CN103260239B - A kind of positioning of mobile equipment method based on WIFI - Google Patents

Abstract

The present invention relates to a mobile device positioning method based on WIFI, comprising the following steps: 1. Evenly sow hypothetical points within the range where the mobile device may move to form a set of hypothetical points; 2. Periodically monitor the mobile device to obtain the location of the device The relative movement distance vector, based on which the current position of each hypothetical point in the hypothetical point set is updated, and the position information of the hypothetical point and the received signal strength of each WIFI source are recorded after each update; 3. For each hypothetical point, the Each position data on the trajectory is fitted with the signal strength data of N WIFI sources 1 to form a fitting function, and N variances are obtained, based on which the unreliable value of the hypothetical point is obtained, and deleted according to the threshold value preset by the user Assume points with high unreliability values, and repeat step 3 until user requirements are met. This method can realize effective positioning in an indoor environment, and can make up for the practical difficulty of achieving accurate device positioning in current environments such as large buildings.

Description

A WIFI-based mobile device positioning method

technical field

The invention relates to a mobile device positioning technology, in particular to a mobile device positioning method based on WIFI signals.

Background technique

In the GPS system, the satellite uses short-wave signals to enable the receiver to receive the position broadcast by the satellite, thereby determining the location of the receiver. Due to the use of short-wave communication means, GPS has deficiencies in global positioning, that is, short-wave signals are difficult to penetrate obstacles and cannot complete the positioning of indoor venues.

In order to overcome the shortcomings of the GPS system, people are looking for a tool that can complete indoor positioning, which is used to supplement the satellite positioning system and improve global positioning. Since IEEE established the 802.11 standard in 1999, WIFI has become one of the most widely used wireless technologies and device communication technologies in the world. In the early days, WIFI users were roughly positioned by the physical location of WIFI, which was used to advertise avatars and broadcast targeted programs. In recent years, through the improvement of WIFI equipment, researchers have found that the received WIFI signal strength is different from the distance of WIFI receiving equipment. There is an obvious relationship between them, which makes it possible to accurately locate mobile devices through WIFI.

Through statistics on different WIFI sources and mobile devices, the researchers found that after large-scale data input, there is the following relationship between WIFI signal strength and the distance of WIFI receiving devices:

H=K+E*log ₁₀ (dis)

Among them, dis is the distance, H is the signal strength, K and E are natural values, and their values are closely related to the WIFI signal source equipment, mobile equipment, and environment. Usually when the WIFI source, mobile device, and environment are fixed, the KE value does not fluctuate much. Therefore, the environmental factor KE of a WIFI source can be determined after a large number of measurements of H-dis pairs.

The WIFI positioning method based on offline learning is to measure the signal strength of the device at a standard length under a fixed environment and a WIFI source. After collecting a large number of distances and their corresponding signal strengths, they are fitted and determined. The environmental factor KE value in Equation 1. Then, put the mobile device into an unknown environment, and determine the distance between the device and the WIFI source through the signal strength. At present, WIFI positioning mostly adopts the offline learning scheme.

However, offline learning schemes also have problems. The relationship between WIFI signal strength and distance is not static, and its relationship is affected by the performance of mobile devices, WIFI source status, and even ambient temperature, which is specifically shown in the distance-strength formula (H=K+E*log ₁₀ (dis )) on the two environmental factors K and E. Since the offline learning time is not the same as the actual positioning time, after the device completes the environment-based offline learning, it is likely that the device will change due to a long interval or other reasons, which will make the KE environment factor unpredictable. Fluctuations, such as under such conditions, continue to carry out post-location positioning, although the intensity information does not change, but the estimated distance value cannot be obtained, resulting in errors.

Moreover, the WIFI positioning method with offline learning steps is difficult to apply in the actual positioning process. It is difficult for the system to require each device that expects to obtain geographic location information to perform an offline learning process. Because this offline learning requires manual intervention by the user at a specific location, it is difficult to achieve. And this process of offline learning or manual intervention is the only process for this equipment to confirm the KE value of the environmental factor.

Contents of the invention

The purpose of the present invention is to address the problems of the prior art and propose a WIFI positioning method, which can weaken the impact of the dynamically changing environment on the WIFI positioning accuracy in the original "offline positioning" method, and solve the problem of "offline positioning" to a certain extent. Positioning" method is difficult to apply to the problem.

The present invention is realized by adopting the following technical solutions:

A WIFI-based mobile device positioning method, comprising the following steps:

1. In the range where the mobile device may move, spread the hypothetical points evenly to form a set of hypothetical points; within this range, there are N WIFI sources;

2. According to the time interval set by the user, the mobile device is regularly monitored. The data measured each time includes the measurement time, the speed of the device, and the signal strength of each WIFI source received; after each measurement, through the device speed and measurement time, Obtain the relative movement distance vector of the device, update the current position of each hypothetical point in the hypothetical point set accordingly, and record the position information of the hypothetical point and the received signal strength of each WIFI source after each update;

3. When the record length, that is, the number of measurements reaches the length S preset by the user, execute the following steps in a loop until it converges to the range permitted by the user:

1) For each hypothetical point, fit each position data on the recorded track with the signal strength data of WIFI source 1 at that position to form a fitting function and obtain a variance of 1;

2) Fit each position data on the hypothetical point record track with the signal strength data of WIFI source i at this position to form a fitting function and obtain variance i; where i=1,2,...,N;

3) Add variance 1, variance 2, ..., variance N to get the unreliable value of the hypothetical point;

4) Calculate the unreliable value of each hypothetical point, sort the hypothetical points according to the unreliable value, and delete the hypothetical point with high unreliable value according to the threshold value preset by the user;

5) For the retained hypothetical points, record the latest measurement data and discard the oldest measurement data.

As an optimization scheme of the present invention, after the point set is converged in step 3, it also includes the step of diverging the point set, centering on the simulated point reserved in step 3, performing near-intensity within the range where the equipment may move Distant points are scattered to form a new hypothetical point set, and the convergence process described in step 3 is performed on this set until the system requirements are met.

Beneficial effect

This method can realize effective positioning in an indoor environment, and can make up for the practical difficulty of achieving accurate device positioning in the current environment of large buildings, such as stores and gymnasiums.

Description of drawings

Fig. 1 is a network environment diagram based on three WIFI sources;

Fig. 2 is the flowchart of the positioning method of the embodiment;

Figure 3 is a schematic diagram of the track recording format.

Detailed ways

The preferred embodiments of the present invention will be specifically described below in conjunction with the accompanying drawings.

This embodiment constructs a hotspot environment based on three WIFI sources, wherein the relative positions of the three WIFI sources are known, as shown in Figure 1, the three WIFI sources construct a hotspot network of an isosceles right triangle, wherein the right angle sides The length is L.

The method of the present embodiment is shown in Figure 2, comprises the following steps:

1. In the range where the mobile device may move, evenly spread the hypothetical points to form a set of hypothetical points;

The quality of the hypothetical point set is one of the keys to the convergence speed of this method. If the hypothetical point set happens to contain a certain hypothetical point that is closer to the real position of the device, the convergence of the method can be completed within dozens of time periods. If The set of hypothetical points does not contain the real position of the device, and even almost all the hypothetical points are far away from the real point of the device. Although the convergence of the method can still be completed, its convergence speed will be greatly reduced. The set of hypothetical points is critical for this purpose.

In the initialization stage, in order to make the hypothetical point set not concentrate in a certain area and not miss a certain area, the method of evenly spreading points is usually adopted. First determine the length L of the right-angled side of the site according to the environmental conditions, and then determine the number D of midpoints per unit length as required.

According to some actual needs, the probability of equipment appearing in some places is greater, and the number of hypothetical points in this area can be appropriately increased to obtain a faster convergence of point sets.

In this embodiment, the initial area is set as a square with side length L, and the three WIFI sources are respectively located on three vertices of the square. The more assumed points per unit length, the higher the computational complexity, but the more accurate the result. In this embodiment, for a small experimental environment, hypothetical points are set at a distance of 1 cm.

2. According to the time interval set by the user, the mobile device is regularly monitored. The data measured each time includes the measurement time, the speed of the device, and the signal strength of each WIFI source received; after each measurement, through the device speed and measurement time, Obtain the relative movement distance vector of the device, update the current position of each hypothetical point in the hypothetical point set accordingly, and record the position information of the hypothetical point and the received signal strength of each WIFI source after each update.

In this embodiment, each measurement data needs to be sent to the server for real-time calculation and processing, so the time interval is set as the minimum time for synchronization between the mobile device (mobile phone) and the computer.

The speed of the mobile device is obtained through the built-in gyroscope, accelerometer and compass of the mobile device. Three-axis gyroscope, three-axis accelerometer, electronic compass and other devices are widely used in mobile phones and other devices. At present, all mobile devices with this device include Apple mobile phones above Iphone4, Galaxy Note, etc. There are many ways to obtain speed through devices such as gyroscopes and accelerometers, with different complexity and accuracy. The method adopted in this embodiment is:

1. Make the device static;

2. When the device starts to move, use the accelerometer to obtain the acceleration in various directions at a certain moment;

3. Use the gravity value in the accelerometer to judge the attitude of the device (if the device rotates, use the gyroscope to assist at the same time), and project the acceleration of the device on the horizontal direction based on the device;

4. Utilize the northing function of the electronic compass to project the acceleration based on the horizontal direction of the device into the horizontal acceleration based on the true north direction;

5. Continue to record the acceleration information at the next moment;

6. Integrate the acceleration for a certain period of time to obtain the device velocity (velocity and direction of the device) at the end of the time.

Through the obtained device speed and time stamp, the relative movement distance vector can be obtained. According to the relative movement distance vector, the current position of each hypothetical point in the hypothetical point set is updated, and the current position and the strength of each received WIFI source are recorded.

With the continuous transmission of data from the mobile terminal, the server will record the location information and signal strength of the assumed point after each update according to the above method. The form of the record on the server is shown in Figure 3, where the direction of the arrow indicates the contained content.

3. When the record length, that is, the number of measurements reaches the length S preset by the user, perform the following steps for all hypothetical points in the point set:

1. Fit each position of the hypothetical point trajectory with the strength of WIFI source 1 at that position to form a fitting function and obtain a variance of 1;

2. Fit each position of the hypothetical point trajectory with the strength of the WIFI source 2 at that position to form a fitting function and obtain the variance 2;

3. Fit each position of the hypothetical point trajectory with the strength of the WIFI source 3 at that position to form a fitting function and obtain the variance 3;

4. Expand and add variance 1, variance 2, and variance 3 to obtain the unreliable value of the hypothetical point;

5. Sort each hypothetical point according to the value, and exclude those hypothetical points with higher unreliable values.

The above steps are repeated after each measurement until it converges to the range permitted by the user.

At the same time, in order to be able to adapt to the sudden change of the environmental factor KE, the track record should be deleted while the track record is dynamically updated, so the record length S should remain unchanged. The setting of the record length should be able to meet the accuracy requirement of fitting the environmental curve. When the record length reaches S, it is necessary to forget the old measurement data when recording new measurement data.

Any common fitting method can be used for fitting, and the least square method is used for fitting in this example. The threshold for deleting hypothetical points is set by the user according to the actual situation. For example, it can be set to a specific unreliable value, and the hypothetical point will be discarded when it is higher than this value. In this embodiment, it is set to a percentage data of 20%, and the 20% hypothetical points with the highest unreliable value are discarded every time. Using this threshold can achieve convergence simply and quickly.

The end condition of the convergence step can also be set by the user according to the actual situation. For example, for some systems, it is required that the variance of the fit cannot be higher than a preset threshold. In this embodiment, the number of remaining points is used as the condition for ending the convergence, and when the remaining points are less than 10, the convergence process ends.

When the point set converges within a certain range, the fitting result of the point set may remain oscillating within a certain range. This is mainly because the initial hypothetical point density cannot match the actual point well in the process of assuming points.

In order to improve the positioning accuracy on this basis, this embodiment further diverges the point set. Divergence is mainly to re-increase the number of simulation points within the range of the current hypothetical points and supplement the set of points. First select those more excellent simulation points, and then supplement the points around these points. The supplement of points should not be carried out according to the uniform distribution plan, because in the process of initial point collection, we do not know the general position of the points. In order not to miss, so spread evenly. But at this time, we already know the general location of the points, that is, around these excellent simulation points, so we will use these simulation points as the center to scatter close-in and far-away points. In this embodiment, a Gaussian random number method is used, centering on these reserved simulation points, hypothetical points are distributed within the possible activity range of the device. Add these newly created simulation points to the point set, and then continue to perform the convergence and divergence process of the points until the variance meets the system requirements.

This method can realize effective positioning in an indoor environment, and can make up for the practical difficulty of achieving accurate device positioning in the current environment of large buildings, such as stores and gymnasiums.

For example, the Eurasian store in Changchun, China has an indoor business area of 520,000 square meters. It is very difficult to know whether employees are on duty. The positioning model constructed by this method can effectively determine the location of employees, indirectly increase the attendance rate of employees, and improve corporate profits.

It should be understood that this embodiment is only a specific example of the implementation of the present invention, and should not limit the protection scope of the present invention. Without departing from the spirit and scope of the present invention, equivalent modifications or changes to the above contents shall be included in the scope of protection claimed by the present invention.

Claims (4)

1. A mobile device positioning method based on WIFI, comprising the following steps: 1. In the range where the mobile device may move, spread the hypothetical points evenly to form a set of hypothetical points; within this range, there are N WIFI sources; 2. According to the time interval set by the user, the mobile device is regularly monitored. The data measured each time includes the measurement time, the speed of the device, and the signal strength of each WIFI source received; after each measurement, through the device speed and measurement time, Obtain the relative movement distance vector of the device, update the current position of each hypothetical point in the hypothetical point set accordingly, and record the position information of the hypothetical point and the received signal strength of each WIFI source after each update; 3. When the record length, that is, the number of measurements reaches the length S preset by the user, execute the following steps in a loop until the convergence end condition is reached. The convergence end condition is that the variance of the fit cannot be higher than the preset threshold or the number of retained points is less than Default value: 1) For each hypothetical point, each position data on the recorded track is fitted with the signal strength data of WIFI source 1 at this position to form a fitting function and obtain a variance of 1; 2) Fit each position data on the hypothetical point record track with the signal strength data of WIFI source i at this position to form a fitting function and obtain variance i; where i=1,2,...,N; 3) add variance 1, variance 2, ..., variance N to obtain the unreliable value of the hypothetical point; 4) Calculate the unreliable value of each hypothetical point, sort each hypothetical point according to the unreliable value, and delete the hypothetical point with high unreliable value according to the threshold value preset by the user; 5) For the retained hypothetical points, record the latest measurement data and discard the oldest measurement data. 2. A kind of WIFI-based mobile device location method according to claim 1, is characterized in that, after step 3 is converged to point set, also comprises the step that point set is diverged, with the simulation that retains in step 3 Points as the center, within the range where the equipment may move, close-in and far-out scattered points are formed to form a new hypothetical point set, and the convergence process described in step 3 is performed on this set until the system requirements are met. 3. A kind of WIFI-based mobile device location method according to claim 1 or 2, is characterized in that, the speed of mobile device is obtained by gyroscope, accelerometer and compass that mobile device carries, method is: 1) Make the device static; 2) When the device starts to move, use the accelerometer to obtain the acceleration in various directions at a certain moment; 3) Utilize the gravity value in the accelerometer to judge the attitude of the device and project the acceleration of the device on the horizontal direction based on the device; 4) Using the north-pointing function of the electronic compass, the acceleration based on the horizontal direction of the device is projected into the horizontal acceleration based on the true north direction; 5) Continue to record the acceleration information at the next moment; 6) Integrate the acceleration for a certain period of time to obtain the device velocity at the end of the time. 4. A WIFI-based mobile device positioning method according to claim 1 or 2, characterized in that the threshold in step 3 is set to a percentage data of 20%, and the 20% assumption with the highest unreliable value is discarded every time point.