CN106211319B - A kind of non-fingerprint passive type localization method based on WI-FI signal - Google Patents
A kind of non-fingerprint passive type localization method based on WI-FI signal Download PDFInfo
- Publication number
- CN106211319B CN106211319B CN201610575118.6A CN201610575118A CN106211319B CN 106211319 B CN106211319 B CN 106211319B CN 201610575118 A CN201610575118 A CN 201610575118A CN 106211319 B CN106211319 B CN 106211319B
- Authority
- CN
- China
- Prior art keywords
- measured
- value
- csi
- target
- height
- 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.)
- Active
Links
Classifications
-
- 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
- H04W64/006—Locating users or terminals or network equipment for network management purposes, e.g. mobility management with additional information processing, e.g. for direction or speed determination
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Position Fixing By Use Of Radio Waves (AREA)
Abstract
The invention discloses a kind of non-fingerprint passive type localization method based on WI-FI signal, this method comprises the following steps: step 1, constructing WI-FI transceiver network；Step 2, the CSI value in WI-FI signal is acquired；Step 3, filter preprocessing process；Step 4, the effective height and match parameter of object to be measured are estimated；Step 5, the pad value of object to be measured is obtained using match parameter, region locating for object to be measured is determined by its pad value；Step 6, target is positioned.The present invention solves the problems, such as to expend time and manpower during its fingerprint collecting compared to fingerprint positioning method, and robustness is good, practical；The present invention solves the problems, such as that existing non-fingerprint positioning method positioning accuracy is low by extracting the CSI information in WI-FI signal.
Description
Technical field
The present invention relates to the applications of positioning, and especially one kind is towards indoor non-fingerprint passive type localization method.
Background technique
As the activity in people space indoors becomes increasingly complex and finely, the demand day based on Indoor Location Information service
Tend to become strong strong.City positioning at present is broadly divided into active positioning and passive type positions two kinds, and the main distinction is to be positioned target
Whether relevant device is carried, and active positioning requirements are decided to be target Portable device, position to device location, and passive type
Positioning is that target does not need Portable device, is positioned according to target to the interference of characteristic information.Due to applying ring some
Under border, target is not intended to carry relevant device, and therefore, passive type scene has become a hot topic of research.
In passive type positioning, according to the location information of acquisition, positioning can be divided into based on the passive of non-electromagnetic signal
Formula positions and the passive type positioning based on electromagnetic signal.Positioning based on non-electromagnetic signal is using such as infrared ray, camera, biography
Sensor etc. carries out the determination of target position, these equipment since the selection of its feature has certain particularity, application scenarios also by
Many limitations are arrived.And in the positioning based on electromagnetic signal, the most commonly used is using RSS information to be positioned, it is divided into fingerprint side
Method positioning and the positioning of non-fingerprint method.Wherein fingerprint method expends time and people due to needing to carry out fingerprint collecting before positioning
Power, and robustness is bad, therefore not very practical；Rather than fingerprint location is due to the limitation of model and asking for RSS information itself
Topic, precision are poor.It is positioned in this regard, there is researcher to extract stable amplitude and phase information using special equipment, but
Special equipment needs special production (such as aerial array) or fancy price (such as USRP (Universal Software
Radio Peripheral, general software radio peripheral hardware)).Therefore part researcher extracts and determine CSI information
Position, but still have the following problems: 1, the not ready-made model of the positioning based on CSI information uses, and can only carry out fingerprint collecting and mention
It takes；2, based on CSI information since multi-subcarrier is mutually indepedent, general information expressing method can not be found.
Summary of the invention
For above-mentioned problems of the prior art, the present invention provides following scheme:
A kind of non-fingerprint passive type localization method based on WI-FI signal, comprising the following steps:
Step 1, WI-FI transceiver network, including transmitting terminal a and receiving end b are constructed；
Step 2, in the WI-FI transceiver network of building, the CSI value in WI-FI signal, the CSI value of the acquisition are acquired
It is divided into test CSI value and practical CSI value, wherein test CSI value are as follows:
There is no initial CSI value when target in (2-1-1) WI-FI transceiver network；
(2-1-2) is put into the CSI value when target of the first known altitude and position in WI-FI transceiver network；
(2-1-3) is put into the CSI value when target of the second known altitude and position in WI-FI transceiver network, wherein
The height of two known targets is different from the height of the first known target；
Practical CSI value refers to, when there are CSI values collected when object to be measured in WI-FI transceiver network；
Step 3, all CSI values collected in step 2 are filtered using 3 σ filtering methods, after filtering processing
CSI value is pre-processed, and multipath effect is eliminated；
Step 4, using CSI value pretreated in step 3, the effective height of object to be measured is estimated, by estimating out
Object to be measured effective height estimate match parameter；
Step 5, the pad value of object to be measured is obtained using match parameter, area locating for object to be measured is determined by its pad value
Domain；
Step 6, region locating for the object to be measured according to obtained in step 5 positions target.
Further, WI-FI transceiver network described in step 1 at least needs both links.
Further, pre-treatment step described in step 3 includes:
(3-1) carries out enhancing processing to the multipath fading signal in the filtered signal, and processing formula is as follows:
Wherein, n is the group number of the CSI value acquired in a period of time, and i ∈ (1, n), each group of CSI value is by 30 subcarriers
Composition, k ∈ (1,30), csimkiFor k-th of subcarrier on i-th of time point collected CSI value, ρkiFor k-th of subcarrier
Upper i-th of time point collected CSI value accounts for the specific gravity of all time point collected CSI values on the subcarrier, CSImkpIt is
Enhancing treated CSI value on k subcarrier；
Wherein, m takes 1,2,3,4 respectively, respectively indicates (2-1-1), (2-1-2), (2-1-3) and WI- in step 2
There are four kinds of situations of object to be measured in FI transceiver network；
(3-2) utilizes following formula, rejects the mistake strong signal after (3-1) is handled in CSI value:
Work as εabk< 1 and εabkWhen maximum, take k-th of subcarrier as the input of system；Reject its remaining sub-carriers；
Wherein, ab indicates the receiving-transmitting chain of signal, k ∈ (1,30), m take respectively 2,3,4 respectively indicate in step 2 the
There are three kinds of situations of object to be measured, CSI in (2-1-2), (2-1-3) and WI-FI transceiver network1kpK-th after expression enhancing processing
Initial CSI value on subcarrier, CSImkpIndicate all CSI values when having target jamming after enhancing is handled on k-th of subcarrier,
εabkThe ratio of each CSI value and initial CSI value, ab link are when indicating to have in k-th of subcarrier in ab link target jamming
Direct line of sight link of the finger information transmitting terminal to receiving end.
Further, match parameter δ described in step 41And δt, specific calculating process is as follows:
(4-1) utilizes the height and location information of the first known target, calculates ab chain road theoretical attenuation value Dab, public
Formula is as follows:
Wherein, DabFor the theoretical attenuation value of ab chain road, c (v) and s (v) are fresnel integral, and v is Fresnel-Kiel
Hough diffraction parameter, calculating formula are as follows:
Wherein, (- 1,1) t ∈, λ are signal wavelength, and h1 is the height of the first known target, da2For the first known target away from
Transmitting terminal distance, db2It is the first known target away from receiving end distance, ab link refers to information transmitting terminal to the direct view of receiving end
Away from link；
(4-2) passes through the D that (4-1) is obtainedab, match parameter δ is calculated using following formula1:
Wherein, CSI2kpFor the CSI value of the first known target after pretreatment, CSI1kpFor by pretreated initial
CSI value, DabFor the theoretical attenuation value of ab link, δ1For match parameter；
H1 and δ known to (4-3)1, following ratio relation is obtained according to formula (1):
Wherein, h1 is the height of the first known target, and ω is ratio parameter；
(4-4) is when, there are when object to be measured, the method for utilization index weighted moving average (EWMA) is pre- in WI-FI network
Estimate the height of object to be measured, specific formula is as follows:
Wherein, factor alpha indicates the variation of weight, and α ∈ (0,1), t are that CSI value acquires moment, StIt is estimated for t moment to be measured
The height value of target, h1, h2 are respectively the height of the first known target and the height of the second known target；
(4-5) ratio parameter ω according to obtained in the height for the object to be measured estimated in (4-4) and (4-3) is utilized
Following formula obtains the match parameter of different height target:
Wherein, δtFor the match parameter for estimating different height target.
Further, the specific steps of step 5 include:
(5-1) sets the variances sigma of initial CSI value after pretreatment1As threshold value, by be measured obtained in step 4
The height and match parameter of target, the D being calculated using formula (1)abIf DabGreater than σ1When, indicate that the object to be measured is in
Detection zone；Otherwise, it means that the object to be measured is in outside detection zone；
(5-2) is handled as follows if object to be measured is in detection zone:
Wherein, (1, t1) f ∈, t1 are the number of ab chain road object to be measured, DabfIndicate f-th of ab chain road to
Survey the theoretical attenuation value of target, min (Dab) indicate the 8th chain road object to be measured in theoretical attenuation value minimum value.
Further, the process of positioning target includes: in step 6
(6-1) is if the range information d of object to be measured is calculated using following formula in the region MTA for object to be measureda4、 db4, then
According to the deployment scenario of transmitting terminal receiving end, target position information is obtained, calculation formula is as follows:
Wherein, v ' is Fresnel-Kirchhoff diffraction parameter, Dabf1For the reality of f-th of the object to be measured in ab chain road
Pad value, c (v) and s (v) are fresnel integral, calculating formula are as follows:
Wherein, h is to estimate height S in step (4-4)t, λ is signal wavelength, da4It is object to be measured away from transmitting terminal distance, db4
It is object to be measured away from receiving end distance；
(6-2) positions object to be measured using the emphasis algorithm of influence area in the area OMTA when target is at the region OMTA
Position in domain.
Further, the emphasis algorithm of influence area described in step (6-2) specifically:
There are L links in (6-2-1) setting WI-FI transceiver network, and region can be surrounded by determining in L link
Chain travel permit number n1, calculate the pad value D of each of the linksq, then its weight are as follows:
Wherein, q ∈ (1, n1), DqFor the pad value of the q articles link, εqFor the q articles link pad value in n1In link
Shared weight；
(6-2-2) is weighted according to the midpoint of each of the links, is shown below, and the position of OMTA zone location is obtained
Confidence breath:
Wherein, xobj、yobjFor the position coordinates of object to be measured, xq、yqFor the q articles link midpoint coordinate.
Compared with prior art, the present invention has following technical effect:
1. the present invention, which compared to fingerprint positioning method, solves, expends asking for time and manpower during its fingerprint collecting
Topic, and robustness is good, it is practical；
2. the present invention solves existing non-fingerprint positioning method positioning accuracy by extracting the CSI information in WI-FI signal
Low problem.
Detailed description of the invention
Fig. 1 is the non-fingerprint location system flow chart of CSI；
Fig. 2 is the non-fingerprint location system hardware schematic diagram of CSI；
Fig. 3 is CSI positioning device schematic diagram, wherein (a) is industrial control mainboard schematic diagram, it (b) is transmission antenna schematic diagram；
Fig. 4 is the non-fingerprint location actual deployment figure of CSI, wherein (a) is floor map, it (b) is realistic picture；
Fig. 5 is the non-fingerprint location statistical results chart of CSI；
Fig. 6 is the non-fingerprint experiment positioning accuracy figure under distinct device quantity of CSI；
Fig. 7 is the non-fingerprint location experiment omission factor figure of CSI.
Specific embodiment
The present invention is described in detail below with reference to the accompanying drawings and embodiments.
The present embodiment describes a kind of non-fingerprint passive type localization method based on WI-FI signal comprising following steps:
Step 1, according to indoor situations, WI-FI transceiver network, including transmitting terminal a and receiving end b are constructed, and the WI-FI is received
Hairnet network includes at least both links；In order to ensure experiment effect, which needs information transmitting apparatus not conflict with each other, and
And it tries not and other WI-FI equipment clash.
Step 2, in the WI-FI transceiver network of building, the CSI value in WI-FI signal, the CSI value of the acquisition are acquired
It is divided into test CSI value and practical CSI value, wherein test CSI value are as follows:
There is no initial CSI value when target in (2-1-1) WI-FI transceiver network；
(2-1-2) is put into the CSI value when target of the first known altitude and position in WI-FI transceiver network；
(2-1-3) is put into the CSI value when target of the second known altitude and position in WI-FI transceiver network, wherein
The height of two known targets is different from the height of the first known target；
Practical CSI value refers to that, when object to be measured enters any position of WI-FI transceiver network, pause 10 seconds is collected
CSI value, and this method also shows how to position to the object to be measured emphatically；
Above-mentioned collected all data are all sent into data processor.
Step 3, all CSI values collected in step 2 are filtered using 3 σ filtering methods, after filtering processing
CSI value is pre-processed, and multipath effect is eliminated；
Aiming at the problem that collected data probably wave distortion occur, this programme is made abnormal using 3 σ filtering methods
Become waveform to try not to occur in the next steps, wherein 3 σ filtering methods specifically:
The data received are in rayleigh distributed, but for each subcarrier, value be theoretically it is certain, be exactly
Say the subcarrier data being an actually-received be theoretical value and noise and.And in the case where environment is more stable, multipath value
Can it is more stable, in this way, receive the result is that the sum of a stationary value and noise.As a result, meeting Gaussian Profile.Cause
Here select the 3 σ filtering methods generally used in engineering to be filtered waveform.
Since during receiving signal, interior has many reflections, it may appear that the multipath effect of transmission, the effect
It is divided into multipath enhancing and multipath fading.This programme is existing to enhance multipath signal, then enhancing signal is rejected, specific method
Are as follows:
(3-1) carries out enhancing processing to the multipath fading signal in the filtered signal, and processing formula is as follows:
Wherein, n is the group number of the CSI value acquired in a period of time, and i ∈ (1, n), each group of CSI value is by 30 subcarriers
Composition, k ∈ (1,30), csimkiK-th of subcarrier collected CSI value, ρ on i-th of time pointkiFor k-th of subcarrier
Upper i-th of time point collected CSI value accounts for the specific gravity of all time point collected CSI values on the subcarrier, CSImkpIt is
Enhancing treated CSI value on k subcarrier；
Wherein, m takes 1,2,3,4 respectively, respectively indicates (2-1-1), (2-1-2), (2-1-3) and WI- in step 2
There are four kinds of situations of object to be measured in FI transceiver network；
(3-2) utilizes following formula, rejects the mistake strong signal after (3-1) is handled in CSI value:
Work as εabk< 1 and εabkWhen maximum, take k-th of subcarrier as the input of system；Reject its remaining sub-carriers；
Wherein, ab indicates the receiving-transmitting chain of signal, k ∈ (1,30), m take respectively 2,3,4 respectively indicate in step 2 the
There are three kinds of situations of object to be measured, CSI in (2-1-2), (2-1-3) and WI-FI transceiver network1kpK-th after expression enhancing processing
Initial CSI value on subcarrier, CSImkpIndicate all CSI values when having target jamming after enhancing is handled on k-th of subcarrier,
εabkThe ratio of each CSI value and initial CSI value, ab link are when indicating to have in k-th of subcarrier in ab link target jamming
Direct line of sight link of the finger information transmitting terminal to receiving end.
Step 4, using CSI value pretreated in step 3, the effective height of object to be measured is estimated, by estimating out
Object to be measured effective height estimate match parameter δ1And δt；
Since tooth shape model is the model based on RSS information, and used herein is CSI information, here according to first
Position and the target effective elevation information for knowing target, determine match parameter δ1, pass through match parameter δ1Obtain match parameter with it is to be measured
The ratio relation of object height is estimated object to be measured height with EWMA method, is finally obtained corresponding to different height object to be measured
Match parameter, specific calculating process is as follows:
(4-1) utilizes the height and location information of the first known target, calculates ab chain road theoretical attenuation value Dab, public
Formula is as follows:
Wherein, DabFor the theoretical attenuation value of ab chain road, c (v) and s (v) are fresnel integral, and v is Fresnel-Kiel
Hough diffraction parameter, calculating formula are as follows:
Wherein, (- 1,1) t ∈, λ are signal wavelength, and h1 is the height of the first known target, da2For the first known target away from
Transmitting terminal distance, db2It is the first known target away from receiving end distance, ab link refers to transmitting terminal to the direct view between receiving end
Away from link；
(4-2) passes through the D that (4-1) is obtainedab, match parameter δ is calculated using following formula1:
Wherein, CSI2kpFor the CSI value under the single height target conditions after pretreatment, CSI1kpFor after pretreatment
Initial CSI value, DabFor the theoretical attenuation value of ab link, δ1For match parameter；
H1 and δ known to (4-3)1, following ratio relation is obtained according to formula (1):
Wherein, h1 is the height of the first known target, and ω is ratio parameter；
(4-4) is when, there are when object to be measured, the method for utilization index weighted moving average (EWMA) is pre- in WI-FI network
Estimate the height of object to be measured, specific formula is as follows:
Wherein, factor alpha indicates the variation of weight, and α ∈ (0,1), t are that CSI value acquires moment, StIt is estimated for t moment to be measured
The height value of target, h1, h2 are respectively the height of the first known target and the height of the second known target；
It is distinguished to reach, the height of the first known target and the second known target is preferably according to experiment area here
Altitudes are selected, and such as local average height is 1.7m, and height variance is 0.2m, then select a height in 1.7m or so
Target and a height tested in the people of 1.9m or so height；
(4-5) can be to be measured below by being derived by later height h value after carrying out known acquisition twice
H value when target position calculates, and the h value by deriving derives new δtValue makes theoretical attenuation and actually declining at that time
Subtract matching.
According to ratio parameter ω obtained in the height for the object to be measured estimated in (4-4) and (4-3), following public affairs are utilized
Formula obtains the match parameter of different height target:
Wherein, δtFor the match parameter for estimating different height target.
There is a certain error and cumulative errors for this estimation method, it is therefore proposed that carrying out after carrying out 10 times or so experiments
Error concealment, to reach positioning accuracy, specific removing method is exactly to carry out initialization height acquisition again, can directly be picked in this way
Except error.
Step 5, the pad value of object to be measured is obtained using match parameter, area locating for object to be measured is determined by its pad value
Domain；
Before positioning, can not intuitively it provide according to the collected data with the presence or absence of object to be measured, it is therefore desirable to before positioning
First judge whether there is object to be measured, judgment method are as follows:
(5-1) sets the variances sigma of initial CSI value after pretreatment1As threshold value, by be measured obtained in step 4
The height and match parameter of target, the D being calculated using formula (1)abIf DabGreater than σ1When, indicate that the object to be measured is in
Detection zone；Otherwise, it means that the object to be measured is in outside detection zone；
It is hypothetical due to using tooth shape model: when the 55% of object to be measured in first Fresnel zone just to other
Region it is effective, i.e., above method could be used when the half of object to be measured will be in first Fresnel zone.Therefore it needs here
When provide can be used, and be handled as follows:
Wherein, (1, t1) f ∈, t1 are the number of ab chain road object to be measured, DabfIndicate f-th of ab chain road to
Survey the theoretical attenuation value of target, min (Dab) indicate the 8th chain road object to be measured in theoretical attenuation value minimum value；
Wherein: if model can be used in object to be measured positioning, model Free Region is defined as model can localization region
(Model Touchable Area, MTA) is model unusable area (Out if object to be measured positioning cannot use model
Of Model Touchable Area, OMTA).
Step 6, region locating for the object to be measured according to obtained in step 5 positions object to be measured, wherein positioning object to be measured
Process include:
(6-1) is if the range information d of target is calculated using following formula in the region MTA for object to be measureda4、db4, further according to
The deployment scenario of transmitting terminal receiving end obtains target position information, and calculation formula is as follows:
Wherein, v ' is Fresnel-Kirchhoff diffraction parameter, Dabf1For the reality of f-th of the object to be measured in ab chain road
Pad value, c (v) and s (v) are fresnel integral, calculating formula are as follows:
Wherein, h is to estimate height S in step (4-4)t, λ is signal wavelength, da4It is object to be measured away from transmitting terminal distance, db4
It is object to be measured away from receiving end distance, DabfFor the theoretical attenuation value of f-th of the object to be measured in ab chain road, Dabf1For ab chain
The actual attenuation value of f-th of object to be measured on the road；
By the above calculation method it is found that the core concept of this method is the actual attenuation value in known link, pass through matching
Parameter δtIt can make theoretical attenuation value DabfWith actual attenuation value Dabf1It is approximately equal, match parameter δtExist with object to be measured height special
Fixed ratio relation, if capableing of the height of known object to be measured, then can be obtained by the ginseng of matching corresponding to the object to be measured
Number, this method use known known target height and position twice, estimate the method for the height of subsequent object to be measured to obtain
Match parameter, to position object to be measured.
In order to verify the feasibility of this method, applicant has also carried out confirmatory experiment, and applicant is put into WI-FI network
One height is known, Location-Unknown object to be measured, and the position of the object to be measured is positioned by this method, verifies above-mentioned algorithm
Accuracy.
In the present embodiment, using the WI-FI signal of 2.4GHz, therefore λ is 0.125m, can release d according to above-mentioned formulaaWith
dbAnd and product, when there are two with uplink, available specific daAnd db, in the deployment scenario according to sending and receiving end,
Obtain the position of object to be measured；
(6-2) positions object to be measured using the emphasis algorithm of influence area in the area OMTA when target is at the region OMTA
Position in domain；
The wherein emphasis algorithm of influence area specifically:
There are L links in (6-2-1) setting WI-FI transceiver network, and region can be surrounded by determining in L link
Chain travel permit number n1, calculate the pad value D of each of the linksq, then its weight are as follows:
Wherein, q ∈ (1, n1), DqFor the pad value of the q articles link, εqFor the q articles link pad value in n1In link
Shared weight；
(6-2-2) is weighted according to the midpoint of each of the links, is shown below, and the position of OMTA zone location is obtained
Confidence breath:
Wherein, xobj、yobjFor the position coordinates of object to be measured, xq、yqFor the q articles link midpoint coordinate.
The present embodiment additionally provides experimental verification:
This sample plot point is selected in eight buildings laboratories of Northwest University's Information Institute, and Experimental Area size is 6*12m, such as Fig. 3 institute
Show, subscriber station five-pointed star point in figure carries out positioning experiment.In order to facilitate experiment, the transmitting and receiving device selected here is all
Carry the machine of Intel5300AGN wireless network card, operating system Ubuntu10.04.4, may be used as signal send and
It receives, such as Fig. 3, data processor is the computer of windowsOS.When signal is sent, wlan device is only simulated, is not done
Other changes.And in order to guarantee to cover, only choose one group of transmission every time here, remaining is respectively as receiving device.This experiment
The case where acquiring 88 positions altogether.The position for carrying out distinct methods determines.
Simultaneously here in order to which the effect for embodying this method and existing method are compared.
For RASS system, RTI system and Alico system, RSS information is used, this information is in CSI information
It can also obtain, and calculating can be extracted.RASS system is that equipment is divided into delta-shaped region, is analyzed, in experiment
The CSI transmitter-receiver of 3 antennas is modeled to 3 sending and receiving devices, multiple delta-shaped regions can be formed in this way, and in this way
What the software of design also can be achieved on.RTI system is divided on link influence, and the variation of RSS information is used in experiment
Obtain as a result, and since RTI system is to obtain result according to the figure for influencing link, due to there was only the portion on both sides in experiment
Administration, it is assumed that the midpoint for influencing link is final goal point.And Alico system is and the immediate system of article, this inner part
Administration and its deployment way are also about the same, can not excessively dense deployment only because plant issue, it may appear that do not examine in article
The region outside first Fresnel zone considered, and here for convenience of calculation and effectively comparison, all equipment positions are known here
It sets, not to the estimation of position.Pilot system is progress Characteristic Contrast positioning after being learnt the information of CSI, is characterized in son
The cross-correlation function of carrier wave.And CWF system is solved using CSI information, carries out data acquisition by the way of poll here, is made
There are links as much as possible under the conditions of limited deployment facility for system, and determine each 20 data of link acquisition, complete
After replace transmitting terminal.
Experimental result is as shown in figure 4, the statistical result of distinct device quantity is as shown in Figure 5.
RTI system effect is worst as seen from the figure, and middle position error reaches 3.5m, and RASS system is secondly, middle position error is
The effect of 3m, Alico system be it is relatively good in RSS information, middle position error reaches 2m, Pilot system in curve,
Precision highest, but middle position error ratio CWF system is lower.Pilot system is 1.4m, and CWF system is 1.3m.And according to preceding
The narration in face, the positioning of CWF system is divided into the region MTA and OMTA, and the region OMTA is estimated with midpoint method, in this way
Error it is very big, meanwhile, some point belonging to region only have 1 group of link that can cover, will cause in this way accuracy rate decline.Therefore
In CDF curve, some effect can be deteriorated.But its result is much better than the positioning using RSS information fingerprint, also than using
The Alico locating effect of similar approach will be got well, therefore the positioning accuracy of the system is in higher level.
This illustrate indoors no matter multipath how much, CWF system proposed in this paper has relatively good precision, though this method
So than the Pilot system difference based on finger print information, but due to the fingerprint collecting that it does not need Pilot method early period, saving
A large amount of time.
From fig. 5, it can be seen that is compared here is the result of middle position error.By comparison it can be found that distinct methods
Position error is different, and error reduces with the increase of transmission device quantity, this is because with the increase of number of devices,
Overlay area is increasing, and the locating effect of the above several method is all related with overlay area.Under any deployment density,
RTI system is worst, this is because RTI system when proposing, is to need to enclose equipment around region one to dispose, only deploys here
Both sides, and there are also 8 points outside region, and in experiment, the mode that RTI system is not directed to two rows of deployment is improved, as a result
It can be far short of what is expected.RASS system effect is slightly good, but still has gap compared to other several methods, and mainly its dispositions method is small by three
It is angular, although triangle can also be formed here, the hexagon that cannot be formed as the equilateral triangle in text and especially.Remaining three
Kind method effect has similar effect in some points, is since on certain points, three kinds of methods can be obtained more accurately
As a result.Alico system is unstable due to multipath and RSS information, and effect is general, but due to there is certain processing, than using RSS
The method of information fingerprint detection will be got well.And for Pilot system, effect is generally best, reason and upper experiment phase
Together, due to acquiring fingerprint, without the influence for considering multipath.But it is also due to acquisition fingerprint, very labor intensive.Finally, right
In CWF system proposed in this paper, effect can be poorer than Pilot under low-density deployment, when being disposed this is mainly due to low-density,
CWF system has many positions to be in the region OMTA, will cause error increase in this way.And Pilot system is finger print information, it is this
The effective position region of system is bigger than the CSI ellipse localization region generally drawn, therefore its effect ratio in low-density deployment
CWF system is good.
Here it defines: being defined as False Negative (FN) in Computer Subject:
POM=∑ target is in region but is judged as the probability not in region
As a result as shown in fig. 6, discovery is with the reduction of coverage area, the result of CWF system, which does not occur, to be sharply increased,
The reason is that the positioning in its region OMTA, in an experiment, many points near deployment region do not meet positioning requirements, but
There are certain variations, therefore can not detected in Pilot and can detecte out in CWF.
Under maximum coverage rate, the omission factor ratio Pilot high of CWF is mainly used due to CWF in positioning here
Boundary be with object variations, and Pilot is by set in advance.The localization method of Pilot can detecte more
Point, but positioning accuracy can also decline simultaneously.
After it compared existing method, it can be deduced that the available relatively high positioning accuracy of system.
Claims (6)
1. a kind of non-fingerprint passive type localization method based on WI-FI signal, which comprises the following steps:
Step 1, WI-FI transceiver network, including transmitting terminal a and receiving end b are constructed；
Step 2, in the WI-FI transceiver network of building, the CSI value in WI-FI signal is acquired, the CSI value is divided into test
CSI value and practical CSI value, wherein test CSI value are as follows:
There is no initial CSI value when target in (2-1-1) WI-FI transceiver network；
(2-1-2) is put into the CSI value when target of the first known altitude and position in WI-FI transceiver network；
(2-1-3) is put into the CSI value when target of the second known altitude and position in WI-FI transceiver network, wherein second
Know that the height of target is different from the height of the first known target；
Practical CSI value refers to: when there are CSI values collected when object to be measured in WI-FI transceiver network；
Step 3, all CSI values collected in step 2 are filtered using 3 σ filtering methods, to the CSI after filtering processing
Value is pre-processed, and multipath effect is eliminated；The pre-treatment step includes:
(3-1) carries out enhancing processing to the multipath fading signal in the filtered signal, and processing formula is as follows:
Wherein, group number of the n for the CSI value of acquisition in a period of time, i ∈ (1, n), each group of CSI value are made of 30 sub- carrier waves,
K ∈ (1,30), csimkiIndicate k-th of subcarrier collected CSI value, ρ on i-th of time pointkiFor on k-th of subcarrier
I-th of time point collected CSI value accounts for the specific gravity of all time point collected CSI values on the subcarrier, CSImkpFor kth
Enhancing treated CSI value on a subcarrier；
Wherein, m takes 1,2,3,4 respectively, respectively indicates (2-1-1), (2-1-2), (2-1-3) and WI-FI transmitting-receiving in step 2
There are four kinds of situations of object to be measured in network；
(3-2) utilizes following formula, rejects the mistake strong signal after (3-1) is handled in CSI value:
Work as εabk< 1 and εabkWhen maximum, take k-th of subcarrier as the input of system；Reject its remaining sub-carriers；
Wherein, ab indicates that the receiving-transmitting chain of signal, k ∈ (1,30), m take 2,3,4 to respectively indicate (2-1- in step 2 respectively
2), there are three kinds of situations of object to be measured, CSI in (2-1-3) and WI-FI transceiver network1kpK-th of son carries after indicating enhancing processing
Initial CSI value on wave, CSImkpIndicate all CSI values when having target jamming after enhancing is handled on k-th of subcarrier, εabkTable
Show that the ratio of each CSI value and initial CSI value when having target jamming in ab link in k-th of subcarrier, ab link refer to letter
Direct line of sight link of the breath transmitting terminal to receiving end；
Step 4, using CSI value pretreated in step 3, estimate the effective height of object to be measured, by estimate out to
The effective height for surveying target estimates match parameter；
Step 5, the pad value of object to be measured is obtained using match parameter, region locating for object to be measured is determined by its pad value；
Step 6, region locating for the object to be measured according to obtained in step 5 positions target.
2. the non-fingerprint passive type localization method based on WI-FI signal as described in claim 1, which is characterized in that in step 1
The WI-FI transceiver network at least needs both links.
3. the non-fingerprint passive type localization method based on WI-FI signal as described in claim 1, which is characterized in that in step 4
The specific calculating process of the match parameter is as follows:
(4-1) utilizes the height and location information of the first known target, calculates ab chain road theoretical attenuation value Dab, formula is such as
Under:
Wherein, DabFor the theoretical attenuation value of ab chain road, c (v) and s (v) are fresnel integral, and v is Fresnel-Kirchhoff
Diffraction parameter, calculating formula are as follows:
Wherein, (- 1,1) t ∈, λ are signal wavelength, and h1 is the height of the first known target, da2It is the first known target away from transmission
Hold distance, db2It is the first known target away from receiving end distance, ab link refers to information transmitting terminal to the direct sighting distance chain of receiving end
Road；
(4-2) passes through the D that (4-1) is obtainedab, match parameter δ is calculated using following formula1:
Wherein, CSI2kpFor the CSI value of the first known target after pretreatment, CSI1kpTo pass through pretreated initial CSI
Value, DabFor the theoretical attenuation value of ab link, δ1For match parameter；
H1 and δ known to (4-3)1, following ratio relation is obtained according to formula (1):
Wherein, h1 is the height of the first known target, and ω is ratio parameter；
(4-4) is when, there are when object to be measured, the method for utilization index weighted moving average (EWMA) is estimated to be measured in WI-FI network
The height of target, specific formula is as follows:
Wherein, factor alpha indicates the variation of weight, and α ∈ (0,1), t are that CSI value acquires moment, StObject to be measured is estimated for t moment
Height value, h1, h2 are respectively the height of the first known target and the height of the second known target；
(4-5) ratio parameter ω according to obtained in the height for the object to be measured estimated in (4-4) and (4-3), using as follows
Formula obtains the match parameter δ of different height targett:
Wherein, h is the height S for the object to be measured estimated in step (4-4)t, δtFor the match parameter for estimating different height target.
4. the non-fingerprint passive type localization method based on WI-FI signal as claimed in claim 3, which is characterized in that step 5
Specific steps include:
(5-1) sets the variances sigma of initial CSI value after pretreatment1As threshold value, pass through object to be measured obtained in step 4
Height and match parameter, are calculated D using formula (1)abIf DabGreater than σ1When, indicate that the object to be measured is in detection zone；
Otherwise, it means that the object to be measured is in outside detection zone；
(5-2) is handled as follows if object to be measured is in detection zone:
Wherein, (1, t1) f ∈, t1 are the number of ab chain road object to be measured, DabfIndicate f-th of the mesh to be measured in ab chain road
Target theoretical attenuation value, min (Dab) indicate ab chain road object to be measured in theoretical attenuation value minimum value.
5. the non-fingerprint passive type localization method based on WI-FI signal as claimed in claim 4, which is characterized in that in step 6
Positioning target process include:
(6-1) is if the range information d of object to be measured is calculated using following formula in the region MTA for object to be measureda4、db4, further according to hair
The deployment scenario of sending end receiving end obtains target position information, and calculation formula is as follows:
Wherein, v ' is Fresnel-Kirchhoff diffraction parameter, Dabf1For the actual attenuation of f-th of the object to be measured in ab chain road
Value, c (v) and s (v) are fresnel integral, calculating formula are as follows:
Wherein, h is to estimate height S in step (4-4)t, λ is signal wavelength, da4It is object to be measured away from transmitting terminal distance, db4For to
Target is surveyed away from receiving end distance；
(6-2) positions object to be measured using the emphasis algorithm of influence area in the region OMTA when target is at the region OMTA
Position.
6. the non-fingerprint passive type localization method based on WI-FI signal as claimed in claim 5, which is characterized in that step (6-
2) the emphasis algorithm of the influence area described in specifically:
There are L link in (6-2-1) setting WI-FI transceiver network, the chain that region can be surrounded in L link is determined
Travel permit number n1, calculate the pad value D of each of the linksq, then its weight are as follows:
Wherein, q ∈ (1, n1), DqFor the pad value of the q articles link, εqFor the q articles link pad value in n1It is shared in link
Weight；
(6-2-2) is weighted according to the midpoint of each of the links, is shown below, and the position letter of OMTA zone location is obtained
Breath:
Wherein, xobj、yobjFor the position coordinates of object to be measured, xq、yqFor the q articles link midpoint coordinate.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610575118.6A CN106211319B (en) | 2016-07-20 | 2016-07-20 | A kind of non-fingerprint passive type localization method based on WI-FI signal |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610575118.6A CN106211319B (en) | 2016-07-20 | 2016-07-20 | A kind of non-fingerprint passive type localization method based on WI-FI signal |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106211319A CN106211319A (en) | 2016-12-07 |
| CN106211319B true CN106211319B (en) | 2019-10-01 |
Family
ID=57490945
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610575118.6A Active CN106211319B (en) | 2016-07-20 | 2016-07-20 | A kind of non-fingerprint passive type localization method based on WI-FI signal |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106211319B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109059895B (en) * | 2018-03-28 | 2021-05-25 | 南京航空航天大学 | Multi-mode indoor distance measurement and positioning method based on mobile phone camera and sensor |
| CN109316192B (en) * | 2018-09-13 | 2020-07-17 | 清华大学 | Sitting posture detection method and device based on mobile network channel state information |
| CN109617591B (en) * | 2019-01-21 | 2021-11-23 | 重庆邮电大学 | WiFi-based link-level moving target tracking method |
| CN110363947B (en) * | 2019-07-11 | 2021-11-23 | 京东方科技集团股份有限公司 | Method for judging illegal intrusion based on WIFI (Wireless Fidelity) reflected signal and security system |
| CN111770436B (en) * | 2020-05-07 | 2023-05-16 | 普联国际有限公司 | Indoor Wi-Fi positioning method and device based on CSI and storage medium |
| CN112738749B (en) * | 2020-12-25 | 2021-10-01 | 武汉理工大学 | Multi-target indoor positioning method for shipborne environment based on Fresnel zone |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20070079705A (en) * | 2006-02-03 | 2007-08-08 | 삼성전자주식회사 | Method and apparatus for informing csi ue which can communicate with voip ue whether csi ue can use cs service or not |
| KR20130012207A (en) * | 2011-07-01 | 2013-02-01 | 주식회사 케이티 | Web messaging system and method based on context awareness for providing social commerce service broadening marketing campaign channel |
-
2016
- 2016-07-20 CN CN201610575118.6A patent/CN106211319B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20070079705A (en) * | 2006-02-03 | 2007-08-08 | 삼성전자주식회사 | Method and apparatus for informing csi ue which can communicate with voip ue whether csi ue can use cs service or not |
| KR20130012207A (en) * | 2011-07-01 | 2013-02-01 | 주식회사 케이티 | Web messaging system and method based on context awareness for providing social commerce service broadening marketing campaign channel |
Non-Patent Citations (1)
| Title |
|---|
| Pilot: Passive Device-free Indoor Localization Using Channel State Information;Jiang Xiao 等;《2013 IEEE 33rd International Conference on Distributed Computing Systems》;20131231;第2-6部分 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106211319A (en) | 2016-12-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106211319B (en) | A kind of non-fingerprint passive type localization method based on WI-FI signal | |
| Gu et al. | Paws: Passive human activity recognition based on wifi ambient signals | |
| Wang et al. | A novel weighted KNN algorithm based on RSS similarity and position distance for Wi-Fi fingerprint positioning | |
| Wang et al. | Toward accurate device-free wireless localization with a saddle surface model | |
| Bocca et al. | Multiple target tracking with RF sensor networks | |
| CN103945332B (en) | A kind of received signal strength and multi-path information united NNs indoor orientation method | |
| CN103713277B (en) | A kind of radiation source localization method of position-based information field | |
| CN104883732B (en) | A kind of enhanced indoor passive passive human body localization method | |
| CN109672485B (en) | Indoor personnel real-time invasion and movement speed detection method based on channel state information | |
| CN112346050B (en) | Fall detection method and system based on Wi-Fi equipment | |
| CN108038419B (en) | Wi-Fi-based indoor personnel passive detection method | |
| CN108650628A (en) | The indoor orientation method that ranging and fingerprint based on Wi-Fi network are combined | |
| Miao et al. | Smart attendance system based on frequency distribution algorithm with passive RFID tags | |
| CN106877947A (en) | A kind of radio-frequency channel parallel detection device and method of unmanned plane | |
| CN109587645A (en) | Personnel's recognition methods under indoor environment based on channel state information | |
| CN106597365B (en) | A kind of sophisticated electronic signal time difference positioning method based on time domain cluster | |
| CN103197302B (en) | Target location extraction method applicable to through-the-wall radar imaging | |
| CN109709512A (en) | A kind of single station of unmanned plane detecting and unmanned plane detecting system | |
| CN107182036A (en) | The adaptive location fingerprint positioning method merged based on multidimensional characteristic | |
| CN110543842A (en) | Target motion identification method and system based on Wi-Fi signals | |
| CN107037422A (en) | A kind of passive type localization method towards multiple application | |
| CN109655790A (en) | Multi-target detection and identification system and method based on indoor LED light source | |
| CN112040400A (en) | Single-site indoor fingerprint positioning method based on MIMO-CSI, storage medium and equipment | |
| CN103945526A (en) | Wireless equipment positioning method and system based on induced detection technology | |
| Hao et al. | CSI‐HC: A WiFi‐Based Indoor Complex Human Motion Recognition Method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |