CN105929388A - Novel indoor positioning method based on WiFi network - Google Patents

Abstract

The present invention discloses a novel indoor positioning method based on a WiFi network. A transmitting antenna irradiates a positioning area where a target is located to form an incident filed, mutual interaction with the target is carried out to form a scattering field, and the scattering echo signal of the target is received. The antenna works in a broadband sweep mode, multiple random frequency point values are divided, the positioning area is subjected to average dispersion to be multiple small areas, and the geometric center of each small area is taken as a discrete point. The linear model of microwave coherent imaging is established, a compressed sensing calculating method is used to carry out matrix inversion on the linear model of microwave coherent imaging, the scattering coefficient of each small area is obtained, the small area where the target is located is obtained as a target area, and a final discrete point is taken as the position of the target. According to an existing indoor WiFi antenna and microwave coherent imaging method, a novel indoor positioning mode is obtained, and the method has the advantages of fast and accurate positioning, simple system deployment and high applicability.

Description

A kind of new indoor localization method based on Wi-Fi network

Technical field

The present invention relates to a kind of space target positioning method, especially related to a kind of based on Wi-Fi network New indoor localization method.

Background technology

Novel microwave coherent imaging technology is that single reception antenna obtains thing in the way of Non-scanning mode from target scattering field Shape and the new technique of position thereof.Make owing to various media and metal object can produce scattering in electromagnetic field With, its scattered field contains the surface characteristic information of object.According to the terrible imaging theory in optics, by known Many groups random distribution electromagnetic field of characteristic is radiated on object and is received by the reception antenna that single position is fixing. After obtaining many group measurement data, scatterer can be extracted after the correlation computations with scattered field of the in-field Profile and surface information, thus only carry out magnetography location with single antenna.This novel method is not required to mobile spy Observation line also avoids complicated antenna array, reduces system cost and improves the ability of identification target simultaneously, Suffer from huge at the civil areas such as the detection of electromagnetism partition wall, medical imaging and radar fix and military field Application potential.

Compressed sensing is that a kind of new acquisition of information is theoretical, is built upon sparse signal representation, calculation matrix A kind of signals collecting on non-correlation and approximation theory and the method for reconstruction.This theory is pointed out, if letter Number it is sparse or moment compression under certain base, it is possible to by far below nyquist sampling theorem The sample rate required obtains the structural information of signal, then is completed the Accurate Reconstruction of signal by restructing algorithm.Its Basic thought is first high dimensional signal to be reduced to low bit space, obtains observation, this mistake by accidental projection Journey completes sampling and compression, the measured value then utilizing sparse prior knowledge processing to receive simultaneously, finally leads to Cross and solve convex optimization problem to rebuild primary signal.It is currently based on algorithm for reconstructing theoretical for CS and mainly has orthogonal Join algorithm (OMP), base tracing algorithm (BP), iteration threshold algorithm (LHT) etc..

Along with the drawback of wireless network is gradually overcome, radio network technique reaches its maturity with universal, based on IEEE The Wi-Fi wireless local area network technology of 802.11 standard bricks be widely used in business, live, each side such as study Face.One typical radio router using 802.11b or 802.11g and antenna, without under any barrier Coverage can reach indoor 50 square metres.IEEE 802.11b/g standard is operated in 2.4G frequency range, frequency Scope is 2.400-2.4835GHz, altogether 83.5M bandwidth frequency range will divide into 14 repetitions, the frequency of labelling Road；The mid frequency of each channel differs 5 megahertzs.

Summary of the invention

In order to cover Wi-Fi network the interior space carry out target location, it is an object of the invention to provide A kind of new indoor localization method based on Wi-Fi network, launches antenna and works in broadband frequency sweep with reception antenna Pattern, single or multiple transmitting antenna radiation chamber region of interest within, individual antenna receives strong scattering target and produces Raw scattered signal, in conjunction with known target area radiation field information, uses compressed sensing algorithm to carry out target Rebuild location.

In order to achieve the above object, the technical solution used in the present invention is:

The present invention use single or multiple transmitting antenna irradiation to the region, location at target place, at positioning area Formation in-field, territory, in-field forms scattered field with objectives interation, by single reception antenna to target Scatter echo signal is received；Antenna can random installation in the interior space.

Launching antenna can be as shown in Figure 3 multiple, or as shown in Figure 2 single, and reception antenna is only There is one.

Launch antenna and work in broadband frequency sweep mode with reception antenna, broadband range is divided into m individual random Value of frequency point, m is frequency dispersion degree, region, location average discrete is become n zonule, with each zonule Geometric center as discrete point, obtain n discrete point；

As it is shown in figure 1, set up the linear of microwave coherent imaging according to electromagnetic scattering principle and Inverse Scattering Theory Model, uses compressed sensing computational methods that the linear model of microwave coherent imaging is carried out matrix inversion, it is thus achieved that The scattering coefficient of each zonule, and then obtain the zonule at target place, and as target area, finally Obtain the discrete point position as target of target area.

In the inventive method, target is the sparse target in a spatial domain for region, location, uses compression Perception algorithm in the case of lack sampling to signal Fast Restoration.

Use compressed sensing computational methods that the linear model of microwave coherent imaging is carried out matrix inversion specifically to select Select Regularization function and solve object function, and optimization object function minima.

Described single reception antenna can also be to launch antenna.

The linear model of described microwave coherent imaging is expressed as:

$Sca\left(r_r\right)=E_{rad}^t\left(r\right)E_{rad}^r\left(r\right)\mathrm{\σ}\left(r\right)$

Wherein, Sca (r_r) for being positioned at positioning area overseas coordinate r_rThe reception antenna at place receives the scatter echo letter obtained Number；WithIt is expressed as the radiation field distribution launching antenna and reception antenna at location region r； σ (r) is the backscattering coefficient in the region r of location；

Described employing compressed sensing computational methods carry out matrix inversion to the linear model of microwave coherent imaging, tool Body process is: the linear model of described microwave coherent imaging is carried out on frequency domain and spatial domain discretization respectively Process, obtain matrix equation:

$\left[\begin{array}{c}Sca{\left(r_r\right)}_1\\ Sca{\left(r_r\right)}_2\\ C\\ Sca{\left(r_r\right)}_m\end{array}\right]=\left[\begin{array}{cccc}{E}_{rad}^t{\left(r\right)}_{11}& E_{rad}^t{\left(r\right)}_{12}& B& E_{rad}^t{\left(r\right)}_{1n}\\ E_{rad}^t{\left(r\right)}_{21}& E_{rad}^t{\left(r\right)}_{22}& & \\ C& & E& \\ E_{rad}^t{\left(r\right)}_{m1}& & & E_{rad}^t{\left(r\right)}_{mn}\end{array}\right]\left[\begin{array}{cccc}{E}_{rad}^t{\left(r\right)}_{11}& E_{rad}^t{\left(r\right)}_{12}& B& E_{rad}^t{\left(r\right)}_{1n}\\ E_{rad}^t{\left(r\right)}_{21}& E_{rad}^t{\left(r\right)}_{22}& & \\ C& & E& \\ E_{rad}^t{\left(r\right)}_{m1}& & & E_{rad}^t{\left(r\right)}_{mn}\end{array}\right]\left[\begin{array}{c}\mathrm{\σ}{\left(r\right)}_1\\ \mathrm{\σ}{\left(r\right)}_2\\ C\\ \mathrm{\σ}{\left(r\right)}_n\end{array}\right]$

Wherein, subscript m represents the electromagnetic field information of m-th Frequency point, and subscript n represents the n-th positioning area Electromagnetic field information at discrete point in territory.

Above-mentioned matrix equation is reduced to:

G=Hf

Wherein,Represent and launch antenna and the reception antenna radiation field information in target area, g For scatter echo signal, f is the backscattering coefficient variable of target.

Wherein, g is the calculation matrix of m × 1, the scattered signal that i.e. m frequency dispersion point is corresponding.H is one The calculation matrix of individual m × n, the wherein measured value of corresponding all of m the different frequent points of row vector.The f of n × 1 Vector represents certain characteristic information in the region, location being separated into n point, such as back scattering value.Radar mesh Target echo-signal can be regarded as the result of several strong scattering point echo-signal superposition, therefore original echo letter Number also have certain openness on certain transform domain.It is clear that microwave coherent imaging is to isolated target Detection location, target to be positioned is exactly sparse signal in spatial domain.

Use compressed sensing computational methods to solve the minima of object function by below equation to carry out matching and optimize Arrive, until the minima of object function minimizes value threshold value:

$F\left(f\right)=\left|\right|g-Hf|{|}_p^2+\mathrm{\λ}R\left(f\right)$

Wherein, F (f) represents the object function of the backscattering coefficient variable f of target, and p represents the exponent number of norm, | | | | for p rank norm sign, R (f) is Regularization function, λ for regulation norm and Regularization function each other Weight parameter.

It is embodied as the compressed sensing algorithm two step iteration threshold fast algorithm (TwIST) of middle a kind of Fast Convergent, But it is not limited to this.

Described Antenna Operation is in Wi-Fi operating frequency, and frequency dispersion degree m is dilute more than the target in region, location Dredge degree.The target sparse degree in region, location refers to position the quantity of the target of domestic demand location, region.

Described region, location is the indoor covering Wi-Fi signal, launches antenna and reception antenna uses Wi-Fi Signal carries out irradiation and reception, and location target is strong scattering object.

Strong scattering object refers to electromagnetic wave irradiation and produces the strongest scattered field to this object, and metal object is permissible It is considered as a kind of typical strong scattering object.

Described scatter echo signal is measured by vector network analyzer and is obtained, i.e. the amplitude of S parameter and phase Position information.

Described antenna irradiation of launching is obtained to the radiation field information positioning region formation by antenna damnification Or transform to target area by antenna damnification to obtain.

The invention has the beneficial effects as follows:

Present invention utilizes compressed sensing so that it carries out data acquisition and processing (DAP) in the case of lack sampling, Its amount of calculation is greatly reduced, and locating speed is fast and accurate.

And the present invention can be applied to be disposed with the indoor of wifi easily and position, system deployment letter Single, it is not necessary to extra dual-mode antenna is installed, it is possible to realizing accurately and quickly positioning, the suitability is strong.

Accompanying drawing explanation

Fig. 1 is the flow chart of the inventive method.

Fig. 2 is that principle schematic arranged by embodiment of the present invention dual-mode antenna.

Fig. 3 is the embodiment of the present invention applicable positioning system structure schematic diagram.

Fig. 4 is embodiment of the present invention emulation experiment schematic diagram.

Fig. 5 is back scattering sparse distribution result figure in region, embodiment of the present invention location.

Fig. 6 is the position location that the embodiment of the present invention obtains according to back scattering distribution coordinate transform.

Fig. 7 is the embodiment of the present invention used convergence of algorithm speed and operation time result.

Detailed description of the invention

Below in conjunction with the accompanying drawing in the embodiment of the present invention, describe the implementation process of the present invention in detail.

The linear model of the microwave coherent imaging of the present invention is obtained by following principle process:

Region, location is the region Q of XOY two dimensional surface,|²Represent two dimensional surface space, r ∈ Q For the vector position of any point in region, location, D represents antenna aperture.It is largely divided into three processes:

First process: launch antenna and the reception antenna formation radiation field distribution in target area.Assume to send out Penetrate antenna T at position r_tMouth scene distribution table be shown asTheoretical according to electromagnetic wave free-space propagation, can To obtain the radiation field distribution at the r of far field:

$E_{rad}^t\left(r\right)=\underset{D_t}{\∫}{E}_{ape}^t\left(r_t\right)\·g(r,r_t){\mathrm{dr}}_t$

Wherein c is the light velocity,R |=| r-r_t| for free space Green's function.In like manner can obtain position r_rThe reception antenna R at place radiation field distribution function at r:

$E_{rad}^r\left(r\right)=\underset{D_r}{\∫}{E}_{ape}^r\left(r_r\right)\·g(r,r_r){\mathrm{dr}}_t$

Second process, being formed after incident radiation field, incident radiation field is to be imaged with position in region Target interacts to form scattered field.This scattering process and the electrical quantity characteristic of target, geometry are the most relevant, For single target, the target backscattering coefficient σ of equivalence is utilized to describe the interaction of in-field and target.

For the static target in this stationary positioned region, in region, the backscattering coefficient of optional position r represents For σ (r), it is assumed that this coefficient is unrelated with frequency, t change the most in time.For the sake of Jian Dan, ignore multiple scattering, Born first approximation is utilized to obtain target back scattering expression formula:

$E_{Sca}\left(r_r\right)=E_{rad}^t\left(r\right)\·\mathrm{\σ}\left(r\right)$

Wherein, E_sca(r_r) for launching antenna at position r_rThe scattered field distribution at place.

3rd process, target scattering signal is as Secondary Emission source, and by being in position r_rThe reception sky at place Line receives, and obtains scatter echo signal:

$Sca\left(r_r\right)=\underset{D_r}{\∫}{E}_{ape}^r\·E_{sca}\left(r\right)\·g(r_r,r)dr$

In conjunction with above-mentioned radiation field expression formula, readjusting and simplifying obtains the linear model expression formula of coherent imaging:

$Sca\left(r_r\right)=E_{rad}^t\left(r\right)E_{rad}^r\left(r\right)\mathrm{\σ}\left(r\right).$

Above-mentioned relation formula illustrates the pass received between scatter echo signal and radiation field of aerial, target scattering field System, launches antenna with reception antenna in the radiation field positioning regionSystem for this imaging system passes Defeated function H, echo-signal Sca is the system response for object σ (r) to be imaged of this system.Scatter echo In contain the scattered information of location region internal object, this information is radiation field of aerial and objectives interation As a result, the integral characteristic of above formula represents that target information is coupling among reception echo, and image-forming principle is namely based on This relational expression, from receiving extraction target information scatter echo, obtains inverting value σ (r) of target scattering coefficient.Cause This, in the case of known radiation field information and scattered signal, select suitable matrix inversion technique to obtain mesh Mark scattered information σ.

Embodiments of the invention are as follows:

Embodiment disposes the indoor locating system framework of Wi-Fi antenna as shown in Figure 2.

The present invention first pass through near field measurement obtain launch antenna divide in the radiation field of target area with reception antenna Cloth, i.e. the transmitting radiation field of aerial of m × n sizeWith reception antenna radiation fieldOperating frequency Counting as m, target area discretization is counted as n, m=32, n=64.The transmitting antenna T randomly placed₁、 T₂……T_kTransmitting electromagnetic wave radiation is to indoor objects region the most successively, and the reception antenna R randomly placed depends on Secondary reception scattered signal, launches antenna T_i(i=1,2 ... k) illuminated target area, reception antenna R obtain The calculation matrix g of scattered signal m × 1.Launch and the process of reception, make g=Sca (r_r)_{I=1,2B k}, Invert according to model g=Hf and obtain desired value f.

Embodiment uses the compressed sensing algorithm two step iteration threshold fast algorithm (TwIST) of Fast Convergent, Microwave coherent imaging model is solved by TwIST algorithm as a Sparse Problems, i.e. by solving object function Minima carry out matching and obtain:

$\hat{f}=\arg {\min }_f\left(\right||g-Hf|{|}_2^2+\mathrm{\λ}\left|\right|f\left|{|}_1^2\right)$

Wherein, | | | | for norm sign, λ is regulation norm and Regularization function weight parameter each other, Regularization function R is chosen as single order norm.

It is two-dimentional fixed to have carried out the single transmitting shown in accompanying drawing 2 and reception antenna location schematic diagram as such scheme Bit emulator is tested；Operating frequency frequency dispersion point m=32, target area Q is separated into zonule, n=16 × 16, Two one one, transmitting antennas are received and are irradiated area to be targeted, and radiation field distribution schematic diagram is as shown in Figure 4.Adopt When carrying out inverting location with traditional least-square fitting approach, the inaccurate situation of positioning result occurs；And Use matrix inversion algorithm based on compressed sensing algorithm provided by the present invention, it is possible in the feelings of super lack sampling Quickly location is realized under condition.As it is shown in figure 5, use compressed sensing algorithm, matrix inversion is calculated The backscattering coefficient distribution in discrete placement region, the subscript that nonzero value is corresponding is strong scattering target place Discrete point coordinate.The physical location in region, location of target can be obtained, such as Fig. 6 by coordinate transform Shown in.Fig. 7 shows, localization method provided by the present invention has convergence rate and shorter computing faster Time, greatly reducing amount of calculation and be greatly reduced, the suitability is strong, has prominent significant technique effect.

The above, the only detailed description of the invention of the present invention, but protection scope of the present invention is not limited to This, any those familiar with the art, in the technical scope of present disclosure, can readily occur in Change or replacement, all should contain within protection scope of the present invention.Therefore, the protection model of the present invention Enclose and be as the criterion with the protection domain of claims.

Claims (9)

1. a new indoor localization method based on Wi-Fi network, it is characterised in that said method comprising the steps of:

Using single or multiple transmitting antenna irradiation to the region, location at target place, in formation in-field, region, location, in-field forms scattered field with objectives interation, single reception antenna is received the scatter echo signal of target；

Launching antenna and reception antenna and work in broadband frequency sweep mode, broadband range is divided into m random value of frequency point, m is frequency dispersion degree, and region, location average discrete is become n zonule, using the geometric center of each zonule as discrete point；

Set up the linear model of microwave coherent imaging, use compressed sensing computational methods that the linear model of microwave coherent imaging is carried out matrix inversion, obtain the scattering coefficient of each zonule, and then obtain the zonule at target place, and as target area, the final discrete point position as target obtaining target area.

A kind of new indoor localization method based on Wi-Fi network the most according to claim 1, it is characterised in that:

Using compressed sensing computational methods that the linear model of microwave coherent imaging carries out matrix inversion specifically selects Regularization function to solve object function, and optimization object function minima.

A kind of new indoor localization method based on Wi-Fi network the most according to claim 1, it is characterised in that: described single reception antenna can also be to launch antenna.

A kind of new indoor localization method based on Wi-Fi network the most according to claim 1, it is characterised in that: the linear model of described microwave coherent imaging is expressed as:

Wherein, Sca (r_r) for being positioned at positioning area overseas coordinate r_rThe reception antenna at place receives the scatter echo signal obtained；WithIt is expressed as the radiation field distribution launching antenna and reception antenna at location region r；σ (r) is the backscattering coefficient in the region r of location.

A kind of new indoor localization method based on Wi-Fi network the most according to claim 1, it is characterised in that:

Described employing compressed sensing computational methods carry out matrix inversion to the linear model of microwave coherent imaging, and detailed process is:

The linear model of described microwave coherent imaging is carried out sliding-model control on frequency domain and spatial domain respectively, obtains matrix equation:

Wherein, subscript m represents the electromagnetic field information of m-th Frequency point, and subscript n represents the electromagnetic field information in the n-th region, location at discrete point.

Above-mentioned matrix equation is reduced to:

G=Hf

Wherein,Representing and launch antenna and the reception antenna radiation field information in target area, g is scatter echo signal, and f is the backscattering coefficient variable of target.

Use compressed sensing computational methods to solve the minima of object function by below equation to carry out matching optimization and obtain, until the minima of object function minimizes value threshold value:

Wherein, F (f) represents the object function of the backscattering coefficient variable f of target, and p represents the exponent number of norm, | | | | for p rank norm sign, R (f) is Regularization function, and λ is regulation norm and Regularization function weight parameter each other.

A kind of new indoor localization method based on Wi-Fi network the most according to claim 1, it is characterised in that: described Antenna Operation is in Wi-Fi operating frequency, and frequency dispersion degree m is more than the target sparse degree in region, location.

A kind of new indoor localization method based on Wi-Fi network the most according to claim 1, it is characterized in that: described region, location is the indoor covering Wi-Fi signal, launching antenna and reception antenna uses Wi-Fi signal to carry out irradiation and reception, location target is strong scattering object.

A kind of new indoor localization method based on Wi-Fi network the most according to claim 1, it is characterised in that: described scatter echo signal is measured by vector network analyzer and is obtained.

A kind of new indoor localization method based on Wi-Fi network the most according to claim 1, it is characterised in that: described antenna irradiation of launching obtains or transforms to target area by antenna damnification obtaining to the radiation field information positioning region formation by antenna damnification.