CN103298103A - Wi-Fi positioning method and device - Google Patents

Abstract

The invention provides a Wi-Fi positioning method. The Wi-Fi positioning method comprises the steps that Wi-Fi information in an environment to be positioned is scanned; according to the Wi-Fi information, an address with the largest possibility is matched in a data record and used as a result of Wi-Fi positioning, wherein a known address and Wi-Fi information corresponding to the known address are saved in the data record. The invention further provides a Wi-Fi positioning device. The Wi-Fi positioning device comprises a scanning unit and a searching unit, wherein the scanning unit is suitable for scanning the Wi-Fi information in an ambient environment, the searching unit is suitable for matching the address with the largest possibility in the data record, the address with the largest possibility is used as the result of the Wi-Fi positioning, and the known address and the Wi-Fi information corresponding to the known address are saved in the data record. The invention further provides an indoor and outdoor seamless positioning system. The indoor and outdoor seamless positioning system comprises an outdoor positioning module, an indoor positioning module and a switching module. The Wi-Fi positioning method and the Wi-Fi positioning device improve Wi-Fi positioning accuracy and achieve a seamless positioning effect.

Description

Wi-Fi localization method and device

Technical field

The present invention relates to satellite and wireless location technology, relate in particular to a kind of Wi-Fi localization method and device.

Background technology

In recent years, LBS (Location Based Service) is that location-based service has obtained application more and more widely in daily life.For example: destination address inquiry and the inquiry of the optimum line from the current location to the destination; Can calculate the taxi charge system of the range of driving, fare, time and route in the city in addition; And can provide mobile phone application software of searching near restaurant, hotel, subway station, park etc. etc.

The application example of this LBS is too numerous to enumerate, and realizes that the technology of the fundamental sum core of LBS is indoor, outdoor location technology.

In outdoor location technology, the Beidou satellite navigation system similar to gps system (BeiDou (COMPASS) Navigation Satellite System) is the Chinese independent development capability of implementing, the global positioning system of independent operating.The Beidou satellite navigation system space segment that China is building is made up of 5 satellites and 30 non-geo satellites, provides two kinds of methods of service to comprise open service and authorization service (belonging to second-generation system).Open service is to provide free the location at service area, test the speed and the time service service, and positioning accuracy is 10 meters, and time service precision was 50 nanoseconds, rate accuracy 0.2 meter per second.Authorization service be to authorized user safer location be provided, test the speed, time service and communication service and system health information.

To the indoor positioning technology, existing navigation system comprises transducer navigation system, wireless local network positioning system etc.Sensing positioning system needs special equipment and network support, and is subject to sensing range and the sensing network deployment of transducer, is difficult to finish accurate location and positioning service covering on a large scale.Consider in recent years, wireless lan (wlan) based on the IEEE802.11b/g agreement is universal day by day, begin built-in wireless network card in all kinds of portable terminals gradually, simultaneously " wireless city " theory makes and increasing wireless access point AP (Access Point) occurs in the environment that this all provides convenience for the Wi-Fi location technology in the world.Can utilize corresponding relation between these Wi-Fi signals and its residing geographical position as foundation, realize non-calibrating and positioning; Simultaneously do not need extra hardware device again, reduced the cost of navigation system largely, good universality is arranged.This location technology based on WLAN utilizes portable terminal (notebook, mobile phone, PDA etc.) from access point (Access Point, abbreviation AP) signal strength signal intensity (the Received Signal Strength that receives, be called for short RSS) position of inferring terminal, the general method that adopts has two kinds at present: based on the method for signal propagation model with based on the method for fingerprint model.

The former is based on signal propagation model, is the nonlinear attenuation that presents when propagating in air according to wireless signal, adopts the propagation model of signal that signal strength signal intensity is converted to distance, and then obtains positional information.The employing linear polynomial is represented the relation between signal strength signal intensity and the propagation distance.Complexity and the power consumption of this method are lower, but positioning accuracy is not high, and reason is in actual environment because barriers such as wall are very big for the loss of signal strength signal intensity, and the multipath effect that signal is propagated makes that the location is inaccurate.

The latter mainly is based on the various statistical and analytical tools of a large amount of actual observation The data based on the localization method of fingerprint model and positions model training, with the location model that obtains impact point is carried out location estimation then.Generally be divided into two stages: the phase I is off-line training step, gather each reference node of required locating area (Reference Point, RP) signal characteristic parameter with the corresponding certain location of one group of wireless fingerprint information, forms the location fingerprint database thus; Second stage is online positioning stage, and the parameter of collecting test location signal adopts the data in pattern matching algorithm and the fingerprint database to be complementary, and optimizes the physical location of estimating user.The off-line training step of phase I needs very big manpower and materials to carry out the staking-out work in early stage.

Summary of the invention

The technical problem to be solved in the present invention provides a kind of Wi-Fi localization method and device, improves the Wi-Fi positioning accuracy.

According to an aspect of the present invention, provide a kind of Wi-Fi localization method, comprising:

Wi-Fi information in step 1, the scanning environment to be positioned;

The result that locate as Wi-Fi step 2, the address that utilizes described Wi-Fi information to match the possibility maximum in data record preserves known address and corresponding Wi-Fi information thereof in the wherein said data record.

Optionally, described data record is set up based on following steps:

Steps A, the Wi-Fi information in the scanning circumstance;

Step B receives the current address title that the user imports; With

Step C, storage current address title and Wi-Fi information are to data record.

Optionally, described data record is database or data file.

Optionally, step 2 adopts the arest neighbors matching algorithm to mate, and wherein, the distance definition of the alternative address in the access point signals intensity that scans and address-signal map database or the data file is as follows:

$d=\sqrt{\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{({\mathrm{RSSI}}_{\mathrm{ci}}-{\mathrm{RSSI}}_{\mathrm{pi}})}^2}$

Wherein, RSSI _CiThe signal strength values of representing i access point of alternative address; RSSI _PiThe RSSI value of representing i access point of environment to be positioned; N represents the number of access point; D is more low, and the expression similitude is more high.

Optionally, described arest neighbors matching algorithm is the nearest neighbor algorithm of non-isometric data, wherein:

$d=\sqrt{\frac{1}{m}\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{({\mathrm{RSSI}}_{\mathrm{ci}}-{\mathrm{RSSI}}_{\mathrm{pi}})}^2}$

M represents the number of the access point that alternative address can both be detected in environment to be positioned and the data record.

Optionally, before step 2, also comprise: utilize predefined signal strength signal intensity, coupling number and/or weights threshold value to screen described data record.

According to a further aspect of the present invention, provide a kind of Wi-Fi positioner, comprising:

Scanning element is suitable for scanning the Wi-Fi information in the surrounding environment;

Retrieval unit is suitable for utilizing described Wi-Fi information to match the address of possibility maximum as the result of Wi-Fi location in data record; Preserve known address and corresponding Wi-Fi information thereof in the wherein said data record.

Optionally, described device also comprises:

Display unit, being suitable for result for retrieval is that positioning result is shown to the user.

Optionally, described device also comprises:

Input unit is suitable for receiving the corresponding current address title of the Wi-Fi information with scanning element scanning that the user imports; With

Memory cell is suitable for the Wi-Fi information of current address title with scanning element scanning is stored in the data record.

Another aspect according to the present invention provides a kind of indoor and outdoor seamless navigation system, comprising:

Outdoor locating module, indoor positioning module, and handover module;

Described outdoor locating module is suitable for carrying out satellite fix;

Described indoor positioning module is Wi-Fi positioner as claimed in claim 9;

Described handover module is suitable for asking satellite fix, if described satellite fix request meets with a response, priming chamber outside fix module if described satellite fix request does not meet with a response, starts the indoor positioning module; Also be suitable for after starting the indoor positioning module, continue the request satellite fix, and when the satellite fix request meets with a response, priming chamber outside fix module.

Compared with prior art, the invention has the advantages that:

(1) improves the Wi-Fi positioning accuracy;

(2) overcome in the past that indoor locating system and outdoor navigation system can't can't automatically switch when environment change, thereby reached the problem of seamless location.

Description of drawings

Fig. 1 is the structural representation of Wi-Fi positioner according to an embodiment of the invention;

Fig. 2 is an example position fixes process schematic diagram of Wi-Fi locating module shown in Figure 1;

Fig. 3 is the structural representation of the seamless navigation system of indoor and outdoor according to an embodiment of the invention;

Fig. 4 is a course of work schematic diagram of the seamless navigation system of indoor and outdoor shown in Figure 3.

Embodiment

In order to make purpose of the present invention, technical scheme and advantage clearer, below in conjunction with accompanying drawing, the present invention is described in more detail.Should be appreciated that specific embodiment described herein only in order to explaining the present invention, and be not used in restriction the present invention.

In order to improve indoor position accuracy, in one embodiment of the invention, provide a kind of Wi-Fi positioner.As shown in Figure 1, this Wi-Fi positioner 300 comprises: address registration submodule 310 and real-time locator module 320.

Wherein, address registration submodule 310 further comprises: scanning element 311, input unit 312, and memory cell 313.

Address registration submodule 310 is carried out following steps:

Wi-Fi information in step 201, scanning element 311 scanning circumstances, i.e. the MAC Address of all Wi-Fi access points (AP:Access Point), signal strength signal intensity, and title;

Step 202, input unit 312 receive the current address title of user's input;

Step 203, memory cell 313 store the current address title into data record 400 with Wi-Fi information.

According to one embodiment of present invention, data record 400 is address-signal map database; Wherein, the data in the database and form can be routine as shown in table 1.

Table 1

In addition, according to one embodiment of present invention, locator module 320 further comprises in real time: scanning element 321, retrieval unit 322 and display unit 323.

The execution in step of locator module 320 is as follows in real time:

Wi-Fi information in step 301, the scanning element 321 scanning surrounding environment comprises the MAC Address of the Wi-Fi that all can search, signal strength signal intensity, and the title of AP;

Step 302, retrieval unit 322 according to matching algorithm, match the address of possibility maximum based on address-signal map database.

According to one embodiment of present invention, described address-signal map database can be positioned on the server, and step 302 can also comprise:

Step 3021, retrieval unit 322 send to server and request location with described Wi-Fi information;

Step 3022, server are received after this Location Request, based on address-signal map database, according to matching algorithm, match the address of possibility maximum, return to retrieval unit 322;

According to one embodiment of present invention, in real time locator module 320 can also execution in step:

Step 303, be that positioning result is shown to the user by display unit 323 with result for retrieval.

At system's initial operating stage, because the registered address is sparse, might locatees inaccurate or can't locate, if user's this moment knows the exact address of present position, can carry out address registration/modification by address registration submodule 310, to enrich address information in the database gradually.Along with being on the increase of the data in the data record 400, system will provide more and more accurate in locating service, can reach the purpose of gradual raising positioning accuracy.

A location instantiation procedure of above-mentioned Wi-Fi locating module as shown in Figure 2.

According to another embodiment of the invention, scanning element 321 is same scanning element with scanning element 311.

According to one embodiment of present invention, preferred, in step 3022, described matching algorithm is arest neighbors matching algorithm (being nearest neighbor method).Nearest neighbor method (Nearest Neighbor, NN) be a kind of sorting technique, the basic thought of nearest neighbor method is that a sample is not known its classification at first, in feature space, find the sample point the most close with it, the classification of these nearest sample neighbours' classification as this unknown classification sample.Wherein, the tolerance of the distance between the sample (similitude) generally can adopt common Euclidean distance.

The distance definition of certain alternative address is as follows in the AP that scans and the database:

$d=\sqrt{\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{({\mathrm{RSSI}}_{\mathrm{ci}}-{\mathrm{RSSI}}_{\mathrm{pi}})}^2}---\left(1\right)$

Wherein, RSSI _Ci: the RSSI value of i AP of the alternative address of lane database; RSSI _Pi: the RSSI value of i AP of the current scanning of finder in real time; N: the number of the AP that receives; D is more low, and the expression similitude is more high.

A plurality of addresses are arranged in the database, and there are a plurality of AP each address, the corresponding RSSI of each AP.RSSI _CiWith RSSI _PiDoing difference, to refer to that Rssi with a certain identical AP in address in the current AP that scans and the database does poor, namely earlier the identical AP in both sides is found out, then one by one correspondence do poor.

Dynamically in the environment that changes, because locating area changes at the AP data set that calibration phase and positioning stage receive between each calibration point, perhaps for some special point, perhaps also have the AP number of different acceptance, the AP number that is adapted to change so the calculating of d is had to.

So described arest neighbors matching algorithm is the nearest neighbor algorithm of non-isometric data.And use standardized Euclidean distance d, standardization to consider that the AP that receives counts the value that m is a variation between calibration point.

$d=\sqrt{\frac{1}{m}\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{({\mathrm{RSSI}}_{\mathrm{ci}}-{\mathrm{RSSI}}_{\mathrm{pi}})}^2}---\left(2\right)$

Wherein, m: the number of the AP that the alternative address of current address and database can both be detected, unfixing.

According to one embodiment of present invention, preferably, in order further to improve positioning accuracy, can will only comprise the one-dimensional characteristic vector of signal strength signal intensity originally, be transformed into the multidimensional characteristic vectors that comprises signal strength signal intensity, coupling number, signal strength signal intensity variance, weights, frequency domain character etc.To select by coupling number, signal strength signal intensity variance, these features of weights and can screen the calculating that participates in d with the AP that represents address information, such as signal strength values is greater than-95.Do further below and elaborate.

The coupling number: the coupling number is in order to reflect in the AP number that searches and the database coincidence degree of AP number in a certain the registered address.Use two sign amount: M (the AP number on the coupling accounts for the percentage of the total number of AP that searches) and N (the AP number on the coupling accounts for the percentage of the total number of AP of this record in the database) to represent.Have only the M of working as and N value to reach certain value, this records just to have and is taken as the meaning that mate alternative address.The N empirical value of M is M=45%, N=45%.Cross when low when M value and N value, illustrate that the address itself that this record registers is just far apart from current location, should not continue participation relatively.Be worth determining and to go to determine according to experiment for M value and N under the varying environment.

The signal strength signal intensity variance: the variance of signal strength signal intensity equally also is to screen for the record that will not have matching condition.By adding up in unified place, the signal strength signal intensity of same AP.Can calculate its average and variance.Thereby set up the Gaussian distribution model of this signal strength signal intensity.According to probability distribution, a scope R is set.When the coupling of location, have only the calculating that when alternative AP intensity is in this scope, just is used for participating in Euclidean distance, do not consider otherwise get rid of.

Weights: because the high dynamic of Wi-Fi signal, same AP As time goes on, variation that may occurrence positions.The AP that can search that is in the same localities also might increase or reduce.So the address information of registration is according to the difference of time, the useful information that provides for the location is different.The data reference value more near apart from the time is more big, and the data reference value more far away apart from the time is more little.In order to embody the difference of this reference value, according to the time of address registration, add a weights W, further improve locating accuracy.

Determining of W value can be according to following method: define the total n bar of record of the registration in same place, by present time recently to be followed successively by farthest T1, T2 ..., Tn}, then the k bar (weights of 1≤k≤n) record are defined as follows:

$W_k=\frac{1}{2^k}(1\≤k\≤n-1)$

$W_k=\frac{1}{2^{k-1}}(k=n)$ (3 and 4)

More than these features constituted the sign vector Vector={d of matching addresses, M, N, R, W} jointly.In location process, each that characterizes in the vector has all played the effect of screening.Be exemplified below:

S401, data scanning; The data that scan are as shown in table 2.

Table 2

The AP number	AP1	RSSI1	AP2	RSSI2	...
						7	aa:aa:aa:aa:aa:aa	-58	bb:bb:bb:bb:bb:bb	-70	...

Address stored in the database-signal map data (address of back, expression registration recently) as shown in table 1.

S402, contrast scans to data in AP and the AP in the record in the database; Find to have gone up 5 AP (comprising AP1, AP2 etc.) with the address for " main building " this record coupling, it is identical that 5 AP are namely arranged.

The total number of the AP that searches is 7, and the total number of AP of this address record is 8.So M is 5/7, N is 5/8.Satisfy condition: M＞45% and N＞40%.

S403, determine scope R value; Statistics AP (aa:aa:aa:aa:aa:aa) signal strength values calculates its average and variance, sets up Gaussian distribution model then, can get the scope of the signal of certain probability that comprises this Gaussian Profile (for example 80%), be R～55 ,-65}.

Because the AP signal strength values-58 of scanning drops in this scope, so the signal strength values of this AP can participate in the calculating of Euclidean distance.

S404, determine weights W; In table 1, because the information that an address name of back is called " main building " is more recent registration, so when address choice, have bigger weight.

Determining of W value can be according to following method: define the total n bar of record of the registration in same place, by present time recently to be followed successively by farthest T1, T2 ..., Tn}, then the k bar (weights of 1≤k≤n) record are defined as follows:

$W_k=\frac{1}{2^k}(1\≤k\≤n-1)$

$W_k=\frac{1}{2^{k-1}}(k=n)$

Above-mentioned weights definition is an empirical formula, in other embodiments of the invention, also can adopt other empirical equations to calculate weights.

S405, calculating Euclidean distance d; The AP that goes out in order to last screening conditional filtering participates in the calculating of Euclidean distance d.

In order between indoor positioning and outdoor location, to accomplish seamless switching, according to one embodiment of present invention, provide a kind of indoor and outdoor seamless navigation system.As shown in Figure 3, this system 100 comprises: location handover module 101, outdoor locating module 102 and indoor positioning module 103 (being the Wi-Fi positioner that provides in above-described embodiment).

The course of work of system 100 comprises as shown in Figure 4:

Step 101, the navigation system 100 that starts;

Step 102, location handover module 101 request GPS or big-dipper satellite location;

Step 103, location handover module 101 check whether described Location Request obtains significant response; Described significant response refers to receive abundant satellite and sufficiently high signal strength signal intensity, to realize positive locating function;

If the described Location Request of step 104 obtains significant response, location handover module 101 determines that user, terminal or navigation system 100 are positioned at outdoor, if and current outdoor locating module 102 is not activated, priming chamber outside fix module 102 carries out GPS or follow the tracks of the big-dipper satellite location; Turn to step 2;

If the described Location Request of step 105 does not obtain significant response, location handover module 101 determines that user, terminal or navigation system 100 are positioned at indoor, if and current indoor positioning module 103 is not activated, starts indoor positioning module 103 and carry out indoor positioning; Turn to step 2.

For example, when terminal or navigation system 100 after starting under the outdoor environment, execution in step 102～104 is utilized the location of big-dipper satellite, be shown as a point on a map at locating terminal, along with the movement of terminal, the point on the map also is corresponding movement.

When terminal or navigation system 100 enter an indoor environment, when for example entering a mansion, terminal positioning goes out the mansion title at place, but can lose the framing signal of big-dipper satellite simultaneously.After the execution in step 2,3,5, terminal or navigation system 100 start indoor positioning module 103, and are invoked at the internal map of this mansion of building that navigates to when Big Dipper satellite signal is lost, thereby use indoor Wi-Fi signal to realize the location.

When locating terminal or navigation system 100 by indoor when outdoor, owing to can regularly initiate connection request to big-dipper satellite after the terminal execution in step 2,3,5, this connection request can't obtain significant response under indoor environment, when terminal by indoor when outdoor, since can receive the signal of big-dipper satellite, so this request will obtain significant response at once, thus the outdoor locating module 102 of system restart, and call outdoor map, carry out the big-dipper satellite location.Close indoor positioning module 103 simultaneously, to save electric weight.

In other embodiments of the invention, can also comprise the manual switchover function on terminal or the navigation system 100, satisfy the user and initiatively switch needs.

Should be noted that and understand, under the situation that does not break away from the desired the spirit and scope of the present invention of accompanying Claim, can make various modifications and improvement to the present invention of foregoing detailed description.Therefore, the scope of claimed technical scheme is not subjected to the restriction of given any specific exemplary teachings.

Claims (10)

1. Wi-Fi localization method comprises:

Wi-Fi information in step 1, the scanning environment to be positioned;

The result that locate as Wi-Fi step 2, the address that utilizes described Wi-Fi information to match the possibility maximum in data record preserves known address and corresponding Wi-Fi information thereof in the wherein said data record.

2. method according to claim 1, wherein, described data record is set up based on following steps:

Steps A, the Wi-Fi information in the scanning circumstance;

Step B receives the current address title that the user imports; With

Step C, storage current address title and Wi-Fi information are to data record.

3. method according to claim 1 and 2, wherein, described data record is database or data file.

4. method according to claim 1, wherein, step 2 adopts the arest neighbors matching algorithm to mate, and wherein, the distance definition of the alternative address in the access point signals intensity that scans and address-signal map database or the data file is as follows:

$d=\sqrt{\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{({\mathrm{RSSI}}_{\mathrm{ci}}-{\mathrm{RSSI}}_{\mathrm{pi}})}^2}$

Wherein, RSSI _CiThe signal strength values of representing i access point of alternative address; RSSI _PiThe RSSI value of representing i access point of environment to be positioned; N represents the number of access point; D is more low, and the expression similitude is more high.

5. method according to claim 4, wherein, described arest neighbors matching algorithm is the nearest neighbor algorithm of non-isometric data, wherein:

$d=\sqrt{\frac{1}{m}\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{({\mathrm{RSSI}}_{\mathrm{ci}}-{\mathrm{RSSI}}_{\mathrm{pi}})}^2}$

M represents the number of the access point that alternative address can both be detected in environment to be positioned and the data record.

6. method according to claim 1 wherein, also comprised before step 2: utilize predefined signal strength signal intensity, coupling number and/or weights threshold value to screen described data record.

7. Wi-Fi positioner comprises:

Scanning element is suitable for scanning the Wi-Fi information in the surrounding environment;

Retrieval unit is suitable for utilizing described Wi-Fi information to match the address of possibility maximum as the result of Wi-Fi location in data record; Preserve known address and corresponding Wi-Fi information thereof in the wherein said data record.

8. device according to claim 7 also comprises:

Display unit, being suitable for result for retrieval is that positioning result is shown to the user.

9. device according to claim 7 also comprises:

Input unit is suitable for receiving the corresponding current address title of the Wi-Fi information with scanning element scanning that the user imports; With

Memory cell is suitable for the Wi-Fi information of current address title with scanning element scanning is stored in the data record.

10. seamless navigation system of indoor and outdoor comprises:

Outdoor locating module, indoor positioning module, and handover module;

Described outdoor locating module is suitable for carrying out satellite fix;

Described indoor positioning module is Wi-Fi positioner as claimed in claim 9;

Described handover module is suitable for asking satellite fix, if described satellite fix request meets with a response, priming chamber outside fix module if described satellite fix request does not meet with a response, starts the indoor positioning module; Also be suitable for after starting the indoor positioning module, continue the request satellite fix, and when the satellite fix request meets with a response, priming chamber outside fix module.

Images