A WLAN signal-based indoor positioning method, device, medium and equipment

Abstract

This invention discloses an indoor positioning method, apparatus, medium, and device based on WLAN signals. The method includes: acquiring multiple reference points with known locations to obtain a first set of reference points; randomly selecting several reference points from the first set of reference points to obtain a second set of reference points; measuring the signal strength of the WLAN signal at each reference point in the second set of reference points; and obtaining a radio frequency fingerprint database based on the measured signal strength and the location of the corresponding reference points; constructing a transformation matrix based on the first set of reference points and an observation matrix based on the second set of reference points; expanding the radio frequency fingerprint database based on the transformation matrix and the observation matrix to obtain an expanded radio frequency fingerprint database; and using the expanded radio frequency fingerprint database to locate a terminal to be positioned, thereby obtaining the location of the terminal. This invention enables the construction of a more accurate indoor positioning system.

Description

Technical Field

[0001] This invention relates to the field of indoor positioning technology, and in particular to an indoor positioning method, apparatus, medium and device based on WLAN signals. Background Technology

[0002] In existing technologies, indoor positioning technology based on WLAN (Wireless Local Area Network) signals usually requires measuring a sufficient number of radio frequency signal features and the one-to-one correspondence between them and reference points, and then constructing a radio frequency fingerprint indoor positioning system based on this. The terminal to be positioned measures the radio frequency signal features of its own location, and searches for the reference point location information with the highest similarity radio frequency signal features in the radio frequency fingerprint indoor positioning system, which is used as the positioning result of the terminal to be positioned.

[0003] It is evident that in the construction of existing RF fingerprint indoor positioning systems, it is necessary to measure a large number of RF signal features and the one-to-one correspondence between reference points. However, manual measurement makes it difficult to accurately measure the signals of very dense and different reference points, resulting in an inaccurate RF fingerprint indoor positioning system. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention proposes an indoor positioning method, apparatus, medium, and device based on WLAN signals. By expanding the radio frequency fingerprint database according to the transformation matrix and the observation matrix, an expanded radio frequency fingerprint database is obtained. This allows the construction of the entire expanded radio frequency fingerprint database to be completed by measuring the signal strength of a selected subset of reference points, and enables more accurate indoor positioning based on the expanded radio frequency fingerprint database.

[0005] To achieve the above objectives, embodiments of the present invention provide an indoor positioning method based on WLAN signals, comprising:

[0006] Obtain reference points at multiple known locations to form a first set of reference points, and randomly select several reference points from the first set of reference points to form a second set of reference points;

[0007] The signal strength of the WLAN signal at each reference point in the second set of reference points is measured, and an RF fingerprint database is obtained based on the measured signal strength and the location of the corresponding reference point.

[0008] Based on the first set of reference points, construct a transformation matrix, and based on the second set of reference points, construct an observation matrix;

[0009] The radio frequency fingerprint database is expanded based on the transformation matrix and the observation matrix to obtain an expanded radio frequency fingerprint database.

[0010] The extended radio frequency fingerprint database is used to locate the terminal to be located, and the location of the terminal to be located is obtained.

[0011] Furthermore, the step of obtaining multiple reference points at known locations to obtain a first set of reference points, and randomly selecting several reference points from the first set of reference points to obtain a second set of reference points, includes: obtaining N reference points at known locations and their corresponding coordinates to obtain a first set of reference points; and selecting M reference points from the first set of reference points to obtain a second set of reference points; wherein M satisfies the following equations (1) and (2):

[0012] 0.5≤λ≤1 (1)

[0013] λ(log 10 (N)) 6 ≤M≤N (2)

[0014] Where λ is a preset constant.

[0015] Furthermore, the measurement of the signal strength of the WLAN signal at each reference point in the second set of reference points, and the acquisition of the radio frequency fingerprint database based on the measured signal strength and the location of the corresponding reference point, includes: measuring the signal strength of signals emitted by multiple WLAN signal sources at each reference point in the second set of reference points, wherein the signal strength is stored in vector form as shown in the following equation (3):

[0016] r i =[RSS i1 RSS i2 ,...,RSS iM ] T (3)

[0017] Among them, RSS iM Let r be the signal strength emitted by the i-th WLAN signal source measured at the M-th reference point in the second set of reference points. i It is the vector form of the signal strength emitted by the i-th WLAN signal source, measured at each reference point of the second set of reference points;

[0018] Based on the signal strength and the location of the corresponding reference point, the radio frequency fingerprint database is obtained as follows (4):

[0019]

[0020] Among them, (x M y M ) represents the coordinates of the Mth reference point in the second set of reference points, and RF represents the radio frequency fingerprint database.

[0021] Furthermore, the step of constructing a transformation matrix based on the first set of reference points and an observation matrix based on the second set of reference points includes: assigning binary numbers to each reference point in the first set of reference points to obtain a binary vector of the corresponding reference point; and defining a feature mapping based on the first set of reference points as shown in equation (5):

[0022] t α (x)=(-1) (α,x) (5)

[0023] Where α is the binary vector of one reference point in the first set of reference points, x is the binary vector of another reference point in the first set of reference points, (α,x) is the inner product of α and x, and t α (x) is the feature mapping;

[0024] Based on the first column vectors obtained by performing the feature mapping operation on each reference point in the first reference point set, a transformation matrix is ​​constructed; wherein, the transformation matrix is ​​as follows (6):

[0025]

[0026] Among them, A G Let G be the transformation matrix, G be the first set of reference points, N be the number of reference points in the first set of reference points, k be the binary vector of the kth reference point, and d be the binary vector of the kth reference point. N-1 The binary vector of the Nth reference point;

[0027] Based on the one-to-one corresponding second column vectors obtained after each reference point in the second reference point set has undergone the feature mapping operation, an observation matrix is ​​constructed; wherein, the observation matrix is ​​as follows (7):

[0028]

[0029] Among them, A LFP Let LFP be the observation matrix, M be the second set of reference points, k be the number of reference points in the second set of reference points, and l be the binary vector of the kth reference point. M Let be the binary vector of the Mth reference point in the second set of reference points.

[0030] Furthermore, the step of expanding the radio frequency fingerprint database based on the transformation matrix and the observation matrix to obtain an expanded radio frequency fingerprint database includes: optimizing the signal intensity measured at each reference point in the second set of reference points based on the observation matrix to obtain several candidate signal intensities; expanding the multiple candidate signal intensities based on the transformation matrix to obtain corresponding expanded signal intensities; wherein, the function expression of the expanded signal intensities is as follows (8):

[0031]

[0032] Among them, A G Let be the transformation matrix. α represents the extended signal strength corresponding to the i-th WLAN signal source. i The candidate signal strength corresponding to the i-th WLAN signal source;

[0033] An extended radio frequency fingerprint database is obtained based on the extended signal strength and the coordinates of the reference points in the first set of reference points; wherein, the extended radio frequency fingerprint database is as follows (9):

[0034]

[0035] Where (x, y) are the coordinates of the reference points in the first set of reference points, RF ex To expand the radio frequency fingerprint database.

[0036] Furthermore, the step of using the extended radio frequency fingerprint database to locate the terminal to be located and obtain the location of the terminal to be located includes: measuring the signal strength to be located at the terminal to be located; searching the extended radio frequency fingerprint database for the p extended signal strengths with the highest similarity to the signal strength to be located and the coordinates of the corresponding reference points; and calculating the location of the terminal to be located based on the coordinates of the reference points corresponding to the p extended signal strengths found.

[0037] Furthermore, the step of searching for the p extended signal strengths and their corresponding reference point coordinates that have the highest similarity to the signal strength to be located from the extended radio frequency fingerprint database includes: searching for the p extended signal strengths and their corresponding reference point coordinates that have the highest similarity to the signal strength to be located from the extended radio frequency fingerprint database using a nearest neighbor algorithm.

[0038] This invention also provides an indoor positioning device based on WLAN signals, comprising:

[0039] The reference point acquisition module is used to acquire reference points at multiple known locations to obtain a first set of reference points, and to randomly select several reference points from the first set of reference points to obtain a second set of reference points.

[0040] The database construction module is used to measure the signal strength of WLAN signals at each reference point in the second set of reference points, and to obtain an RF fingerprint database based on the measured signal strength and the location of the corresponding reference point.

[0041] A matrix construction module is used to construct a transformation matrix based on the first set of reference points and an observation matrix based on the second set of reference points.

[0042] The database expansion module is used to expand the radio frequency fingerprint database according to the transformation matrix and the observation matrix to obtain an expanded radio frequency fingerprint database.

[0043] The positioning module is used to locate the terminal to be located using the extended radio frequency fingerprint database to obtain the positioning location of the terminal to be located.

[0044] This invention also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the WLAN signal-based indoor positioning method as described in any of the preceding embodiments.

[0045] This invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executed by the processor, wherein the processor executes the program to implement the steps of the WLAN signal-based indoor positioning method as described in any of the preceding embodiments.

[0046] The present invention has the following beneficial effects:

[0047] In this embodiment of the invention, a first set of reference points is obtained by acquiring multiple reference points at known locations. A second set of reference points is then obtained by randomly selecting several reference points from the first set. The signal strength of the WLAN signal at each reference point in the second set is measured, and an RF fingerprint database is obtained based on the measured signal strength and the corresponding reference point location. A transformation matrix is ​​constructed based on the first set of reference points, and an observation matrix is ​​constructed based on the second set of reference points. The RF fingerprint database is then expanded based on the transformation matrix and the observation matrix to obtain an expanded RF fingerprint database. The expanded RF fingerprint database is used to locate the terminal to be located, thus obtaining the location of the terminal. This embodiment of the invention expands the RF fingerprint database based on the transformation matrix and the observation matrix to obtain an expanded RF fingerprint database. This makes the WLAN signal strength sparse under the action of the transformation matrix. That is, only the signal strength of a selected subset of reference points needs to be measured, and the signal strength of this subset of reference points can be expanded through the action of the transformation matrix. The expanded signal strength is then correlated with the coordinates of all reference points, thereby obtaining a sufficiently dense signal strength corresponding to reference points. This allows for the construction of a more accurate WLAN signal-based indoor positioning system. Attached Figure Description

[0048] Figure 1 This is a flowchart illustrating an embodiment of an indoor positioning method based on WLAN signals provided by the present invention;

[0049] Figure 2 This is a schematic diagram of an embodiment of an indoor positioning device based on WLAN signals provided by the present invention. Detailed Implementation

[0050] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0051] See Figure 1 This is a flowchart illustrating an embodiment of the indoor positioning method based on WLAN signals provided by the present invention. The method includes steps S1 to S5, as detailed below:

[0052] S1, obtain multiple reference points with known locations to obtain a first set of reference points, and randomly select several reference points from the first set of reference points to obtain a second set of reference points;

[0053] Specifically, the multiple reference points obtained are evenly distributed on the map.

[0054] Specifically, a second set of reference points is obtained by uniformly and randomly selecting several reference points from the first set of reference points.

[0055] Preferably, the step of obtaining multiple reference points at known locations to obtain a first set of reference points, and randomly selecting several reference points from the first set of reference points to obtain a second set of reference points, includes: obtaining N reference points at known locations and their corresponding coordinates to obtain a first set of reference points; and selecting M reference points from the first set of reference points to obtain a second set of reference points; wherein M satisfies the following equations (1) and (2):

[0056] 0.5≤λ≤1 (1)

[0057] λ(log 10 (N)) 6 ≤M≤N (2)

[0058] Where λ is a preset constant.

[0059] Specifically, λ is a preset constant, usually taken as λ = 1.

[0060] S2, Measure the signal strength of the WLAN signal at each reference point in the second set of reference points, and obtain the radio frequency fingerprint database based on the measured signal strength and the location of the corresponding reference point;

[0061] Preferably, the step of measuring the signal strength of the WLAN signal at each reference point in the second set of reference points, and obtaining the radio frequency fingerprint database based on the measured signal strength and the location of the corresponding reference point, includes: measuring the signal strength of signals emitted by multiple WLAN signal sources at each reference point in the second set of reference points, wherein the signal strength is stored in vector form as shown in the following formula (3):

[0062] r i =[RSS i1 RSS i2 ,...,RSS iM ] T (3)

[0063] Among them, RSS iM Let r be the signal strength emitted by the i-th WLAN signal source measured at the M-th reference point in the second set of reference points. i It is the vector form of the signal strength emitted by the i-th WLAN signal source, measured at each reference point of the second set of reference points;

[0064] Based on the signal strength and the location of the corresponding reference point, the radio frequency fingerprint database is obtained as follows (4):

[0065]

[0066] Among them, (x M y M ) represents the coordinates of the Mth reference point in the second set of reference points, and RF represents the radio frequency fingerprint database.

[0067] Specifically, the signal strength of the WLAN signal at each reference point in the second set of reference points is measured, and an RF fingerprint database is constructed based on the measured signal strength and the specific location coordinates of the corresponding reference points. This database is an initial, unexpanded database, in which only the signal strength at a small number of reference points needs to be measured to complete the construction of the database, thereby reducing the workload of manual measurement. Furthermore, since the locations of the manually measured reference points are relatively sparse, the error of the signal strength obtained by manual measurement at different reference points is smaller. Thus, the embodiments of the present invention can construct a more accurate RF fingerprint database.

[0068] Specifically, the signal strength measured in this embodiment of the invention is the signal strength of the WLAN signal, and WLAN signal sources have been widely deployed in indoor scenarios, so the cost is relatively low.

[0069] S3, construct a transformation matrix based on the first set of reference points, and construct an observation matrix based on the second set of reference points;

[0070] Preferably, the step of constructing a transformation matrix based on the first set of reference points and constructing an observation matrix based on the second set of reference points includes: assigning binary numbers to each reference point in the first set of reference points to obtain a binary vector of the corresponding reference point; and defining a feature mapping based on the first set of reference points as shown in equation (5):

[0071] t α (x)=(-1) (α,x) (5)

[0072] Where α is the binary vector of one reference point in the first set of reference points, x is the binary vector of another reference point in the first set of reference points, (α,x) is the inner product of α and x, and t α (x) is the feature mapping;

[0073] Based on the first column vectors obtained by performing the feature mapping operation on each reference point in the first reference point set, a transformation matrix is ​​constructed; wherein, the transformation matrix is ​​as follows (6):

[0074]

[0075] Among them, A GLet G be the transformation matrix, G be the first set of reference points, N be the number of reference points in the first set of reference points, k be the binary vector of the kth reference point, and d be the binary vector of the kth reference point. N-1 The binary vector of the Nth reference point;

[0076] Based on the one-to-one corresponding second column vectors obtained after each reference point in the second reference point set has undergone the feature mapping operation, an observation matrix is ​​constructed; wherein, the observation matrix is ​​as follows (7):

[0077]

[0078] Among them, A LFP Let LFP be the observation matrix, M be the second set of reference points, k be the number of reference points in the second set of reference points, and l be the binary vector of the kth reference point. M Let be the binary vector of the Mth reference point in the second set of reference points.

[0079] Specifically, each reference point in the first set of reference points is assigned a binary number to obtain a binary vector for that reference point. In one specific embodiment, the reference points in the first set are first vectorized and arranged in a top-down or left-to-right order. Then, each reference point is numbered using an n-bit binary code according to the arrangement order. After numbering, a corresponding binary vector is constructed based on the numerical value of each reference point's number. For example, if the first reference point is 00…01 and the second reference point is 0…010, then the binary vector of the first reference point is [0,0,……,0,1]. T The binary vector of the second reference point is [0,0,……,1,0]. T Then, the transformation matrix and observation matrix are constructed using binary vectors. After the operation of feature mapping, the binary vectors with all elements of 0 or 1 have all elements of 1 or -1. Thus, the transformation matrix and observation matrix in this embodiment of the invention also have all elements of 1 or -1, which is simpler in calculation and easier to implement in hardware.

[0080] Specifically, the feature map (5) is an additive group [GF(2)]. n In constructing the above additive group, binary vectors are regarded as elements and binary addition of vectors are regarded as addition operation rules. Furthermore, the feature mapping (5) is a linear transformation defined on the real number field. Compared with the Fourier transform, it does not generate real and imaginary parts during the operation process. Therefore, the operation result can be obtained without further splicing the real and imaginary parts. Thus, the calculation of the embodiment of the present invention is simpler and has smaller error.

[0081] Specifically, the constructed transformation matrix A G It is a symmetric matrix that satisfies And matrix A G It is an N-order unitary matrix, that is, it satisfies Here I N It is an N-order identity matrix.

[0082] S4. Based on the transformation matrix and the observation matrix, the radio frequency fingerprint database is expanded to obtain an expanded radio frequency fingerprint database.

[0083] Preferably, the step of expanding the radio frequency fingerprint database according to the transformation matrix and the observation matrix to obtain an expanded radio frequency fingerprint database includes: optimizing the signal intensity measured at each reference point in the second set of reference points according to the observation matrix to obtain several candidate signal intensities; expanding the multiple candidate signal intensities according to the transformation matrix to obtain corresponding expanded signal intensities; wherein, the function expression of the expanded signal intensities is as follows (8):

[0084]

[0085] Among them, A G Let be the transformation matrix. For the extended signal strength corresponding to the i-th WLAN signal source, a i The candidate signal strength corresponding to the i-th WLAN signal source;

[0086] An extended radio frequency fingerprint database is obtained based on the extended signal strength and the coordinates of the reference points in the first set of reference points; wherein, the extended radio frequency fingerprint database is as follows (9):

[0087]

[0088] Where (x, y) are the coordinates of the reference points in the first set of reference points, RF ex To expand the radio frequency fingerprint database.

[0089] Specifically, the signal strength measured at each reference point in the second set of reference points is optimized and solved according to the observation matrix to obtain several candidate signal strengths;

[0090] The first embodiment of optimizing the calculation formula in the above steps is as follows: (10)

[0091]

[0092] Among them, A LFP Let α be the observation matrix. i For the candidate signal strength corresponding to the i-th WLAN signal source, r iIt is the vector form of the signal strength emitted by the i-th WLAN signal source, measured at each reference point of the second set of reference points;

[0093] The second embodiment is as follows (11):

[0094]

[0095] Among them, A LFP Let α be the observation matrix. i For the candidate signal strength corresponding to the i-th WLAN signal source, r i η is the vector form of the signal strength emitted by the i-th WLAN signal source measured at each reference point of the second set of reference points, where η is the noise threshold.

[0096] The third embodiment is as follows (12):

[0097]

[0098] Among them, A LFP Let α be the observation matrix. i For the candidate signal strength corresponding to the i-th WLAN signal source, r i λ is the vector form of the signal strength emitted by the i-th WLAN signal source measured at each reference point of the second set of reference points, where λ is a preset constant.

[0099] The fourth embodiment is as follows (13):

[0100]

[0101] Among them, A LFP Let α be the observation matrix. i For the candidate signal strength corresponding to the i-th WLAN signal source, r i τ is the vector form of the signal strength emitted by the i-th WLAN signal source, measured at each reference point in the second set of reference points, where τ is a preset constant.

[0102] Specifically, the radio frequency fingerprint database is expanded based on the transformation matrix and the observation matrix to obtain an expanded radio frequency fingerprint database. In fact, the signal strength of the manually measured WLAN signal is expanded using the transformation matrix and the observation matrix, so that the signal strength has sparsity under the action of the transformation matrix. Then, the coordinates corresponding to the expanded signal strength are assigned, thereby obtaining the coordinates of sufficiently dense reference points and their corresponding signal strengths. That is, it combines the algorithm idea of ​​compressed sensing, and reconstructs the complete signal by measuring discrete samples of the signal, which can build a more accurate indoor positioning system based on WLAN signal.

[0103] S5, the extended radio frequency fingerprint database is used to locate the terminal to be located, and the location of the terminal to be located is obtained.

[0104] Preferably, the step of using the extended RF fingerprint database to locate the terminal to be located and obtain the location of the terminal to be located includes: measuring the signal strength to be located at the terminal to be located; searching the extended RF fingerprint database for the p extended signal strengths with the highest similarity to the signal strength to be located and the coordinates of the corresponding reference points; and calculating the location of the terminal to be located based on the coordinates of the reference points corresponding to the p extended signal strengths found.

[0105] Preferably, the step of finding the p extended signal strengths and their corresponding reference point coordinates that have the highest similarity to the signal strength to be located from the extended RF fingerprint database includes: finding the p extended signal strengths and their corresponding reference point coordinates that have the highest similarity to the signal strength to be located from the extended RF fingerprint database using a nearest neighbor algorithm.

[0106] Specifically, the system finds the p extended signal strengths with the highest similarity to the signal strength to be located and the coordinates of their corresponding reference points. Then, it calculates the average value of the coordinates of the reference points corresponding to the p extended signal strengths. This average value is used as the location coordinates of the terminal to be located. Since the reference points with corresponding signal strength values ​​recorded in the extended RF fingerprint database are sufficiently dense, the error between the coordinates of the p reference points and the actual location coordinates of the terminal to be located is very small. Therefore, the average value of the calculated coordinates of the p reference points is accurate enough as the location coordinates of the terminal to be located.

[0107] Specifically, a concrete implementation of the nearest neighbor algorithm is the KNN (K-Nearest Neighbor) method.

[0108] Accordingly, embodiments of the present invention also provide an indoor positioning device based on WLAN signals, used to implement all the processes of the indoor positioning method based on WLAN signals provided in the above embodiments.

[0109] See Figure 2 This is a schematic diagram of an embodiment of the indoor positioning device based on WLAN signal provided by the present invention.

[0110] An indoor positioning device based on WLAN signal is provided in this embodiment of the invention, comprising:

[0111] The reference point acquisition module 101 is used to acquire reference points at multiple known locations to obtain a first set of reference points, and to randomly select several reference points from the first set of reference points to obtain a second set of reference points.

[0112] Specifically, the multiple reference points obtained are evenly distributed on the map.

[0113] Specifically, a second set of reference points is obtained by uniformly and randomly selecting several reference points from the first set of reference points.

[0114] Preferably, the reference point acquisition module 101 further includes: acquiring N reference points at known locations and their corresponding coordinates to obtain a first reference point set; and extracting M reference points from the first reference point set to obtain a second reference point set; wherein M satisfies the following equations (1) and (2):

[0115] 0.5≤λ≤1 (1)

[0116] λ(log 10 (N)) 6 ≤M≤N (2)

[0117] Where λ is a preset constant.

[0118] Specifically, λ is a preset constant, usually taken as λ = 1.

[0119] The database construction module 102 is used to measure the signal strength of WLAN signals at each reference point in the second set of reference points, and to obtain an RF fingerprint database based on the measured signal strength and the location of the corresponding reference point.

[0120] Preferably, the database construction module 102 further includes: measuring the signal strength of signals emitted by multiple WLAN signal sources at each reference point of the second set of reference points, wherein the signal strength is stored in vector form as shown in equation (3):

[0121] r i =[RSS i1 RSS i2 ,...,RSS iM ] T (3)

[0122] Among them, RSS iM Let r be the signal strength emitted by the i-th WLAN signal source measured at the M-th reference point in the second set of reference points. i It is the vector form of the signal strength emitted by the i-th WLAN signal source, measured at each reference point of the second set of reference points;

[0123] Based on the signal strength and the location of the corresponding reference point, the radio frequency fingerprint database is obtained as follows (4):

[0124]

[0125] Among them, (x M y M ) represents the coordinates of the Mth reference point in the second set of reference points, and RF represents the radio frequency fingerprint database.

[0126] Specifically, the signal strength of the WLAN signal at each reference point in the second set of reference points is measured, and an RF fingerprint database is constructed based on the measured signal strength and the specific location coordinates of the corresponding reference points. This database is an initial, unexpanded database, in which only the signal strength at a small number of reference points needs to be measured to complete the construction of the database, thereby reducing the workload of manual measurement. Furthermore, since the locations of the manually measured reference points are relatively sparse, the error of the signal strength obtained by manual measurement at different reference points is smaller. Thus, the embodiments of the present invention can construct a more accurate RF fingerprint database.

[0127] Specifically, the signal strength measured in this embodiment of the invention is the signal strength of the WLAN signal, and WLAN signal sources have been widely deployed in indoor scenarios, so the cost is relatively low.

[0128] The matrix construction module 103 is used to construct a transformation matrix based on the first set of reference points and to construct an observation matrix based on the second set of reference points.

[0129] Preferably, the matrix construction module 103 further includes: assigning binary numbers to each reference point in the first reference point set to obtain a binary vector of the corresponding reference point; and defining a feature mapping based on the first reference point set as shown in equation (5):

[0130] t α (x)=(-1) (α,x) (5)

[0131] Where α is the binary vector of one reference point in the first set of reference points, x is the binary vector of another reference point in the first set of reference points, (α,x) is the inner product of α and x, and t α (x) is the feature mapping;

[0132] Based on the first column vectors obtained by performing the feature mapping operation on each reference point in the first reference point set, a transformation matrix is ​​constructed; wherein, the transformation matrix is ​​as follows (6):

[0133]

[0134] Among them, A G Let G be the transformation matrix, G be the first set of reference points, N be the number of reference points in the first set of reference points, k be the binary vector of the kth reference point, and d be the binary vector of the kth reference point.N-1 The binary vector of the Nth reference point;

[0135] Based on the one-to-one corresponding second column vectors obtained after each reference point in the second reference point set has undergone the feature mapping operation, an observation matrix is ​​constructed; wherein, the observation matrix is ​​as follows (7):

[0136]

[0137] Among them, A LFP Let LFP be the observation matrix, M be the second set of reference points, k be the number of reference points in the second set of reference points, and l be the binary vector of the kth reference point. M Let be the binary vector of the Mth reference point in the second set of reference points.

[0138] Specifically, each reference point in the first set of reference points is assigned a binary number to obtain a binary vector for that reference point. In one specific embodiment, the reference points in the first set are first vectorized and arranged in a top-down or left-to-right order. Then, each reference point is numbered using an n-bit binary code according to the arrangement order. After numbering, a corresponding binary vector is constructed based on the numerical value of each reference point's number. For example, if the first reference point is 00…01 and the second reference point is 0…010, then the binary vector of the first reference point is [0,0,……,0,1]. T The binary vector of the second reference point is [0,0,……,1,0]. T Then, the transformation matrix and observation matrix are constructed using binary vectors. After the operation of feature mapping, the binary vectors with all elements of 0 or 1 have all elements of 1 or -1. Thus, the transformation matrix and observation matrix in this embodiment of the invention also have all elements of 1 or -1, which is simpler in calculation and easier to implement in hardware.

[0139] Specifically, the feature map (5) is an additive group [GF(2)]. n In constructing the above additive group, binary vectors are regarded as elements and binary addition of vectors are regarded as addition operation rules. Furthermore, the feature mapping (5) is a linear transformation defined on the real number field. Compared with the Fourier transform, it does not generate real and imaginary parts during the operation process. Therefore, the operation result can be obtained without further splicing the real and imaginary parts. Thus, the calculation of the embodiment of the present invention is simpler and has smaller error.

[0140] Specifically, the constructed transformation matrix A G It is a symmetric matrix that satisfies And matrix A G It is an N-order unitary matrix, that is, it satisfies Here IN It is an N-order identity matrix.

[0141] The database expansion module 104 is used to expand the radio frequency fingerprint database according to the transformation matrix and the observation matrix to obtain an expanded radio frequency fingerprint database.

[0142] Preferably, the database expansion module 104 further includes: optimizing the signal intensity measured at each reference point in the second reference point set according to the observation matrix to obtain several candidate signal intensities; expanding the multiple candidate signal intensities according to the transformation matrix to obtain corresponding expanded signal intensities; wherein, the function expression of the expanded signal intensities is as follows (8):

[0143]

[0144] Among them, A G Let be the transformation matrix. α represents the extended signal strength corresponding to the i-th WLAN signal source. i The candidate signal strength corresponding to the i-th WLAN signal source;

[0145] An extended radio frequency fingerprint database is obtained based on the extended signal strength and the coordinates of the reference points in the first set of reference points; wherein, the extended radio frequency fingerprint database is as follows (9):

[0146]

[0147] Where (x, y) are the coordinates of the reference points in the first set of reference points, RF ex To expand the radio frequency fingerprint database.

[0148] Specifically, the signal strength measured at each reference point in the second set of reference points is optimized and solved according to the observation matrix to obtain several candidate signal strengths;

[0149] The first embodiment of optimizing the calculation formula in the above steps is as follows: (10)

[0150]

[0151] Among them, A LFP Let α be the observation matrix. i For the candidate signal strength corresponding to the i-th WLAN signal source, r i It is the vector form of the signal strength emitted by the i-th WLAN signal source, measured at each reference point of the second set of reference points;

[0152] The second embodiment is as follows (11):

[0153]

[0154] Among them, A LFP Let α be the observation matrix. i For the candidate signal strength corresponding to the i-th WLAN signal source, r i η is the vector form of the signal strength emitted by the i-th WLAN signal source measured at each reference point of the second set of reference points, where η is the noise threshold.

[0155] The third embodiment is as follows (12):

[0156]

[0157] Among them, A LFP Let α be the observation matrix. i For the candidate signal strength corresponding to the i-th WLAN signal source, r i λ is the vector form of the signal strength emitted by the i-th WLAN signal source measured at each reference point of the second set of reference points, where λ is a preset constant.

[0158] The fourth embodiment is as follows (13):

[0159]

[0160] Among them, A LEP Let α be the observation matrix. i For the candidate signal strength corresponding to the i-th WLAN signal source, r i τ is the vector form of the signal strength emitted by the i-th WLAN signal source, measured at each reference point in the second set of reference points, where τ is a preset constant.

[0161] Specifically, the radio frequency fingerprint database is expanded based on the transformation matrix and the observation matrix to obtain an expanded radio frequency fingerprint database. In fact, the signal strength of the manually measured WLAN signal is expanded using the transformation matrix and the observation matrix, so that the signal strength has sparsity under the action of the transformation matrix. Then, the coordinates corresponding to the expanded signal strength are assigned, thereby obtaining the coordinates of sufficiently dense reference points and their corresponding signal strengths. That is, it combines the algorithm idea of ​​compressed sensing, and reconstructs the complete signal by measuring discrete samples of the signal, which can build a more accurate indoor positioning system based on WLAN signal.

[0162] The positioning module 105 is used to locate the terminal to be located using the extended radio frequency fingerprint database to obtain the location of the terminal to be located.

[0163] Preferably, the positioning module 105 further includes: measuring the signal strength to be located at the terminal to be located; searching the extended radio frequency fingerprint database for the p extended signal strengths with the highest similarity to the signal strength to be located and the coordinates of the corresponding reference points; and calculating the positioning location of the terminal to be located based on the coordinates of the reference points corresponding to the p extended signal strengths found.

[0164] Preferably, the positioning module 105 further includes: searching the extended radio frequency fingerprint database for the p extended signal strengths and the coordinates of their corresponding reference points that have the highest similarity to the signal strength to be located, using a nearest neighbor algorithm.

[0165] Specifically, the system finds the p extended signal strengths with the highest similarity to the signal strength to be located and the coordinates of their corresponding reference points. Then, it calculates the average value of the coordinates of the reference points corresponding to the p extended signal strengths. This average value is used as the location coordinates of the terminal to be located. Since the reference points with corresponding signal strength values ​​recorded in the extended RF fingerprint database are sufficiently dense, the error between the coordinates of the p reference points and the actual location coordinates of the terminal to be located is very small. Therefore, the average value of the calculated coordinates of the p reference points is accurate enough as the location coordinates of the terminal to be located.

[0166] Specifically, a concrete implementation of the nearest neighbor algorithm is the KNN (K-Nearest Neighbor) method.

[0167] Furthermore, embodiments of the present invention also provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the WLAN signal-based indoor positioning method as described in any of the preceding claims.

[0168] Furthermore, embodiments of the present invention also provide a computer device, including a memory, a processor, and a computer program stored in the memory and executed by the processor, wherein the processor executes the program to implement the steps of the WLAN signal-based indoor positioning method as described in any of the preceding embodiments.

[0169] Through the above description of the embodiments, those skilled in the art can clearly understand that the present invention can be implemented by means of software plus necessary hardware platforms, and of course, it can also be implemented entirely by hardware. Based on this understanding, all or part of the technical solution of the present invention that contributes to the background art can be embodied in the form of a software product. This computer software product can be stored in a storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in various embodiments or some parts of the embodiments of the present invention.

[0170] In summary, the present invention has the following beneficial effects:

[0171] In this embodiment of the invention, a first set of reference points is obtained by acquiring multiple reference points at known locations. A second set of reference points is then obtained by randomly selecting several reference points from the first set. The signal strength of the WLAN signal at each reference point in the second set is measured, and an RF fingerprint database is obtained based on the measured signal strength and the corresponding reference point location. A transformation matrix is ​​constructed based on the first set of reference points, and an observation matrix is ​​constructed based on the second set of reference points. The RF fingerprint database is then expanded based on the transformation matrix and the observation matrix to obtain an expanded RF fingerprint database. The expanded RF fingerprint database is used to locate the terminal to be located, thus obtaining the location of the terminal. This embodiment of the invention expands the RF fingerprint database based on the transformation matrix and the observation matrix to obtain an expanded RF fingerprint database. This makes the WLAN signal strength sparse under the action of the transformation matrix. That is, only the signal strength of a selected subset of reference points needs to be measured, and the signal strength of this subset of reference points can be expanded through the action of the transformation matrix. The expanded signal strength is then correlated with the coordinates of all reference points, thereby obtaining a sufficiently dense signal strength corresponding to reference points. This allows for the construction of a more accurate WLAN signal-based indoor positioning system.

[0172] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. An indoor positioning method based on WLAN signals, characterized in that, include: Obtain reference points at multiple known locations to form a first set of reference points, and randomly select several reference points from the first set of reference points to form a second set of reference points; The signal strength of the WLAN signal at each reference point in the second set of reference points is measured, and an RF fingerprint database is obtained based on the measured signal strength and the location of the corresponding reference point. Based on the first set of reference points, construct a transformation matrix, and based on the second set of reference points, construct an observation matrix; The radio frequency fingerprint database is expanded based on the transformation matrix and the observation matrix to obtain an expanded radio frequency fingerprint database. The extended radio frequency fingerprint database is used to locate the terminal to be located, and the location of the terminal to be located is obtained. The step of obtaining multiple reference points at known locations to form a first set of reference points, and then randomly selecting several reference points from the first set to form a second set of reference points, includes: Obtain N reference points at known locations and their corresponding coordinates to form the first set of reference points; M reference points are uniformly and randomly selected from the first set of reference points to obtain the second set of reference points; where M satisfies the following equations (1) and (2): （1） （2） in This is a preset constant; The step of constructing a transformation matrix based on the first set of reference points and an observation matrix based on the second set of reference points includes: Each reference point in the first set of reference points is assigned a binary number to obtain the binary vector of the corresponding reference point. Based on the first set of reference points, the feature mapping is defined as follows (5): （5） Where α is the binary vector of one reference point in the first set of reference points, and x is the binary vector of another reference point in the first set of reference points. Let α and x be the inner product. For the feature mapping; Based on the first column vectors obtained by performing the feature mapping operation on each reference point in the first reference point set, a transformation matrix is ​​constructed. The transformation matrix is ​​as follows (6): （6） in, Let G be the transformation matrix, G be the first set of reference points, N be the number of reference points in the first set of reference points, and k be the binary vector of the kth reference point. The binary vector of the Nth reference point; Based on the one-to-one corresponding second column vectors obtained after each reference point in the second reference point set has undergone the feature mapping operation, an observation matrix is ​​constructed. The observation matrix is ​​as follows (7): （7） in, Let LFP be the observation matrix, M be the second set of reference points, and k be the binary vector of the k-th reference point. Let M be the binary vector of the Mth reference point in the second set of reference points; The feature mapping operation is used to map a binary vector to 1 or -1, such that the elements of the transformation matrix and the observation matrix are both 1 or -1.

2. The indoor positioning method based on WLAN signals as described in claim 1, characterized in that, The measurement of WLAN signal strength at each reference point in the second set of reference points, and the generation of an RF fingerprint database based on the measured signal strength and the location of the corresponding reference point, includes: At each reference point in the second set of reference points, the signal strength of signals emitted by multiple WLAN signal sources is measured, and the signal strength is stored in vector form as shown in equation (3) below: （3） in, Let be the signal strength emitted by the i-th WLAN signal source, measured at the M-th reference point in the second set of reference points. It is the vector form of the signal strength emitted by the i-th WLAN signal source, measured at each reference point of the second set of reference points; Based on the signal strength and the location of the corresponding reference point, the radio frequency fingerprint database is obtained as follows (4): （4） in, Let M be the coordinates of the Mth reference point in the second set of reference points, and RF be the radio frequency fingerprint database.

3. The indoor positioning method based on WLAN signals as described in claim 2, characterized in that, The step of expanding the radio frequency fingerprint database based on the transformation matrix and the observation matrix to obtain an expanded radio frequency fingerprint database includes: The signal strength measured at each reference point in the second set of reference points is optimized and solved based on the observation matrix to obtain several candidate signal strengths; The intensity of the candidate signals is extended according to the transformation matrix to obtain the corresponding extended signal intensity; wherein the function expression of the extended signal intensity is as follows (8): （8） in, Let be the transformation matrix. The extended signal strength corresponding to the i-th WLAN signal source, The candidate signal strength corresponding to the i-th WLAN signal source; An extended radio frequency fingerprint database is obtained based on the extended signal strength and the coordinates of the reference points in the first set of reference points; wherein, the extended radio frequency fingerprint database is as follows (9): （9） in, The coordinates of the reference points in the first set of reference points. To expand the radio frequency fingerprint database.

4. The indoor positioning method based on WLAN signals as described in claim 3, characterized in that, The step of using the extended radio frequency fingerprint database to locate the terminal to be located, and obtaining the location of the terminal to be located, includes: Measure the signal strength at the terminal to be located; From the extended radio frequency fingerprint database, find the p extended signal strengths with the highest similarity to the signal strength to be located and the coordinates of the corresponding reference points; The location of the terminal to be located is calculated based on the coordinates of the reference points corresponding to the p extended signal strengths found.

5. The indoor positioning method based on WLAN signals as described in claim 4, characterized in that, The step of searching the extended radio frequency fingerprint database for the p extended signal strengths and their corresponding reference point coordinates that have the highest similarity to the signal strength to be located includes: From the extended RF fingerprint database, the nearest neighbor algorithm is used to find the p extended signal strengths with the highest similarity to the signal strength to be located and the coordinates of the corresponding reference points.

6. An indoor positioning device based on WLAN signals, characterized in that, include: The reference point acquisition module is used to acquire reference points at multiple known locations to obtain a first set of reference points, and to randomly select several reference points from the first set of reference points to obtain a second set of reference points. The database construction module is used to measure the signal strength of WLAN signals at each reference point in the second set of reference points, and to obtain an RF fingerprint database based on the measured signal strength and the location of the corresponding reference point. A matrix construction module is used to construct a transformation matrix based on the first set of reference points and an observation matrix based on the second set of reference points. The database expansion module is used to expand the radio frequency fingerprint database according to the transformation matrix and the observation matrix to obtain an expanded radio frequency fingerprint database. The positioning module is used to locate the terminal to be located using the extended radio frequency fingerprint database to obtain the positioning location of the terminal to be located. The step of obtaining multiple reference points at known locations to form a first set of reference points, and then randomly selecting several reference points from the first set to form a second set of reference points, includes: Obtain N reference points at known locations and their corresponding coordinates to form the first set of reference points; M reference points are uniformly and randomly selected from the first set of reference points to obtain the second set of reference points; where M satisfies the following equations (1) and (2): （1） （2） in This is a preset constant; The step of constructing a transformation matrix based on the first set of reference points and an observation matrix based on the second set of reference points includes: Each reference point in the first set of reference points is assigned a binary number to obtain the binary vector of the corresponding reference point. Based on the first set of reference points, the feature mapping is defined as follows (5): （5） Where α is the binary vector of one reference point in the first set of reference points, and x is the binary vector of another reference point in the first set of reference points. Let α and x be the inner product. For the feature mapping; Based on the first column vectors obtained by performing the feature mapping operation on each reference point in the first reference point set, a transformation matrix is ​​constructed. The transformation matrix is ​​as follows (6): （6） in, Let G be the transformation matrix, G be the first set of reference points, N be the number of reference points in the first set of reference points, and k be the binary vector of the kth reference point. The binary vector of the Nth reference point; Based on the one-to-one corresponding second column vectors obtained after each reference point in the second reference point set has undergone the feature mapping operation, an observation matrix is ​​constructed. The observation matrix is ​​as follows (7): （7） in, Let LFP be the observation matrix, M be the second set of reference points, and k be the binary vector of the k-th reference point. Let M be the binary vector of the Mth reference point in the second set of reference points; The feature mapping operation is used to map a binary vector to 1 or -1, such that the elements of the transformation matrix and the observation matrix are both 1 or -1.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps of the indoor positioning method based on WLAN signals as described in any one of claims 1 to 5.

8. A computer device comprising a memory, a processor, and a computer program stored in the memory and executed by the processor, characterized in that, When the processor executes the program, it implements the steps of the indoor positioning method based on WLAN signal as described in any one of claims 1 to 5.

Images