CN103957503A - Method improving robustness of WiFi fingerprint location by utilizing sensor to count steps - Google Patents

Abstract

The invention relates to a method improving robustness of WiFi fingerprint location by utilizing a sensor to count steps. According to the method, at first, a wireless access point of each sampling point is detected to obtain a signal intensity value, the signal intensity value and an MAC address of each sampling point together form a position fingerprint of the position, and the position fingerprints are stored into a fingerprint database; then the wireless access points are scanned at an unknown position A to obtain the position fingerprints, the position fingerprints and the fingerprints in the fingerprint database are matched, and a geographical coordinate corresponding to the proximate fingerprint is the current position of the position A; finally, whether the terminal position changes is judged according to the change of the accelerated speed in the vertical direction of a tested person, and therefore whether to carry out step counting operation is decided. According to the method, the sensor step counting detection method is used for detecting whether the terminal position changes and judging whether position information needs to be upgraded, the problem of terminal position information errors caused by unstable WiFi signals under the static state is solved, and therefore the robustness of WiFi fingerprint location is improved.

Description

A kind of method of utilizing transducer meter step to improve WiFi fingerprint location robustness

Technical field

The invention belongs to wireless positioning field, be specifically related to a kind of method of utilizing transducer meter step to improve WiFi fingerprint location robustness.

Background technology

In recent years, along with the fast development of wireless network communication technique, the market demand of indoor positioning is also increasing.For example, in underground parking, by location, can vehicle location be inquired about and be followed the tracks of; Under mine, can position trapped personnel, thereby can complete fast rescue duty.

At present, most widely used location technology is " GPS (Global Position System) (GPS) ".But it is outdoor that GPS technology is mainly used in, spacious region, such as desert, ocean etc.If between skyscraper inside or high building, GPS locating effect will be unsatisfactory, cannot meet people's demand.Therefore, in application, there is significant limitation.

In recent years, the application of WiFi technology on navigation system market develops rapidly.The access point of WiFi network is dispersed throughout the places such as market, school, hospital, parking lot, utilizes WiFi technology in limited region, property, vehicle and personnel to be located in real time and to be followed the tracks of.Thereby solving GPS technology cannot complete indoor or barrier more regions and locate this problem.In addition,, due to the spreading all over property of WiFi access point, location cost is also reduced greatly.

Navigation system based on WiFi technology has had good application at home and abroad at present.For example Ghent, Belgium university hospital utilizes this technology can also provide patient's position data with the medical communication information that patient is provided, and has greatly improved hospital work efficiency.

Wi-Fi location technology has multiple, and adopting is at present the fingerprint positioning method based on Wi-Fi the most widely, and Wi-Fi fingerprint refers in a certain position, signal strength signal intensity indication (RSSI) value that terminal receives from a plurality of Wi-Fi WAP (wireless access point) (AP).Before positioning, first need target area to carry out fingerprint sampling, and set up fingerprint database.In actual location process, adopt specific algorithm to contrast the actual received signals fingerprint recording of terminal and the signal strength signal intensity in fingerprint database, be user's current location with the immediate sampled point of received signals fingerprint numerical value.

In adopting the indoor positioning technology of RSSI, WiFi jitter and being easily disturbed, therefore, when the signal strength signal intensity that terminal receives and the Data Comparison in fingerprint database, the positional information of returning there will be saltus step, causes position error, affects positioning precision.

Summary of the invention

For WiFi jitter, be easily disturbed and cause poor robustness to affect the problems such as positioning precision, the present invention proposes a kind of method of utilizing transducer meter step to improve WiFi fingerprint location robustness.

Basic principle of the present invention is: vertical direction acceleration value cyclic variation during human body walking, utilize the transducer human body in terminal whether to walk about, when detecting while not having position to move, do not change positional information, thereby improved the robustness of WiFi location.Described terminal refers to have WiFi function, and with the smart mobile phone of the transducers such as accelerometer, gyroscope, magnetic compass.

Utilize transducer meter step to improve a method for WiFi fingerprint location robustness, comprise the following steps:

Step 1: determine all sampled point number N in whole locating area;

Step 2: terminal WiFi function is opened, detected the signal strength values that obtains k wireless access point AP, i.e. RSSI n sample point ₁, RSSI ₂..., RSSI _k; N=1,2 ..., N;

Step 3: each RSSI MAC Address ID corresponding with it formed together to the location fingerprint matrix F P of sampled point, the location fingerprint of n sampled point ${\mathrm{FP}}_n={\left[\begin{array}{ccc}{\mathrm{RSSI}}_i& ...& {\mathrm{RSSI}}_j\\ {\mathrm{ID}}_i& ...& {\mathrm{ID}}_j\end{array}\right]}^T,$ ID _ibe RSSI _icorresponding MAC Address;

Step 4: the coordinate FP that writes down n sampled point _n(x, y), and by the coordinate FP of this sampled point _n(x, y) and location fingerprint FP _nadd in fingerprint database;

Step 5: repeated execution of steps 2～4, until all sampled points in traversal sample area, for sample area is set up complete location fingerprint database;

Step 6: in a certain unknown position A place scanning wireless access point AP, obtain the RSSI value of k AP, obtain the location fingerprint of A position according to step 3, obtain RSSI sequence and the MAC Address corresponding to each RSSI at A place, position;

Step 7: the MAC Address sequence of the location fingerprint of all sampled points in the MAC Address sequence of the location fingerprint of A position and fingerprint base is mated;

Step 8: mate the current location that geographical coordinate that consistent sampled point is corresponding is position A;

Step 9: test person handheld terminal when walking, read previous moment human body vertical direction acceleration mean value a by the transducer of terminal ₀(t-1), initial time vertical direction acceleration mean value is set as a ₀(0), size equals gravity acceleration g, the saltus step of shielding WiFi signal, and terminal positional information is constant;

Step 10: detect in real time vertical direction acceleration, select peak value a ₁and low valley a (t) ₂(t);

Step 11: calculate the poor of vertical direction acceleration peak value and low valley and previous moment vertical direction acceleration mean value, i.e. a ₁(t)-a ₀and a (t-1) ₀(t-1)-a ₂(t), if a ₁(t)-a ₀and a (t-1) ₀(t-1)-a ₂(t) be all less than threshold value r, be considered as slight vibrations, go to step 9; Otherwise, execution step 12; R generally gets 0.35;

Step 12: terminal location changes, counts step;

Step 13: calculate current time vertical direction acceleration mean value a ₀(t) size, formula is as follows:

a ₀(t)=[a ₁(t)+a ₂(t)]/2

Step 14: return to step 6, obtain new positional information, repeated execution of steps 6～14, until finish.

Compared with prior art, the present invention has the following advantages:

Whether the method for the invention changes by transducer meter step detection method check terminal location, determine whether and need to upgrade positional information, efficiently solve under inactive state and occur wrong problem because WiFi jitter causes terminal positional information, thereby improved the robustness of WiFi fingerprint location.

Accompanying drawing explanation

Fig. 1 is the flow chart of method involved in the present invention;

Fig. 2 is that indoor positioning sampled point arranges plane graph, and " " represents sampling point position;

Fig. 3 is testing equipment schematic layout pattern.

Embodiment

Below in conjunction with drawings and Examples, the present invention will be further described.

Embodiment experimental situation is everybody rooms of 5 meters of length and width, is furnished with 5 wireless access point AP, as shown in Figure 3.

Utilize transducer meter step improve WiFi fingerprint location robustness method flow chart as shown in Figure 1, comprise the following steps:

(1) in room floor plan, set up rectangular coordinate system, using first sampled point as initial point, and be designated as reference point 1(0,0), then every sampled point of one meter of mark, totally 36 sampled points in whole plane graph, i.e. 36 reference points, as shown in Figure 2;

(2) terminal is in reference point 1(0,0) position, scan 5 wireless access point AP 1～AP5, the signal strength values RSSI of AP1～AP5 detected, be respectively :-66 ,-73 ,-100 ,-90 ,-80, the dBm of unit;

(3) 5 the signal strength signal intensity RSSI values and the corresponding MACID thereof that step (2) are obtained ₁, ID ₂, ID ₃, ID ₄, ID ₅be expressed as the location fingerprint of reference point 1:

${\mathrm{FP}}_1={\left[\begin{array}{c}-66,-73,-100,-90,-80\\ {\mathrm{ID}}_1,{\mathrm{ID}}_2,{\mathrm{ID}}_3,{\mathrm{ID}}_4,{\mathrm{ID}}_5\end{array}\right]}^T$

(4) terminal is with reference to the geographical coordinate FP of point 1 ₁(0,0) and corresponding location fingerprint FP ₁be saved in fingerprint database, complete the fingerprint sampling of reference point 1;

(5) to 36 sampled point repeated execution of steps (2)～(4) in sample area, set up the fingerprint database of sample area, by fingerprint database data upload to server;

(6) experimenter's handheld terminal, at unknown position A place scanning wireless access point AP 1～AP5, is write down the RSSI value of five AP that scan ,-71 ,-75 ,-105,-95 ,-82, and corresponding MAC Address [ID1, ID2, ID3, ID4, ID5], obtain the fingerprint of current position: ${\mathrm{FP}}_A={\left[\begin{array}{c}-71,-75,-105,-95,-82\\ \mathrm{ID}1,\mathrm{ID}2,\mathrm{ID}3,\mathrm{ID}4,\mathrm{ID}5\end{array}\right]}^T;$

(7) fingerprint base in the automatic Download Server of terminal, by the MAC Address sequence FP of the location fingerprint of position A _a[ID ₁, ID ₂, ID ₃, ID ₄, ID ₅] with server fingerprint base in MAC Address sequence in the location fingerprint of 36 sampled points carry out fingerprint matching;

(8) by the geographical coordinate FP of coupling _a(0,0) is as the current location of unknown position A.

(9) end sensor starts to read the acceleration initial average output value a of vertical direction ₀(0) a, obtaining first ₀(0) value is 10, the saltus step of shielding WiFi signal, and terminal display position Information preservation is constant.

(10) experimenter steps forward a step, and transducer obtains the peak value a of vertical direction acceleration ₁(1) with low ebb a ₂(1) value is respectively 12 and 9.

(11) calculate the excursion of accekeration, the variable quantity that obtains acceleration peak value and low ebb is respectively: 12-10=2,10-9=1.

(12) acceleration change amount 2>r=0.35,1>0.35, result of determination is that terminal is subjected to displacement.

(13) calculate current time vertical direction acceleration mean value a ₀(1), a ₀(1)=(12+9)/2=10.5.

(14) return to step (6), obtain new positional information, repeated execution of steps (6)～(14), until finish.

Claims (1)

1. utilize transducer meter step to improve a method for WiFi fingerprint location robustness, application has WiFi function and measures with the intelligent mobile phone terminal of acceleration transducer, it is characterized in that, said method comprising the steps of:

Step 1: determine all sampled point number N in whole locating area;

Step 2: terminal WiFi function is opened, detected the signal strength values that obtains k wireless access point AP, i.e. RSSI n sample point ₁, RSSI ₂..., RSSI _k; N=1,2 ..., N;

Step 3: each RSSI MAC Address ID corresponding with it formed together to the location fingerprint matrix F P of sampled point, the location fingerprint of n sampled point ${\mathrm{FP}}_n={\left[\begin{array}{ccc}{\mathrm{RSSI}}_i& ...& {\mathrm{RSSI}}_j\\ {\mathrm{ID}}_i& ...& {\mathrm{ID}}_j\end{array}\right]}^T,$ ID _ibe RSSI _icorresponding MAC Address;

Step 4: the coordinate FP that writes down n sampled point _n(x, y), and by the coordinate FP of this sampled point _n(x, y) and location fingerprint FP _nadd in fingerprint database;

Step 5: repeated execution of steps 2～4, until all sampled points in traversal sample area, for sample area is set up complete location fingerprint database;

Step 6: in a certain unknown position A place scanning wireless access point AP, obtain the RSSI value of k AP, obtain the location fingerprint of A position according to step 3, obtain RSSI sequence and the MAC Address corresponding to each RSSI at A place, position;

Step 7: the MAC Address sequence of the location fingerprint of all sampled points in the MAC Address sequence of the location fingerprint of A position and fingerprint base is mated;

Step 8: mate the current location that geographical coordinate that consistent sampled point is corresponding is position A;

Step 9: test person handheld terminal when walking, read previous moment human body vertical direction acceleration mean value a by the transducer of terminal ₀(t-1), initial time vertical direction acceleration mean value is set as a ₀(0), size equals gravity acceleration g, the saltus step of shielding WiFi signal, and terminal positional information is constant;

Step 10: detect in real time vertical direction acceleration, select peak value a ₁and low valley a (t) ₂(t);

Step 11: calculate the poor of vertical direction acceleration peak value and low valley and previous moment vertical direction acceleration mean value, i.e. a ₁(t)-a ₀and a (t-1) ₀(t-1)-a ₂(t), if a ₁(t)-a ₀and a (t-1) ₀(t-1)-a ₂(t) be all less than threshold value r, be considered as slight vibrations, go to step 9; Otherwise, execution step 12; R gets 0.35 conventionally;

Step 12: terminal location changes, counts step;

Step 13: calculate current time vertical direction acceleration mean value a ₀(t) size, formula is as follows:

a ₀(t)=[a ₁(t)+a ₂(t)]/2

Step 14: return to step 6, obtain new positional information, repeated execution of steps 6～14, until finish.