WO2017185828A1

WO2017185828A1 - Fingerprint positioning method and apparatus

Info

Publication number: WO2017185828A1
Application number: PCT/CN2017/070501
Authority: WO
Inventors: 向平叶; 陈诗军; 叶小仁
Original assignee: 中兴通讯股份有限公司
Priority date: 2016-04-26
Filing date: 2017-01-06
Publication date: 2017-11-02
Also published as: CN107318084A; CN107318084B

Abstract

A fingerprint positioning method and apparatus, the method comprising: implementing Euclidean distance matching processing and similarity matching processing of signal features of a point to be positioned and signal features of a location fingerprinting reference point in a location fingerprinting database in order to determine a plurality of optimal similarity location fingerprinting reference points at the nearest distance to the point to be positioned; and using the plurality of optimal similarity location fingerprinting reference points to calculate the position coordinates of the point to be positioned.

Description

Fingerprint positioning method and device

Technical field

The present disclosure relates to the field of mobile positioning technologies, for example, to a fingerprint positioning method and apparatus.

Background technique

With the development of the mobile Internet, people are increasingly demanding positioning and navigation functions. Outdoor positioning technology represented by Global Positioning System (GPS) and Beidou has been widely used, but in complex indoor or closed environments, such as large waiting rooms, large venues, stadiums, large office buildings, underground mines, etc. The scene, due to the serious attenuation of the signal occlusion, still cannot be located. However, for these complex indoor environments, the communication network needs to meet the hotspot coverage anytime and anywhere, so the communication system base station can be used for indoor positioning.

To achieve mobile positioning, it can be based on gyroscope positioning, triangulation and fingerprint positioning. There is a problem of error accumulation based on the positioning of the gyroscope, which cannot be used for a long time. Positioning systems based on time measurement generally require multiple base stations to be strictly time synchronized and require high-accuracy arrival time measurement of wireless signals. The requirements for base station equipment are relatively high and difficult to implement. Fingerprint positioning does not need to know the location of the base station and the accurate channel model, so it is superior to triangulation in terms of implementation and positioning performance.

Fingerprint positioning refers to testing the received signal strength (RSS) signal of all reference points in the positioning area and extracting the signal characteristics of the RSS signal, and storing the signal feature together with the position coordinates of the corresponding reference point into the location fingerprint database. Then, the same method is used to obtain the signal characteristics of the point to be located, and matching is performed according to a certain matching algorithm and the location fingerprint database, thereby obtaining an estimated position of the point to be located. Common methods include the Nearest neighbor in signal space (NNSS), the K-nearest neighbor algorithm (KNNSS), and the like.

However, whether it is the NNSS algorithm or the KNNSS algorithm, the reference fingerprint is selected by calculating the distance between the signal strength vector of the point to be located and the signal intensity value vector of all sample points in the fingerprint library, and the signal intensity of the fingerprint does not vary linearly with distance. It is difficult to guarantee the quality of the selected reference fingerprint, which in turn affects the accuracy and stability of the positioning result, so that the accuracy of the positioning result is seriously affected.

Summary of the invention

The present disclosure provides a fingerprint positioning method and apparatus, which can eliminate the diversity error of matching results. Improve fingerprint positioning accuracy.

The fingerprint positioning method provided by the present disclosure includes:

Performing Euclidean distance matching processing and similarity matching processing on the signal feature of the point to be located and the signal feature of the position fingerprint reference point in the location fingerprint database to determine a plurality of optimal position fingerprint reference points, and the plurality of optimal position fingerprint reference points Is a plurality of location fingerprint reference points that are closest to the point to be located and have the highest similarity; and

Using the plurality of optimal location fingerprint reference points, calculating position coordinates of the to-be-positioned point.

Optionally, the signal feature is a received signal strength vector, and the determining a plurality of optimal location fingerprint reference points includes:

And performing a Euclidean distance matching process on the received signal strength vector of the to-be-located point and the received signal strength vector of the plurality of position fingerprint reference points in the location fingerprint database to determine N position fingerprint reference points with the smallest Euclidean distance;

Performing similarity matching processing on the received signal strength vector of the point to be located and the received signal strength vector of the N position fingerprint reference points, respectively, and determining the n position fingerprint reference points with the highest similarity as the optimal similarity position fingerprint Reference point, where 2 ≤ n ≤ N, N > 3.

Optionally, the determining the n location fingerprint reference points with the highest similarity as the optimal location fingerprint reference points includes:

Calculating, respectively, a similarity between the received signal strength vector of the to-be-located point and the received signal strength vector of the N-position fingerprint reference points, to obtain N similarity values;

Sorting the N similarity values to determine a highest n similarity values and n position fingerprint reference points corresponding to the n similarities.

Optionally, the similarity calculation is performed using a cosine similarity algorithm or a modified cosine similarity algorithm or a Pearson similarity algorithm.

Optionally, the calculating the location coordinates of the to-be-located point by using the determined multiple optimal location fingerprint reference points includes:

The position coordinates of the point to be located are obtained by weighting the position coordinates of the plurality of optimal position fingerprint reference points.

A non-transitory computer readable storage medium provided in accordance with an embodiment of the present disclosure is stored by a computer executable program that can be used to perform the fingerprint positioning method described above.

A fingerprint positioning apparatus according to an embodiment of the present disclosure includes:

The optimal fingerprint determining module is configured to perform a Euclidean distance matching process and a similarity matching process on the signal feature of the point to be located and the signal feature of the position fingerprint reference point in the location fingerprint database to determine a plurality of optimal position fingerprint reference points, The plurality of optimal location fingerprint reference points are a plurality of location fingerprint reference points that are closest to the to-be-positioned point and have the highest similarity;

The to-be-positioned position determining module is configured to calculate the position coordinates of the to-be-positioned point by using the plurality of optimal position fingerprint reference points.

Optionally, the signal feature is a received signal strength vector, and the optimal fingerprint determining module may be configured to receive the received signal strength vector of the to-be-located point and the fingerprint reference point of each location in the location fingerprint database. The signal strength vector is respectively subjected to the Euclidean distance matching process, and the N position fingerprint reference points with the smallest Euclidean distance are determined, and the received signal strength vector of the to-be-positioned point and the received signal strength vector of the N position fingerprint reference points are respectively performed. The similarity matching processing determines the n position fingerprint reference points with the highest similarity as the optimal position fingerprint reference point, where 2≤n≤N, N>3.

Optionally, the optimal fingerprint determining module may be configured to separately calculate a similarity between the received signal strength vector of the to-be-located point and the received signal strength vector of the N position fingerprint reference points, to obtain N similarity values. And sorting the N similarity values to determine a maximum n similarity values and n position fingerprint reference points corresponding to the n similarity values.

Optionally, the optimal fingerprint determination module may perform a similarity calculation by using a cosine similarity algorithm or a modified cosine similarity algorithm or a Pearson similarity algorithm.

Optionally, the to-be-positioned position determining module may be configured to obtain a position coordinate of the to-be-positioned point by weighting the position coordinates of the plurality of optimal position fingerprint reference points.

An electronic device according to an embodiment of the present disclosure, the electronic device includes one or more processors, a memory, and one or more programs, the one or more programs being stored in a memory when executed by one or more processors When the fingerprint positioning method described above is performed.

A computer program product according to an embodiment of the present disclosure, the computer program product comprising a computer program stored on a non-transitory computer readable storage medium, the computer program comprising program instructions, when the program instructions are executed by a computer, The computer is caused to perform the fingerprint positioning method described above.

By calculating the Euclidean distance and similarity of the signal features of the location to be located and the signal features of the location fingerprint reference point in the location fingerprint database, the location fingerprint reference point that is optimally similar to the location to be located can be selected to eliminate the diversity error of the matching result. Improve positioning accuracy.

DRAWINGS

FIG. 1 is a first flowchart of a fingerprint positioning method according to an embodiment of the present disclosure.

FIG. 2 is a schematic diagram of a first structure of a fingerprint positioning apparatus according to an embodiment of the present disclosure.

FIG. 3 is a second flowchart of a positioning method according to an embodiment of the present disclosure.

FIG. 4 is a flowchart of a positioning method provided by an embodiment of the present disclosure.

FIG. 5 is a schematic plan view of a shopping mall according to an embodiment of the present disclosure.

FIG. 6 is a schematic diagram of a second structure of a fingerprint positioning apparatus according to an embodiment of the present disclosure.

FIG. 7 is a schematic structural diagram of hardware of an electronic device according to an embodiment of the present disclosure.

detailed description

The embodiments of the present disclosure are described in detail below with reference to the accompanying drawings. The features of the following embodiments and embodiments may be combined with each other without conflict.

FIG. 1 is a first flowchart of a fingerprint positioning method according to an embodiment. As shown in FIG. 1, the method may include S110-S120.

In S110, the signal feature of the point to be located and the signal feature of the position fingerprint reference point in the location fingerprint database are subjected to Euclidean distance matching processing and similarity matching processing to determine a plurality of optimal position fingerprint reference points, and the plurality of optimal positions The fingerprint reference point may refer to a plurality of location fingerprint reference points that are closest to the point to be located and have the highest similarity.

The above signal characteristic may be a received signal strength vector (which may be power information).

Optionally, the received signal strength vector of the point to be located and the received signal strength vector of the plurality of position fingerprint reference points in the location fingerprint database are first subjected to Euclidean distance matching processing to determine N position fingerprint reference points with the smallest Euclidean distance. Then, the received signal strength vector of the point to be located and the received signal strength vector of the N position fingerprint reference points are respectively subjected to similarity matching processing, and the first n position fingerprint reference points with the highest similarity are determined as the optimal similarity position fingerprint reference point. .

Optionally, the similarity of the received signal strength vector between the target location point and the N location fingerprint reference points is calculated separately, and a cosine similarity algorithm or a modified cosine similarity algorithm or Pearson may be used. The similarity algorithm obtains N similarities. By sorting the obtained N similarities, the maximum n similarities and n position fingerprint reference points corresponding to n similarities are determined. Wherein, 2≤n≤N, N>3, n and N are positive integers.

Optionally, the location fingerprint database is a pre-established database in an offline situation, the database containing a fingerprint of each location fingerprint reference point (ie, a reference point or a reference fingerprint point or a sample point), the fingerprint including the location fingerprint reference point Location and signal characteristics.

In S120, the determined position coordinates of the point to be located are calculated by using the determined plurality of optimal position fingerprint reference points.

Optionally, the position coordinates of the point to be located are obtained by weighting the determined position coordinates of the plurality of optimal position fingerprint reference points.

It will be understood by those skilled in the art that all or part of the steps of the foregoing embodiments may be implemented by a program to instruct related hardware, and the program may be stored in a computer readable storage medium. The solution in S110-S120 can be implemented. The computer readable storage medium may be a non-transitory computer readable storage medium such as a ROM/RAM, a magnetic disk, an optical disk, or the like.

FIG. 2 is a schematic diagram of a first structure of a fingerprint positioning apparatus according to an embodiment. As shown in FIG. 2, the method may include an optimal fingerprint determination module and a position determining module to be located.

The optimal fingerprint determination module may be configured to perform Euclidean distance matching processing and similarity matching processing on the signal feature of the point to be located and the signal feature of the position fingerprint reference point in the location fingerprint database, and determine a plurality of optimal position fingerprint reference points, the plurality of The optimal location fingerprint reference point may refer to a plurality of location fingerprint reference points that are closest to the point to be located and have the highest similarity.

Alternatively, the signal characteristic may be a received signal strength vector. The optimal fingerprint determination module may be further configured to perform the Euclidean distance matching processing on the received signal strength vector of the point to be located and the received signal strength vector of each position fingerprint reference point in the location fingerprint database to determine the N position fingerprints with the smallest Euclidean distance. The reference point is used, and the received signal strength vector of the point to be located and the received signal strength vector of the N position fingerprint reference points are respectively subjected to similarity matching processing.

Optionally, a cosine similarity algorithm or a modified cosine similarity algorithm or a Pearson similarity algorithm may be used to perform similarity calculation, and N similarities are obtained, and N similarities are sorted to determine the highest n similarities and corresponding n n position fingerprint reference points of similarity, and n position fingerprint reference points are used as the optimal position fingerprint reference points. Wherein, 2≤n≤N, N>3, n and N are positive integers.

a positioning point location determining module, configured to utilize the determined plurality of optimal location fingerprint reference points, Calculate the position coordinates of the point to be positioned.

Optionally, the to-be-positioned position determining module may further be configured to obtain a position coordinate of the to-be-positioned point by weighting the determined position coordinates of the plurality of optimal position fingerprint reference points.

FIG. 3 is a second flowchart of a positioning method according to an embodiment of the present disclosure. As shown in FIG. 3, the method may include S201-S204.

In S201, an offline fingerprint database is constructed.

The offline fingerprint database may refer to a fingerprint database or a location fingerprint database.

In S202, the feature information of the point to be located is matched with the data in the fingerprint library by Euclidean distance. The feature information may refer to a signal feature.

Through the Euclidean distance matching, several reference fingerprint point data with the smallest Euclidean distance are obtained.

In S203, similarity matching is performed on the feature information of the point to be located and the reference fingerprint point data with the smallest Euclidean distance.

Through the similarity matching, several fingerprint points with the largest similarity are obtained.

In S204, a weighted nearest neighbor algorithm is applied to the plurality of fingerprint points with the highest similarity to estimate a final position of the to-be-located point.

The fingerprint positioning method provided by the embodiment of the present disclosure first constructs an offline fingerprint database; secondly, the feature information of the point to be located is matched with the feature information in the fingerprint database to match the Euclidean distance, and the N fingerprints with the smallest Euclidean distance are taken as new Fingerprint library; then using the feature information of the point to be located and the new fingerprint database to solve the similarity, obtain the n fingerprints with the highest similarity, and finally apply the weighting algorithm such as WKNNSS (Weighted KNNSS) to the n fingerprints to obtain the final Position the results. The implementation includes the construction of an offline fingerprint database stage and an online positioning stage, which may include:

Phase 1: Building an offline fingerprint database.

M base stations are arranged in the positioning environment, K reference points are set in the positioning area, the received signal power of each base station is sampled at each reference point, and the reference point position and power information are combined to form a fingerprint, and the i-th fingerprint is expressed as follows :[xi,yi,zi,p _1i ,p _2i ,...,p _Mi ].

Where xi, yi, zi are the position information of the i-th reference point, and p _1i , p _2i , . . . , p _Mi are the user equipment signals of the M base stations in the fingerprint database construction time to the i-th reference point position. The received power, that is, the received power of the signals from the M base stations received by the user equipment at the i-th reference point position at the time of the fingerprint database construction time. The value of i ranges from 1 to K.

Phase 2: Online positioning

Step 1: Find m valid signal strength values from the signal strength vectors R _x =[p ₁ , p ₂ , . . . , p _M ] received by the M base stations at the point to be located, where m is less than M and A positive integer greater than 2.

Step 2: Performing Euclidean distance matching between the signal strength vector received by the positioning point and the corresponding multiple signal strength vectors in the fingerprint library, respectively, by calculating the signal strength vector received by the to-be-located point and the corresponding multiple signals in the fingerprint database. The Euclidean distance of the intensity vector gives the N fingerprints with the smallest Euclidean distance, and N is a positive integer greater than 3.

Calculate the Euclidean distance using the following formula:

Wherein, P _xj represents the signal strength received by the jth base station of the point to be located, P _ji represents the signal strength vector of the jth base station received by the i th reference point, and m represents the effective signal strength number. Qi represents the Euclidean distance of the signal strength of the i-th reference point to the point to be located.

Step 3: The signal strength vector of the N fingerprints obtained in step 2 is sequentially solved with the signal strength vector of the point to be located. The similarity algorithm may select cosine similarity, Pearson correlation coefficient or modified cosine similarity.

Taking cosine similarity as an example, it is calculated by the following formula:

Where P _x represents the signal strength vector of the point to be located, P _i represents the signal strength vector of the _ith fingerprint of the N fingerprints, ||*|| indicates modulo operation, <*> indicates inner product calculation, CosSim (P _x , P _i ) represent the cosine similarity coefficient of the point to be located and the i-th fingerprint point. The larger the cosine similarity coefficient, the greater the correlation between the two.

Step 4: Sort the N cosine similarity coefficients from large to small, and select the n largest values before the cosine similarity coefficient, and the n fingerprints corresponding to the n cosine similarity coefficients are the closest to the to-be-located point. n fingerprint points.

Step 5: Calculate the final positioning result of the above n fingerprint points by using the WKNNSS algorithm. Calculated as follows:

Where x _k , y _k , and z _k are coordinate information of the kth matching fingerprint. The weight w _{k is} obtained by the weighted nearest neighbor method, and the formula is as follows:

Where ε is a very small real constant used to avoid the case where the denominator is zero. Q _k represents the Euclidean distance of the signal strength of the kth reference point to the point to be located.

FIG. 4 is a flowchart of a positioning method according to an embodiment of the present disclosure. As shown in FIG. 4, the method may include S301-S307.

In S301, sample points are set in the fingerprint positioning area.

In S302, the received signal power information of the sample points in the sampling area with respect to each base station.

In S303, the location information of the sample points and the received power information are combined to form a fingerprint library.

In S304, the received signal strength information of the M base stations related to the point to be located is selected.

In S305, the power information of the point to be located and the power information of the plurality of fingerprints in the fingerprint library are matched by Euclidean distance, and N fingerprint points with the smallest Euclidean distance are selected.

In S306, the N fingerprint points selected in the above step are solved with the cosine similarity degree, and the n corresponding fingerprint points whose cosine similarity is positive and smallest are selected from the N fingerprint points.

In S307, the location of the terminal is estimated from a plurality of candidate points using a specific algorithm such as WKNNSS.

The operations performed in S301 to S303 are the steps of the offline fingerprint library construction phase, and the operations performed in S304 to S307 are the steps of the online positioning phase.

The fingerprint method in the related art only selects the reference fingerprint by the signal strength vector size, and it is difficult to ensure the quality of the selected N fingerprints, thereby affecting the accuracy and stability of the positioning result. In particular, when the vector composed of signal strengths obtained by different base stations has symmetry, the matching result of the points to be located in the sample may be different, and it is impossible to distinguish which sample point is the closest distance from the point to be located, and the selected one is selected. Among the N reference fingerprints, there may be fingerprints with large physical position errors, which makes the positioning error fluctuate greatly, the positioning accuracy is seriously affected, and the user experience is poor.

This embodiment describes the technical method provided by the present disclosure in detail in conjunction with FIG. 5.

FIG. 5 is a schematic plan view of a shopping mall according to an embodiment of the present disclosure, as shown in FIG. 5, at a distance of 12 meters. * 6 access points (APs) or base stations are installed in the 60-meter mall.

Embodiment 1. Fingerprint positioning based on intensity and direction

Phase 1: Building an offline fingerprint database

M=6 APs are arranged in a 12m*60m mall, K=50 reference points are arranged in the mall, and the received signal power of each AP is sampled at each reference point, and the reference point position and power information are combined. The fingerprint is formed, and the i-th fingerprint is expressed as follows: [xi, yi, zi, p _1i , p _2i , ..., p _Mi ]. Where xi, yi, zi are the position information of the i-th reference point, and p _1i , p _2i , . . . , p _Mi are signals of the user equipment of the ith reference point position of the six base stations at the time of the fingerprint database construction time. Receive power. The value of i ranges from 1 to 50.

Phase 2: Online positioning

Step 1: Find the m=3 effective signal strength values from the signal strength vector R _x =[p ₁ ,p ₂ ,...,p _M ] received from the point to be located. The effective signal strength gate valve is used by the shopping mall environment. Ok, here it is set to -100dBm.

Step 2: Find the signal strength vector received by the positioning point and the corresponding signal strength vector in the fingerprint library to perform Euclidean distance matching, and obtain the six fingerprints with the smallest Euclidean distance. Calculate the Euclidean distance using the following formula:

Wherein, P _xj represents the signal strength received by the jth AP of the point to be located, P _ji represents the signal strength vector of the jth AP received by the i th reference point, and m represents the number of effective signal strengths. Qi represents the Euclidean distance of the signal strength of the i-th reference point to the point to be located.

Step 3: The signal strength vector of the six fingerprints with the smallest Euclidean distance obtained in step 2 is sequentially solved by the cosine similarity algorithm with the signal strength vector of the point to be located. The cosine similarity is calculated by the following formula:

Where P _x represents the signal strength vector of the point to be located, P _i represents the signal strength vector of the _ith fingerprint of the 6 fingerprints, ||*|| indicates modulo operation, <*> indicates inner product calculation, CosSim ( P _x , P _i ) represents the cosine similarity coefficient of the point to be located and the i-th fingerprint point. The larger the cosine similarity coefficient, the greater the correlation between the two.

Step 4: Sort the six cosine similarity coefficients from large to small, and select the first three large values of the cosine similarity coefficient. The three fingerprints corresponding to the three cosine similarity coefficients are the closest to the to-be-located point. Fingerprint points.

Step 5: Apply the WKNNSS algorithm to the above three fingerprint points to obtain the final positioning result (x, y, z). The calculation formula is as follows:

Where x _k , y _k , z _k are the coordinate information of the kth matching fingerprint. The weight w _k is obtained by the weighted nearest neighbor method, and the formula is as follows:

Where ε is a very small real constant used to avoid the case where the denominator is zero.

Embodiment 2: Fingerprint localization based on optimal similarity

Phase 1: Building an offline fingerprint database

M=6 base stations are arranged in a 12m*60m shopping mall, K=50 reference points are arranged in the shopping mall, and each reference point samples the current reference point to receive the signal power of 6 base stations, and the reference point position Combined with the power information to form a fingerprint, the ith fingerprint is expressed as follows: [xi, yi, zi, p _1i , p _2i , ..., p _Mi ]. Where xi, yi, zi are the position information of the i-th reference point, and p _1i , p _2i , . . . , p _Mi is the signal power of the user equipment received by the user equipment of the ith point. The value of i ranges from 1 to 50.

Phase 2: Online positioning

Step 1: Find the effective signal strength value of the received m=3 base stations from the signal strength vector R _x =[p ₁ , p ₂ , . . . , p _M ] of the 6 base stations received from the user equipment to be located. The effective signal strength gate valve is determined by the mall environment, here set to -98dBm.

Step 3: The signal strength vector of the six fingerprints with the smallest Euclidean distance obtained in step 2 is sequentially solved by the Pearson similarity algorithm with the signal intensity vector of the point to be located. Pearson similarity is as follows Formula calculation:

Where P _x represents the signal strength vector of the point to be located, and P _i represents the signal strength vector of the _ith fingerprint of the 6 fingerprints.

Indicates the average of the signal strength vector with the anchor point,

Indicates the average value of the signal strength vector of the ith fingerprint, ||*|| indicates the modulo operation, <*> indicates the inner product calculation, and Corr(P _x , P _i ) indicates the point to be located and the ith fingerprint point. Pearson similarity coefficient. The greater the Pearson similarity coefficient, the greater the correlation between the two.

Step 4: Sort the six Pearson similarity coefficients from large to small, and select the first three large values of the cosine similarity coefficient. The three fingerprints corresponding to the three cosine similarity coefficients are the closest to the to-be-positioned point. Fingerprint points.

The embodiments of the present disclosure can eliminate the positioning matching diversity error caused by the signal strength vector symmetry measured by different base stations, and improve the positioning accuracy.

Example 3:

According to the fingerprint positioning method provided in the above embodiments, the embodiment of the present disclosure further provides an apparatus for applying the fingerprint positioning method described above.

FIG. 6 is a second schematic structural diagram of a fingerprint positioning apparatus according to an embodiment of the present disclosure. As shown in FIG. 6, the method includes:

Offline fingerprint database building module, which can be called offline fingerprint database module, set as fingerprint database Establishing, deploying M base stations in the positioning environment, setting K reference points in the positioning area, sampling the received signal power of each base station or terminal relative to each base station at each reference point, and combining the reference point position and power information Form a fingerprint.

The receiving selection module is configured to receive the measured data reported by the user equipment and filter out valid data.

The matching module is configured to perform a European distance calculation on the measured data and the fingerprint database data by the positioning server to find the first N fingerprints with the smallest Euclidean distance.

The determining module is configured to perform a similarity calculation on the measured data and the N fingerprint data by the positioning server, and select n optimal similarity fingerprints.

The positioning module is configured to obtain the positioning result of the user equipment by using the WKNNSS method according to the selected optimal similarity fingerprint information.

The receiving selection module, the matching module and the determining module can jointly implement the function of the optimal similarity fingerprint determining module, and the positioning module can implement the function of the position determining module to be located.

The fingerprint location of the mobile network base station can be divided into two methods: uplink fingerprint location and downlink fingerprint location. The uplink fingerprint location refers to the UE transmitting reference signals, and multiple base stations measure the signal power transmitted by the UE, and the fingerprints are pre-stored in the database. The fingerprint is matched for positioning. The downlink fingerprint location refers to that the UE receives and measures the strength of the transmitted signals of multiple base stations, and the fingerprints are matched and positioned by the fingerprints pre-stored in the database.

The present disclosure is based on the combination of the size and direction of the RSS to determine the optimal similarity between the fingerprint and the measured data, and the positioning accuracy is high. Since the wireless channel has symmetry, the present disclosure is applicable to the upper and lower fingerprint positioning and the downlink fingerprint positioning.

By calculating the Euclidean distance and similarity between the user's measured data and the fingerprint data in the fingerprint database, and ensuring the correlation between the reference fingerprint and the point to be located from the data size and direction, the reference that is optimally similar to the point to be located can be selected. The fingerprint can eliminate the diversity error of the matching of the fingerprint point and the point to be located, and improve the positioning accuracy.

FIG. 7 is a schematic diagram of a hardware structure of an electronic device according to an embodiment of the present disclosure. As shown in FIG. 7, the electronic device may include:

A processor 410 and a memory 420; may also include a communications interface 430 and a bus 440.

The processor 410, the memory 420, and the communication interface 430 can be completed by each other through the bus 440. Communication. Communication interface 430 can be used for information transmission. The processor 410 can call the logic instructions in the memory 420 to perform the fingerprint positioning method of the above embodiment.

In addition, the logic instructions in the memory 420 described above may be implemented in the form of a software functional unit and sold or used as a stand-alone product, and may be stored in a storage medium. Based on such understanding, the technical solution of the present disclosure may be embodied in the form of a software product stored in a storage medium, including a plurality of instructions for causing a computer device (which may be a personal computer, a server, or A network device or the like) performs all or part of the steps of the method described in the embodiments of the present disclosure. The foregoing storage medium may be a non-transitory storage medium, including: a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk. A medium that can store program code, or a transitory storage medium.

Finally, it should be understood that those skilled in the art can understand that all or part of the process of implementing the above embodiments can be completed by executing related hardware by a computer program, and the program can be stored in a computer readable storage medium. The program, when executed, may include the flow of an embodiment of the method described above, wherein the computer readable storage medium may be a magnetic disk, an optical disk, a read only memory (ROM), or a random access memory (RAM). Wait.

Industrial applicability

The present disclosure provides a fingerprint positioning method and device, which can eliminate the diversity error of the matching result and improve the positioning accuracy.

Claims

A fingerprint positioning method includes:

Performing Euclidean distance matching processing and similarity matching processing on the signal feature of the point to be located and the signal feature of the position fingerprint reference point in the location fingerprint database to determine a plurality of optimal position fingerprint reference points, and the plurality of optimal position fingerprint reference points Is a plurality of location fingerprint reference points having the highest similarity to the point to be located; and

Using the plurality of optimal location fingerprint reference points, calculating position coordinates of the to-be-positioned point.
The method of claim 1, wherein the signal characteristic is a received signal strength vector, and the determining the plurality of optimal location fingerprint reference points comprises:

Performing Euclidean distance matching processing on the received signal strength vector of the to-be-located point and the received signal strength vector of each position fingerprint reference point in the position fingerprint database to determine N position fingerprint reference points with the smallest Euclidean distance;

Performing similarity matching processing on the received signal strength vector of the point to be located and the received signal strength vector of the N position fingerprint reference points, respectively, and determining n position fingerprint reference points with the highest similarity as the optimal position fingerprint reference point Where 2 ≤ n ≤ N, N > 3.
The method according to claim 2, wherein the determining the n position fingerprint reference points with the highest similarity as the optimal position fingerprint reference points comprises:

Calculating a similarity between the received signal strength vector of the to-be-located point and the received signal strength vector of the N-position fingerprint reference points, to obtain N similarity values;

Sorting the N similarity values to determine a highest n similarity values and n position fingerprint reference points corresponding to the n similarity values.
The method of claim 3, wherein the similarity calculation is performed using a cosine similarity algorithm or a modified cosine similarity algorithm or a Pearson similarity algorithm.
The method according to claim 1, wherein the calculating the position coordinates of the point to be located by using the plurality of optimal position fingerprint reference points comprises:

The position coordinates of the point to be located are obtained by weighting the position coordinates of the plurality of optimal position fingerprint reference points.
A fingerprint positioning device includes:

An optimal fingerprint determination module, configured to set a signal feature of the point to be located and a location fingerprint database Setting a signal feature of the fingerprint reference point to perform Euclidean distance matching processing and similarity matching processing, and determining a plurality of optimal position fingerprint reference points, wherein the plurality of optimal position fingerprint reference points are closest to the to-be-positioned point and similarity The highest number of location fingerprint reference points;

The to-be-positioned position determining module is configured to calculate the position coordinates of the to-be-positioned point by using the plurality of optimal position fingerprint reference points.
The apparatus of claim 6 wherein said signal characteristic is a received signal strength vector, said optimal fingerprint determination module being configured to:

Performing Euclidean distance matching processing on the received signal strength vector of the to-be-located point and the received signal strength vector of each position fingerprint reference point in the position fingerprint database, and determining N position fingerprint reference points with the smallest Euclidean distance, and The received signal strength vector of the point to be located and the received signal strength vector of the N position fingerprint reference points are respectively subjected to similarity matching processing, and the n position fingerprint reference points with the highest similarity are determined as the optimal position fingerprint reference point. Among them, 2 ≤ n ≤ N, N > 3.
The apparatus according to claim 7, wherein the optimal fingerprint determination module is configured to:

Calculating a similarity between the received signal strength vector of the to-be-located point and the received signal strength vector of the N-position fingerprint reference points, respectively, to obtain N similarity values, and sorting the N similarity values to determine the highest n similarity values and n position fingerprint reference points corresponding to the n similarity values.
The apparatus according to claim 8, wherein the optimal fingerprint determination module performs a similarity calculation using a cosine similarity algorithm or a modified cosine similarity algorithm or a Pearson similarity algorithm.
The device according to claim 6, wherein the to-be-positioned position determining module is configured to:

The position coordinates of the point to be located are obtained by weighting the position coordinates of the plurality of optimal position fingerprint reference points.
A non-transitory computer readable storage medium storing computer executable instructions for performing the fingerprint locating method of any of claims 1-5.