Term " comprising " and/or " having "
There is cited feature, numeral, step, operation, component, element or its combination in regulation, and
Other one or more features, numeral, step, operation, component, member are not precluded the presence or addition of
Element or its combination.
Unless otherwise defined explicitly, otherwise all terms used herein have and exemplary embodiment
The identical implication that those of ordinary skill in art is generally understood.It is also understood that removing
Non- separately to explicitly define herein, this area is general when term should be interpreted as having and invent the time
The consistent implication of implication in the specification that logical technical staff is understood.
Illustrate to position confidence level according to the calculating of the embodiment of the present disclosure below with reference to Fig. 1~Fig. 4 B
Method.
Fig. 1 shows the positioning confidence level of the calculating fingerprint location according to one embodiment of the disclosure
Method 100 flow chart.
When having been obtained for the positioning result of target to be positioned by fingerprint location resolving, Ke Yikai
Beginning calculates the method 100 of the confidence level of the positioning result.Herein, it can be referred to by any of
Line positioning calculation algorithm (such as kNN algorithms) obtains the positioning result of target to be positioned, the present invention
It is unrestricted in this regard.
In step s 110, the actual measurement characteristic vector of target to be positioned is obtained.Target to be positioned
Characteristic vector is typically RSSI characteristic vectors, and it can include one or more dimensions.Hereinafter
The characteristic vector mentioned is often referred to RSSI characteristic vectors, unless otherwise expressing.
In the step s 120, according to the positioning result of target to be positioned, the positioning result is obtained
Peripheral position set of eigenvectors.Specifically, can be using the positioning result of target to be positioned in
The heart, to preset range is around amplified, regard the region after amplification as region to be checked.It is then possible to
The area to be checked is obtained from the fingerprint database of the other equipment storage in local storage or system
The characteristic vector for each grid that domain is included.The characteristic vector of all grids in the region to be checked
Set form the set of eigenvectors of the peripheral position.
It should be understood that being generally previously stored with the fingerprint database of area to be targeted in systems.Such as
Previously mentioned, fingerprint positioning method is generally divided into two steps:The off-line training modelling phase and
Line predicts positioning stage.In the off-line training modelling phase, the off-line training modelling phase can use in advance
Professional measuring apparatus will by certain space density to target area (hereinafter referred to as area to be targeted)
Physical space is divided into several grids, and the RSSI for gathering each grid (also referred to as collection point) is special
Vector (also referred to as finger print information) is levied, then according to the fingerprint of collection and the geographical coordinate number of collection point
According to the fingerprint database for generating whole area to be targeted.Entirely treated that is, fingerprint database will be included
The characteristic vector of all grids of localization region.
In step s 130, the actual measurement characteristic vector and the peripheral position of target to be positioned are calculated
Set of eigenvectors between Euclidean distance.It should be understood that the set of eigenvectors of the peripheral position
Including one or more vectors.The actual measurement characteristic vector of target to be positioned and the spy of the peripheral position
The Euclidean distance levied between vector set also includes one or more Euclidean distances, and it includes mesh to be positioned
Target surveys the Europe of characteristic vector and each characteristic vector of the characteristic vector concentration of the peripheral position
Family name's distance, will form a set.
For example, calculating the actual measurement characteristic vector of target to be positioned and the characteristic vector of the peripheral position
Euclidean distance between collection can include calculating respectively the actual measurement characteristic vector of target to be positioned with it is described
Euclidean distance between the characteristic vector of each grid in region to be checked.
In step S140, the credible of the positioning result is calculated according to the Euclidean distance calculated
Degree.
For example, the actual measurement feature for the target to be positioned that can be calculated in calculation procedure S130 first to
The Euclidean distance sum for each characteristic vector that amount and the characteristic vector of the peripheral position are concentrated, so
Afterwards can estimating using the confidence level reciprocal as the positioning result of the Euclidean distance sum calculated
Value.It should be understood that can also be used as with other functions of the Euclidean distance sum calculated credible
The valuation of degree, can be for example used as with any inverse ratio function of the Euclidean distance sum calculated can
The valuation of reliability.In this case, the valuation of confidence level is higher, shows that confidence level is higher, and
When confidence level valuation is lower, show that confidence level is lower.
Alternatively it is also possible to the Euclidean distance sum calculated is in itself, or its direct ratio function is made
For the valuation of confidence level.But, in this case, valuation is higher, shows that confidence level is lower, and
Valuation is lower, shows that confidence level is higher.
It will be understood by those skilled in the art that method 100 illustrated above is only exemplary.This
The method of invention is not limited to step and order illustrated above.Those skilled in the art are according to institute
Show that the teaching of embodiment can carry out many and change and modifications.
For example, single step can be split as execution of multiple steps, and some steps can also merge
Performed into single step.
And for example, alternatively, method 100 can also include off-line training step, the off-line training
Step is used to generate the fingerprint database in region to be checked.The off-line training step is typically in fingerprint
What the off-line training modelling phase of localization method completed.The off-line training step can include:It will treat
Localization region is divided into several grids according to physical space;For each grid, by instructing offline
Practice generation for the grid training set, and by all features in the training set to
RSSI value in the amount each dimension of sample is averaging to obtain the characteristic vector of the grid.
The method 100 according to the embodiment of the present disclosure is specifically described below with reference to Fig. 2.Fig. 2 is shown
It is specific according to one of the method 100 of the positioning confidence level of the calculating fingerprint location of the embodiment of the present disclosure
Realize 100 ' schematic diagram.
Similar with method 100, method 100 ' starts by fingerprint location resolving and had been obtained for
During the Primary Location result of target to be positioned.
As shown in Fig. 2 the computational methods 100 ' of positioning confidence level mainly include three steps:
1. obtain the actual measurement characteristic vector (the step S110 for being similar to Fig. 1) of target to be positioned
2. according to Primary Location position, obtain the set of eigenvectors of Primary Location position peripheral position
(the step S120 for being similar to Fig. 1).
3. between the actual measurement characteristic vector and peripheral position set of eigenvectors that calculate target to be positioned
Euclidean distance, the confidence level for then calculating positioning result according to Euclidean distance (is similar to Fig. 1 step
Rapid S130 and S140).
These three steps are specifically described respectively below with reference to Fig. 2.
1. obtain the actual measurement characteristic vector of target to be positioned
The RSSI characteristic vectors of target to be positioned are labeled as R_W, wherein including WCIndividual dimension.
2. obtain the set of eigenvectors of Primary Location position peripheral position
As illustrated, the step 2 mainly includes following three sub-steps.
1. save mesh characteristic vector
- area to be targeted according to physical space is divided into several grids, each grid is entered
Line number, then each grid have corresponding numbering.
- in this example, C grid is divided into area to be targeted altogether, each grid
Numbering is designated as:A1, A2, A3..., AC.It can be calculated according to the training set of each grid
Go out the characteristic vector of the grid.
- for the grid A in area to be targetedi(1 <=i <=C), training set is designated as { ai(k)}
(wherein the number of training set is Ki, 1 <=k <=Ki), each in the training set
RSSI characteristic vectors ai(k) W is included inCIndividual dimension.To the K in training setiIndividual spy
The correspondence of the RSSI value in vectorial each dimension is levied to average, obtain the feature of the grid to
Measure Ri, it is expressed as follows:
1. sub-step can complete in the off-line training modelling phase.Grid search-engine in area to be targeted
Vector can be stored as fingerprint database together.
2. region to be checked is obtained
- (obtained with the positioning result L of target to be positioned by " fingerprint location resolving module "
Positioning result L) centered on, the preset range of surrounding is amplified to, by the region after amplification
It is used as region to be checked.The preset range can be the first adjacent ranges as shown in Figure 3A
Or the second adjacent ranges as shown in Figure 3 B.It should be understood that the amplification range is not limited to
Shown example, can be other values.The present invention is unrestricted in this regard.
The grid covered in-region to be checked is as grid to be checked.
3. the set of eigenvectors for the grid that region to be checked is included is obtained
- numbering of the grid that region to be checked is included is obtained in 2., according to the grid
1. numbering is combined can obtain the set of eigenvectors { R } for the grid that region to be checked is included.Example
Such as, the grid that reference numeral is obtained from the fingerprint database of storage is numbered according to grid
Characteristic vector.Fig. 4 A and 4B respectively illustrate treat corresponding with Fig. 3 A and Fig. 3 B
Grid is examined, wherein being labelled with the characteristic vector R of the grid in each gridi。
3. the confidence level of positioning result is calculated according to Euclidean distance
As illustrated, the step 3 mainly includes following two sub-steps.
4. actual measurement characteristic vector and the Euclidean distance of grid search-engine vector set to be checked are calculated
- vector X (X=[X1, X2..., Xn]) and vector Y (Y=[Y1, Y2..., Yn]) between Europe
Family name can be calculated as follows apart from d's:
The actual measurement RSSI characteristic vectors of-target to be positioned obtained in step 1 are R_W
- when selecting the first adjacent ranges, the set of eigenvectors for the grid that region to be checked is included
For Ril, Ri2, Ri3, Ri4, Ri5, Ri6... Ri9
- calculate respectively the actual measurement RSSI characteristic vectors of target to be positioned and above-mentioned 9 features to
Euclidean distance d (R_W, R between amountil), d (R_W, Ri2), d (R_W, Ri3) ...
D (R_W, Ri9)。
5. the confidence level D of current positioning result is calculated
Above combined Fig. 1~Fig. 4 B to according to the present invention location Calculation degree method carry out
Explanation.
Using Credibility judgement proposed by the present invention, locating accuracy can be improved.Then, this public affairs
Open a kind of method that utilization confidence level improves the fingerprint location degree of accuracy that also proposed.Below with reference to Fig. 5
It is described with Fig. 6.
Fig. 5 shows the flow of the fingerprint positioning method 200 according to one embodiment of the disclosure
Figure.
, can when having been obtained for the initial alignment result of target to be positioned by fingerprint location resolving
With start method 200.
In step S210, according to foregoing confidence level computational methods (such as method 100 or 100 ')
Calculate the positioning confidence level of current positioning result.
In step S220, judge whether the confidence level of current positioning result is more than or equal to thresholding.
The thresholding can be default.
If the confidence level that the judged result in step S220 is current positioning result is more than thresholding,
Step S230 is proceeded to, current positioning result is exported.
If the confidence level that the judged result in step S220 is current positioning result is less than or equal to door
Limit, then proceed to step S240.In step S240, obtain peripheral position in working as prelocalization
As a result the minimum grid of Euclidean distance, and using the corresponding position of acquired grid as after renewal
Target to be positioned current positioning result.The peripheral position used in step S240 in calculating with working as
The peripheral position (region i.e. to be checked) determined during the confidence level of prelocalization result is consistent.
In step S240, after the current positioning result that have updated target to be positioned, step is returned to
Rapid S210, recalculates the confidence level of the current positioning result after updating.
Updated by above-mentioned constantly iteration, final output confidence level is more than the positioning result of thresholding.
Fig. 6 shows that one of above-mentioned fingerprint positioning method 200 implements 200 ' schematic diagram.
Similar with method 200, method 200 ' starts by fingerprint location resolving and had been obtained for
During the Primary Location result of target to be positioned.
As shown in fig. 6, the localization method 200 ' mainly includes following four steps:
1. judge whether the confidence level of current positioning result is more than thresholding
2. if greater than thresholding, then current positioning result is exported
3. if less than thresholding, the corresponding grid numbering L ' of Euclidean distance minimum value is obtained
[dmin, GridID]=min ([d (R_W, Ril), d (R_W, Ri2), d (R_W, Ri3) ..., d (R_W, R19)])
L '=GridIDFormula (4)
4. the positioning result of target to be positioned is updated, i.e., regard grid L ' as the positioning knot updated
Really
L=L '
Confidence level calculating is re-started, so that the confidence level of the positioning result after being updated.So
After repeat the above steps.
Updated by above-mentioned continuous iteration, most output confidence level is more than the positioning result of thresholding at last.
So as to which the positioning knot of high accuracy can be provided by the fingerprint positioning method of the embodiment of the present disclosure
Really.
Fig. 7 shows the block diagram of the fingerprint location instrument 300 according to one embodiment of the disclosure.
As illustrated, the fingerprint location instrument 300 can include receiver 310, processing unit 320,
Output unit 330 and memory 340.
Receiver 310 is configured to receive signal.The receiver 310 can be positioned with conventional fingerprint
The function of the signal receiver of instrument is similar.
Processing unit 320 is configured to calculate the confidence level of positioning result.Specifically, processing unit
320 are configurable to:According to signal is received, the actual measurement characteristic vector of target to be positioned is determined;Root
According to the positioning result of target to be positioned, the set of eigenvectors of the peripheral position of the positioning result is obtained;
Calculate the Europe between the actual measurement characteristic vector of target to be positioned and the set of eigenvectors of the peripheral position
Family name's distance;And, the confidence level of the positioning result is calculated according to the Euclidean distance calculated.
Certainly, also the fingerprint location with conventional fingerprint position indicator resolves function to processing unit 320.
Any existing fingerprint computation (such as kNN algorithms) can be used in the present invention.The present invention is at this
Aspect is unrestricted.
Output unit 330 can not only export positioning result, additionally it is possible to be exported with positioning result instrument
The confidence level of positioning result.
Memory 340 is configured to store the grid search-engine vector of area to be targeted, such as area to be targeted
Fingerprint database.
For example, what the grid search-engine vector of the area to be targeted can be determined by following manner:
Area to be targeted is divided into several grids according to physical space;For each grid, by from
Line training generation is directed to the training set of the grid, and by all spies in the training set
The RSSI value in each dimension of vectorial sample is levied to be averaging to obtain the characteristic vector of the grid.
In a preferred embodiment, processing unit 320 is additionally configured to:Judge current positioning result
Confidence level whether be more than thresholding;If the confidence level of current positioning result is more than or equal to thresholding, defeated
Go out current positioning result;If the confidence level of current positioning result is less than thresholding, the periphery is obtained
The grid minimum with the Euclidean distance of current positioning result in position, and by acquired grid pair
The position answered as the target to be positioned after renewal current positioning result.Processing unit 320 will weight
Multiple above-mentioned processing, until current positioning result confidence level is more than thresholding, just passes through output unit 330
Export positioning result.
Fingerprint location instrument 300 according to embodiments of the present invention can be used for performing according to present invention implementation
The computational methods (such as method 100 and 100 ') of the positioning confidence level of example, and/or according to the present invention
The use confidence level implemented improves fingerprint positioning method (such as 200 Hes of positional accuracy
200’).The concrete operations of fingerprint location instrument 300 may be referred to the description of the above method, herein not
Repeat again.
Above combined preferred embodiment invention has been described.Those skilled in the art can
To understand, apparatus and method illustrated above are only exemplary.The equipment of the present invention can include
The more or less parts of part than showing.The method of the present invention is not limited to illustrated above
Step and order.Those skilled in the art can carry out many changes according to the teaching of illustrated embodiment
And modification.
The equipment and its part of the present invention can be by such as super large-scale integration OR gate array, all
It is such as semiconductor of logic chip, transistor or such as field programmable gate array, programmable
The hardware circuit realization of the programmable hardware device of logical device etc., can also be with by various types of
The software of computing device realizes, can also being implemented in combination with by above-mentioned hardware circuit and software.
Although it should be appreciated by those skilled in the art that describe the present invention by specific embodiment,
It is that the scope of the present invention is not limited to these specific embodiments.The scope of the present invention is by appended claims
And its any equivalents are limited.
Claims (15)
1. a kind of computational methods of the positioning confidence level of fingerprint location, including:
Obtain the actual measurement characteristic vector of target to be positioned;
According to the positioning result of target to be positioned, the feature of the peripheral position of the positioning result is obtained
Vector set;
Calculate the actual measurement characteristic vector of target to be positioned and the peripheral position set of eigenvectors it
Between Euclidean distance;And
The confidence level of the positioning result is calculated according to the Euclidean distance calculated.
2. according to the method described in claim 1, wherein the characteristic vector includes the letter received
Number intensity indicates RSSI characteristic vectors.
3. according to the method described in claim 1, wherein being calculated according to the Euclidean distance calculated
The confidence level of the positioning result includes:
Euclidean distance summation to being calculated;And
Using the confidence level reciprocal as the positioning result of the Euclidean distance sum calculated.
4. according to method according to any one of claims 1 to 3, wherein, obtain the positioning
As a result the set of eigenvectors of the peripheral position of indicated position includes:
Centered on the positioning result of target to be positioned, to preset range is around amplified, it will put
Region after big is used as region to be checked;And
The characteristic vector for each grid that the region to be checked is included is obtained, to generate the periphery
The set of eigenvectors of position.
5. method according to claim 4, in addition to off-line training step, described offline
Training step includes:
Area to be targeted is divided into several grids according to physical space;
For each grid, the training set for the grid, Yi Jitong are generated by off-line training
Cross and the RSSI value in all each dimensions of characteristic vector sample in the training set is averaging to come
Obtain the characteristic vector of the grid.
6. method according to claim 4, calculates the actual measurement characteristic vector of target to be positioned
Euclidean distance between the set of eigenvectors of the peripheral position includes:
The actual measurement characteristic vector of target to be positioned and the characteristic vector of the peripheral position are calculated respectively
Euclidean distance between the characteristic vector for each grid concentrated.
7. a kind of fingerprint positioning method, including:
The positioning of current positioning result is calculated according to method according to any one of claims 1 to 6
Confidence level;
Judge whether the confidence level of current positioning result is more than thresholding;
If the confidence level of current positioning result is more than or equal to thresholding, current positioning result is exported;
If the confidence level of current positioning result is less than thresholding, obtain in the peripheral position with
The grid of the Euclidean distance minimum of current positioning result, and the acquired corresponding position of grid is made
For the current positioning result of the target to be positioned after renewal;
Repeat the above, until output confidence level is more than the positioning result of thresholding.
8. a kind of fingerprint location instrument, including:
Receiver, is configured to:Receive signal;
Processing unit, is configured to:
According to signal is received, the actual measurement characteristic vector of target to be positioned is determined;
According to the positioning result of target to be positioned, the peripheral position of the positioning result is obtained
Set of eigenvectors;
Calculate the actual measurement characteristic vector of target to be positioned and the characteristic vector of the peripheral position
Euclidean distance between collection;And
The confidence level of the positioning result is calculated according to the Euclidean distance calculated.
9. fingerprint location instrument according to claim 8, wherein the characteristic vector includes connecing
The signal intensity of receipts indicates RSSI characteristic vectors.
10. fingerprint location instrument according to claim 8, wherein according to the Euclidean calculated away from
Include from the confidence level for calculating the positioning result:
Euclidean distance summation to being calculated;And
Using the confidence level reciprocal as the positioning result of the Euclidean distance sum calculated.
11. the fingerprint location instrument according to any one of claim 8~10, wherein, obtain
The set of eigenvectors of the peripheral position of position includes indicated by the positioning result:
Centered on the positioning result of target to be positioned, to preset range is around amplified, it will put
Region after big is used as region to be checked;And
The characteristic vector for each grid that the region to be checked is included is obtained, to generate the periphery
The set of eigenvectors of position.
12. fingerprint location instrument according to claim 11, in addition to:
Memory, is configured to store the grid search-engine vector of area to be targeted, the area to be targeted
Grid search-engine vector determined by following manner:
Area to be targeted is divided into several grids according to physical space;
For each grid, the training set for the grid is generated by off-line training, with
And by the RSSI value in all each dimensions of characteristic vector sample in the training set
It is averaging to obtain the characteristic vector of the grid.
13. fingerprint location instrument according to claim 11, wherein, calculate target to be positioned
Actual measurement characteristic vector and the peripheral position set of eigenvectors between Euclidean distance include:
The actual measurement characteristic vector of target to be positioned and the characteristic vector of the peripheral position are calculated respectively
The distance between characteristic vector of each grid concentrated.
14. fingerprint location instrument according to claim 8, in addition to:Output unit, configuration
To export the confidence level of positioning result and the positioning result together.
15. fingerprint location instrument according to claim 14,
Wherein described processing unit is additionally configured to:
Judge whether the confidence level of current positioning result is more than thresholding;
If the confidence level of current positioning result is more than or equal to thresholding, prelocalization knot is worked as in output
Really;
If the confidence level of current positioning result is less than thresholding, obtain in the peripheral position
The grid minimum with the Euclidean distance of current positioning result, and by acquired grid correspondence
Position as the target to be positioned after renewal current positioning result;
Repeat the above, until output confidence level is more than the positioning result of thresholding.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610191108.2A CN107295538B (en) | 2016-03-30 | 2016-03-30 | Positioning reliability calculation method, positioning method using reliability and positioning instrument |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610191108.2A CN107295538B (en) | 2016-03-30 | 2016-03-30 | Positioning reliability calculation method, positioning method using reliability and positioning instrument |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107295538A true CN107295538A (en) | 2017-10-24 |
CN107295538B CN107295538B (en) | 2022-03-18 |
Family
ID=60087568
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610191108.2A Active CN107295538B (en) | 2016-03-30 | 2016-03-30 | Positioning reliability calculation method, positioning method using reliability and positioning instrument |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107295538B (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108810819A (en) * | 2018-05-25 | 2018-11-13 | 厦门华方软件科技有限公司 | A kind of earth's surface localization method and medium based on network communication |
CN109982242A (en) * | 2019-03-29 | 2019-07-05 | 深圳市九洲电器有限公司 | A kind of indoor orientation method, device, base station and system |
CN110113708A (en) * | 2018-04-18 | 2019-08-09 | 爱动超越人工智能科技(北京)有限责任公司 | Localization method and device based on Wi-Fi location fingerprint |
CN110493867A (en) * | 2019-06-27 | 2019-11-22 | 湖南大学 | A kind of signal behavior and the wireless indoor location method of position correction |
Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2009068530A1 (en) * | 2007-11-30 | 2009-06-04 | Siemens Aktiengesellschaft | A locating method and a locating system |
CN101639527A (en) * | 2009-09-03 | 2010-02-03 | 哈尔滨工业大学 | K nearest fuzzy clustering WLAN indoor locating method based on REE-P |
US20110250905A1 (en) * | 2008-12-17 | 2011-10-13 | Wigren Karl Torbjoern | Methods and Arrangements for Fingerprinting Positioning |
CN103200678A (en) * | 2013-04-09 | 2013-07-10 | 南京信息工程大学 | Android device wireless fidelity (WiFi) indoor locating method based on position fingerprint identification algorithm |
CN103596267A (en) * | 2013-11-29 | 2014-02-19 | 哈尔滨工业大学 | Fingerprint map matching method based on Euclidean distances |
CN103763769A (en) * | 2013-12-26 | 2014-04-30 | 北京邮电大学 | Indoor fingerprint positioning method based on access point reselection and self-adaptation cluster splitting |
CN104424276A (en) * | 2013-08-30 | 2015-03-18 | 中国电信集团公司 | Method and device for self-updating fingerprint database based on manifold learning |
CN104469676A (en) * | 2014-11-21 | 2015-03-25 | 北京拓明科技有限公司 | Method and system for locating mobile terminal |
CN104502889A (en) * | 2014-12-29 | 2015-04-08 | 哈尔滨工业大学 | Reference point maximum range based positioning reliability calculation method in fingerprint positioning |
CN104703143A (en) * | 2015-03-18 | 2015-06-10 | 北京理工大学 | Indoor positioning method based on WIFI signal strength |
CN105424030A (en) * | 2015-11-24 | 2016-03-23 | 东南大学 | Fusion navigation device and method based on wireless fingerprints and MEMS sensor |
-
2016
- 2016-03-30 CN CN201610191108.2A patent/CN107295538B/en active Active
Patent Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2009068530A1 (en) * | 2007-11-30 | 2009-06-04 | Siemens Aktiengesellschaft | A locating method and a locating system |
US20110250905A1 (en) * | 2008-12-17 | 2011-10-13 | Wigren Karl Torbjoern | Methods and Arrangements for Fingerprinting Positioning |
CN101639527A (en) * | 2009-09-03 | 2010-02-03 | 哈尔滨工业大学 | K nearest fuzzy clustering WLAN indoor locating method based on REE-P |
CN103200678A (en) * | 2013-04-09 | 2013-07-10 | 南京信息工程大学 | Android device wireless fidelity (WiFi) indoor locating method based on position fingerprint identification algorithm |
CN104424276A (en) * | 2013-08-30 | 2015-03-18 | 中国电信集团公司 | Method and device for self-updating fingerprint database based on manifold learning |
CN103596267A (en) * | 2013-11-29 | 2014-02-19 | 哈尔滨工业大学 | Fingerprint map matching method based on Euclidean distances |
CN103763769A (en) * | 2013-12-26 | 2014-04-30 | 北京邮电大学 | Indoor fingerprint positioning method based on access point reselection and self-adaptation cluster splitting |
CN104469676A (en) * | 2014-11-21 | 2015-03-25 | 北京拓明科技有限公司 | Method and system for locating mobile terminal |
CN104502889A (en) * | 2014-12-29 | 2015-04-08 | 哈尔滨工业大学 | Reference point maximum range based positioning reliability calculation method in fingerprint positioning |
CN104703143A (en) * | 2015-03-18 | 2015-06-10 | 北京理工大学 | Indoor positioning method based on WIFI signal strength |
CN105424030A (en) * | 2015-11-24 | 2016-03-23 | 东南大学 | Fusion navigation device and method based on wireless fingerprints and MEMS sensor |
Non-Patent Citations (5)
Title |
---|
XIAOLONG XU等: "Variance-based fingerprint distance adjustment algorithm for indoor localization", 《JOURNAL OF SYSTEMS ENGINEERING AND ELECTRONICS》 * |
傅韬: ""基于测量报告的移动终端定位算法研究"", 《中国博士学位论文全文数据库 信息科技辑》 * |
史伟光等: "基于加权欧式算子的射频识别定位算法", 《计算机工程》 * |
屈阳等: "采用指纹SD权重的FM两级室内定位方法", 《导航定位学报》 * |
牛建伟等: "一种基于Wi-Fi信号指纹的楼宇内定位算法", 《计算机研究与发展》 * |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110113708A (en) * | 2018-04-18 | 2019-08-09 | 爱动超越人工智能科技(北京)有限责任公司 | Localization method and device based on Wi-Fi location fingerprint |
CN108810819A (en) * | 2018-05-25 | 2018-11-13 | 厦门华方软件科技有限公司 | A kind of earth's surface localization method and medium based on network communication |
CN108810819B (en) * | 2018-05-25 | 2021-05-14 | 厦门华方软件科技有限公司 | Earth surface positioning method and medium based on network communication |
CN109982242A (en) * | 2019-03-29 | 2019-07-05 | 深圳市九洲电器有限公司 | A kind of indoor orientation method, device, base station and system |
CN110493867A (en) * | 2019-06-27 | 2019-11-22 | 湖南大学 | A kind of signal behavior and the wireless indoor location method of position correction |
Also Published As
Publication number | Publication date |
---|---|
CN107295538B (en) | 2022-03-18 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106646338B (en) | A kind of quickly accurate indoor orientation method | |
CN109682382A (en) | Global fusion and positioning method based on adaptive Monte Carlo and characteristic matching | |
CN107295538A (en) | Position the computational methods and the localization method and position indicator using confidence level of confidence level | |
CN106199500B (en) | Fingerprint characteristic localization method and device | |
EP1532464A1 (en) | Error estimate concerning a target devic's location operable to move in a wireless environment | |
Ruan et al. | Hi-Loc: Hybrid indoor localization via enhanced 5G NR CSI | |
KR20120019435A (en) | Location detection system and method with fingerprinting | |
CN106658704A (en) | Positioning method and system of starting point of indoor positioning | |
Tao et al. | AIPS: An accurate indoor positioning system with fingerprint map adaptation | |
CN108427941A (en) | Method, method for detecting human face and device for generating Face datection model | |
CN109379711A (en) | A kind of localization method | |
US20130234894A1 (en) | Location estimation using radio scene signatures | |
CN106954190A (en) | A kind of WIFI indoor orientation methods based on index mapping domain | |
CN110837079A (en) | Target detection method and device based on radar | |
CN108111976B (en) | WiFi signal fingerprint data optimization method and device | |
CN105682039B (en) | A kind of RF fingerprint positioning method and system | |
CN111182460A (en) | Hybrid indoor positioning method and device, computer equipment and storage medium | |
CN110118979A (en) | The method of improved differential evolution algorithm estimation multipath parameter based on broad sense cross-entropy | |
Marques et al. | Crater delineation by dynamic programming | |
CN117095360A (en) | Food crop monitoring method and system based on SAR satellite remote sensing technology | |
CN102075970B (en) | Method for detecting sparse event of wireless sensor network by loop restructuring | |
CN110809284A (en) | Positioning method, system, device and readable storage medium based on MR data | |
CN111163514B (en) | Optimal area self-adaptive selection fingerprint positioning method and system | |
Rosić et al. | Hybrid genetic optimization algorithm for target localization using TDOA measurements | |
CN107071901A (en) | A kind of WIFI indoor orientation method the ratio between dual based on received signal strength |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |