WiFi fingerprint positioning method, system and device based on Gaussian process regression and medium

Abstract

The invention relates to the technical field of indoor positioning, in particular to a WiFi fingerprint positioning method, system and device based on Gaussian process regression and a medium, and the method comprises the steps: simulating the overall distribution condition of target indoor WiFi signals in an indoor physical environment of a known target, and dividing an indoor WiFi signal distribution region according to a set threshold value to obtain a plurality of sub-regions; determining a corresponding kernel function for each sub-region so as to construct a Gaussian process regression model; acquiring RSS signals of WiFi of a target indoor environment, inputting the RSS signals into the model, fitting to obtain an RSS signal distribution curve, and constructing a WiFi position fingerprint database according to the distribution curve; and performing similarity calculation on the real-time RSS fingerprints and the position fingerprint database to finally obtain a positioning estimation result of the current position, thereby realizing indoor WiFi accurate positioning.

Description

Technical Field

[0001] The present invention relates to the field of indoor positioning technology, and in particular to a WiFi fingerprint positioning method, system, equipment and medium based on Gaussian process regression. Background Art

[0002] With the rapid development of the Internet of Things and smart homes, indoor positioning technology has a wide range of demands in practical applications. Among them, indoor positioning technology based on WiFi signals has been widely studied and applied due to its advantages such as no need for additional hardware, low cost and flexible deployment. However, in traditional WiFi indoor positioning methods, the construction process of the fingerprint library is cumbersome and time-consuming, which limits its feasibility in large-scale applications.

[0003] Existing WiFi indoor positioning technology mainly collects a large amount of WiFi signal strength data to build a fingerprint library, and then achieves indoor positioning by matching the signal strength of the point to be positioned and combining a certain algorithm. For example, the Chinese invention patent application document with the authorization announcement number CN105338498B discloses a method for constructing a fingerprint library in a WiFi indoor positioning system. This method processes RSSI information data through the quantile elimination method to construct a fingerprint library. However, this method has the following technical problems: the quantile elimination method easily eliminates important data, thereby affecting the accuracy of the fingerprint library; in addition, this method does not consider the real-time update of the fingerprint library, which may lead to a decrease in positioning accuracy.

[0004] The patent application with the authorization announcement number CN106714109B discloses a method for updating a WiFi fingerprint library based on crowdsourced data. The method collects crowdsourced data of indoor environments and performs clustering processing, then dynamically obtains an access point table based on the updated data and calculates the standard fingerprint of each reference point. However, the method has the following technical problems in practical applications: the collection and processing of crowdsourced data may lead to data redundancy, thereby increasing the computational complexity; in addition, the update effect of the fingerprint library may be affected by the quality of crowdsourced data.

[0005] The patent application with the authorization announcement number CN107832834B discloses a method for constructing a WiFi indoor positioning fingerprint library based on a generative adversarial network. The method generates amplitude and phase feature maps of different positions through a generative adversarial network, thereby constructing a WiFi indoor positioning fingerprint library with sufficient samples. However, the method has the following technical problems: the training process of the generative adversarial network may require a large number of data samples, resulting in high computational complexity; in addition, the generated fingerprint library may not fully reflect the complexity of the actual WiFi environment.

[0006] The above methods require a large amount of sampling data when building the fingerprint library, which not only increases the complexity of data processing, but also reduces the efficiency of positioning. Therefore, in the scenario of complex indoor environment and severe WiFi signal interference, how to quickly and accurately build a high-precision fingerprint library is an urgent problem to be solved in the current WiFi indoor positioning technology. Summary of the invention

[0007] The technical problem to be solved by the present invention is to provide a WiFi fingerprint positioning method, system, device and medium based on Gaussian process regression in view of the deficiencies in the above-mentioned prior art, so as to solve the technical problems that the current WiFi fingerprint positioning requires a large number of data samples and the fingerprint positioning is inaccurate.

[0008] The objective of the present invention is achieved by the following technical solutions: In a first aspect, the present invention provides a WiFi fingerprint positioning method based on Gaussian process regression, comprising: In a known indoor physical environment, simulate the overall distribution of the target's indoor WiFi signal; According to the set threshold, the overall distribution area of ​​the indoor WiFi signal is divided into several sub-areas; Obtain corresponding kernel functions according to the plurality of sub-regions, respectively, and construct a Gaussian process regression model based on the kernel functions corresponding to the plurality of sub-regions; Obtain the RSS signal of the WiFi in the target indoor environment, input the RSS signal into the Gaussian process regression model, and fit the RSS signal distribution curve of the WiFi in the target room; Constructing a WiFi location fingerprint database according to the RSS signal distribution curve; The real-time RSS fingerprint is similarly calculated with the WiFi location fingerprint database to obtain the current location positioning estimation result.

[0009] As a further improvement of the present invention, in an indoor physical environment where the target is known, simulating the overall distribution of indoor WiFi signals specifically includes: The overall distribution of indoor WiFi signals is simulated using the WiFi signal path loss model and the indoor wireless signal propagation model; The signal propagation model adopts the Motley-Kennan model, which includes:

[0010] The formula expression of the path loss model is:

[0011] In the formula, is the propagation distance between the transmitting source and the receiving end, is the distance from the emission source The path loss at is the reference distance The path loss at is the path loss exponent, is the number of obstacles, It is The attenuation factor of the obstacle is is the propagation distance, is the signal frequency, is the number of penetrations through the wall, is the amount of penetration through the floor.

[0012] As a further improvement of the present invention, the overall distribution area of ​​the indoor WiFi signal is divided according to the set threshold, specifically including: The fixed threshold is set by the contour method, and the overall distribution area of ​​the indoor WiFi signal is divided according to the strength continuity of the indoor WiFi signal; The area where the indoor WiFi signal is continuous is divided into several sub-areas.

[0013] As a further improvement of the present invention, corresponding kernel functions are obtained respectively according to the several sub-regions, specifically including: the kernel function includes at least one of an exponential kernel function, a square exponential kernel function, and a Matern kernel function.

[0014] As a further improvement of the present invention, a Gaussian process regression model is constructed based on the kernel functions corresponding to the plurality of sub-regions, specifically including: Defining the Mapping :

[0015] Where S is the reference point position, satisfying ; y is the corresponding WiFi signal strength at a specific location, expressed as ; N is the number of reference points in the positioning space; the reference point positions and WiFi signal strengths conform to the multivariate Gaussian distribution:

[0016] In the formula, represents the mean, express The covariance matrix of the RSS values ​​between the reference points.

[0017] As a further improvement of the present invention, a WiFi location fingerprint database is constructed according to the RSS signal distribution curve, specifically including: The WiFi location fingerprint database includes multiple reference point information; the reference point information includes , Expressed as The indoor space coordinate information of the reference point corresponds to the current indoor The RSS signal sequence of the access point is: ,When all reference point information is collected, the location fingerprint database is constructed.

[0018] As a further improvement of the present invention, the real-time RSS fingerprint is similarly calculated with the WiFi location fingerprint database, and the similarity calculation algorithm adopts the K-weighted neighbor method, which specifically includes: Obtain all the test points in the target indoor environment, and use the WKNN algorithm to estimate the Euclidean distance between the test point and each reference point; Sort the reference points according to the distance, select the K closest reference points as neighboring points, and calculate the weights of the neighboring points; According to the weights of the neighboring points, the position coordinates of the point to be measured are obtained by performing weighted average calculation on the coordinates of the neighboring points.

[0019] In a second aspect, the present invention provides a WiFi fingerprint positioning system based on Gaussian process regression, which is used to implement the above-mentioned WiFi fingerprint positioning method based on Gaussian process regression, including: The WiFi distribution simulation module is used to simulate the overall distribution of the target indoor WiFi signal in a known indoor physical environment of the target; The area division module is used to divide the overall distribution area of ​​the indoor WiFi signal into several sub-areas according to the set threshold; A Gaussian process regression model construction module obtains corresponding kernel functions according to the plurality of sub-regions, and constructs a Gaussian process regression model based on the kernel functions corresponding to the plurality of sub-regions; A location fingerprint database acquisition module is used to acquire the RSS signal of the WiFi in the target indoor environment, input the RSS signal into the Gaussian process regression model, and fit the RSS signal distribution curve of the WiFi in the target room; and construct a WiFi location fingerprint database according to the RSS signal distribution curve; The WiFi fingerprint positioning module calculates the similarity between the real-time RSS fingerprint and the WiFi location fingerprint database to obtain the current location positioning estimation result.

[0020] In a third aspect, the present invention provides a computer-readable storage medium storing one or more programs, wherein the one or more programs include instructions, which, when executed by a computing device, enable the computing device to perform the above-mentioned WiFi fingerprint positioning method based on Gaussian process regression.

[0021] In a fourth aspect, the present invention provides a computing device, comprising: One or more processors, a memory and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs include steps for executing the above-mentioned WiFi fingerprint positioning method based on Gaussian process regression.

[0022] The beneficial effects of the present invention are as follows: the WiFi fingerprint positioning method based on Gaussian process regression provided by the present invention can perform nonlinear modeling on signal distribution through Gaussian process regression, capture the spatial variation law of the signal, thereby improving the positioning accuracy, and better adapt to the local signal characteristics by selecting suitable kernel functions for different sub-areas. The indoor environment is divided into several sub-areas according to the signal strength threshold, so that the signal characteristics of each area are more consistent, further improving the positioning accuracy. By dividing the indoor environment into several sub-areas, the computational complexity of the Gaussian process regression model is reduced. Gaussian process regression models are constructed for different sub-areas respectively, which can effectively handle the multi-scale changes of the signal.

[0023] Furthermore, the Motley-Kennan model is used as the signal propagation model, which can more accurately simulate the propagation characteristics of indoor WiFi signals, taking into account factors such as multipath effects and obstacle attenuation, thereby improving the accuracy of the simulation.

[0024] Furthermore, by selecting K nearest neighbor points and performing weighted averaging, the WKNN algorithm can effectively reduce the impact of noise and improve positioning accuracy. The calculation process of the WKNN algorithm is relatively simple, the amount of calculation is moderate, and it is suitable for real-time positioning applications. Through reasonable indexing and optimization, efficient real-time positioning can be achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0026] Figure 1 The present invention is a flow chart of a WiFi fingerprint positioning method based on improved Gaussian process regression.

[0027] Figure 2 It is a system framework diagram of a WiFi fingerprint positioning method based on improved Gaussian process regression in the present invention.

[0028] Figure 3The present invention discloses an indoor WiFi signal distribution simulation diagram of a WiFi fingerprint positioning method based on improved Gaussian process regression.

[0029] Figure 4 The invention discloses an improved Gaussian process regression model fitting diagram of a WiFi fingerprint positioning method based on improved Gaussian process regression.

[0030] Figure 5 This is a positioning performance comparison diagram of a WiFi fingerprint positioning method based on improved Gaussian process regression in the present invention.

[0031] Figure 6 It is a structural schematic diagram of an electronic device of the present invention. DETAILED DESCRIPTION

[0032] In order to make the purpose and technical solution of the present invention clearer and easier to understand, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. The specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0033] RSS (Received Signal Strength): refers to the signal strength of the wireless access point (AP) detected by the receiving device (such as a mobile phone, computer, etc.).

[0034] WKNN (Weighted K-Nearest Neighbors): An improved K-nearest neighbor algorithm for classification or regression problems.

[0035] The technical solution of the present invention will be clearly and completely described below in conjunction with the accompanying drawings and specific embodiments, wherein the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments.

[0036] Example 1 like Figure 1 As shown, this embodiment provides a WiFi fingerprint positioning method based on Gaussian process regression. The indoor WiFi signal propagation is simulated by an improved Gaussian process regression model, and the signal sampling intensity is reduced by predicting the signal strength of the unsampled reference point. Not only can the WiFi fingerprint database be quickly constructed, but the model training does not require a large number of data samples, and has the advantages of accuracy and real-time performance, thereby improving the accuracy of WiFi indoor positioning. The following is a specific implementation method.

[0037] In an indoor physical environment where the target is known, the overall distribution of the target indoor WiFi signal is simulated. The indoor wireless signal propagation model and the WiFi signal path loss model are used to simulate the distribution of indoor WiFi signals.

[0038] The formula expression of the path loss model in this embodiment is:

[0039] The signal propagation model adopts the Motley-Kennan model, and the specific expression of the Motley-Kennan model is:

[0040] In the formula, is the propagation distance between the transmitting source and the receiving end, is the distance from the emission source The path loss at is the reference distance The path loss at is the path loss exponent, is the number of obstacles, It is The attenuation factor of the obstacle is is the propagation distance, is the signal frequency, is the number of penetrations through the wall, is the amount of penetration through the floor.

[0041] According to the set threshold, the overall distribution area of ​​the indoor WiFi signal is divided into several sub-areas. Specifically, a fixed threshold is set according to the contour method, and the overall distribution area of ​​the indoor WiFi signal is divided according to the continuity of the indoor WiFi signal strength. The area where the indoor WiFi signal is continuous is divided to obtain several sub-areas.

[0042] Corresponding kernel functions are obtained according to the sub-regions, and Gaussian process regression models are constructed based on the kernel functions corresponding to the sub-regions.

[0043] For different sub-regions obtained by threshold segmentation, Gaussian process regression models with different kernel functions are used. Computational experiments are carried out on each region separately to verify the applicable kernel functions for different sub-regions. A multi-kernel improved Gaussian process regression model is constructed based on the kernel functions of different sub-regions.

[0044] The kernel function in this embodiment includes at least an exponential kernel function, a square exponential kernel function, and a Matern kernel function.

[0045] Among them, the expression of the exponential kernel function is:

[0046] in, is the characteristic variable, is the target variable. is the exponential kernel function parameter.

[0047] The formula of the square exponential kernel function is:

[0048] in, is the characteristic variable, is the target variable, represents the maximum covariance factor, whose value corresponds to the adaptability to signal fluctuations, represents the length scale parameter.

[0049] The Matern kernel function is:

[0050]

[0051]

[0052] in, is the characteristic variable, is the target variable, represents the maximum covariance factor, which adds an additional parameter , used to control the smoothness of the resulting function.

[0053] The exponential kernel function is effective in processing continuous smooth data and is suitable for large areas and simple scenes. The square exponential kernel function can better fit nonlinear boundaries and is suitable for complex scenes. The Matern kernel function can be used to model functions with different degrees of smoothness and performs better when WiFi signal data is rich. Therefore, the Gaussian process regression model uses different kernel functions in different segmented area environments, which can effectively improve the prediction accuracy of the model.

[0054] The Gaussian process regression model specifically includes the definition of the mapping :

[0055] Where S is the reference point position, satisfying ; y is the corresponding WiFi signal strength at a specific location, expressed as ; N is the number of reference points in the positioning space; the reference point positions and WiFi signal strengths conform to the multivariate Gaussian distribution:

[0056] In the formula, represents the mean, express The covariance matrix of the RSS values ​​between reference points. The covariance matrix is:

[0057] Get the RSS signal of the WiFi in the target indoor environment, input the RSS signal into the Gaussian process regression model, and fit the RSS signal distribution curve of the WiFi in the target room. The obtained WiFi signal distribution simulation diagram is as follows: Figure 3 The model fitting diagram is shown in Figure 4 As shown in the figure, by simulating the propagation of WiFi signals in the physical space and partitioning them, a corresponding Gaussian regression model is created for each sub-area of ​​the environment, avoiding the interference factors caused by wall attenuation and improving the prediction accuracy of the model.

[0058] The WiFi location fingerprint database is constructed based on the RSS signal distribution curve. Specifically, the WiFi location fingerprint database includes multiple reference point information; the reference point information includes , Expressed as The indoor space coordinate information of the reference point corresponds to the current indoor The RSS signal sequence of the access point is: ,When all reference point information is collected, the location fingerprint database is constructed.

[0059] The real-time RSS fingerprint is similarly calculated with the WiFi location fingerprint database to obtain the current location positioning estimation result.

[0060] In this embodiment, the method for calculating similarity adopts the K-weighted neighbor method, which specifically includes: Obtain all the test points in the target indoor environment, and use the WKNN algorithm to estimate the Euclidean distance between the test point and each reference point; Sort the reference points according to the distance, select the K closest reference points as neighboring points, and calculate the weights of the neighboring points; According to the weights of the neighboring points, the coordinates of the neighboring points are weighted averaged to obtain the position coordinates of the point to be measured. This embodiment uses the correlation of the signal strength of adjacent reference points to establish an improved Gaussian process regression model, predicts the signal strength of the reference point to be predicted by the several sampled reference points closest to the reference point to be predicted, and more accurately predicts the signal strength of the unsampled reference point.

[0061] Example 2 like Figure 1-Figure 5 As shown, this embodiment is a preferred implementation in Example 1, and the specific implementation is as follows.

[0062] In an indoor space where the target is known, the path loss model described in the IEEE 802.11ax protocol and the Motley-Kennan indoor signal propagation model are used to simulate the distribution of indoor WiFi signals, such as Figure 3 shown.

[0063] The indoor signal propagation model is the Motley-Kennan model, and the formula is as follows:

[0064] in, is the propagation distance, is the distance from the emission source The path loss at is the reference distance The path loss at is the path loss exponent, is the number of obstacles, It is The attenuation factor of an obstacle.

[0065] The path loss model of the WiFi signal is the path loss model described by the IEEE 802.11ax protocol, and the formula is as follows:

[0066] in, is the propagation distance, is the signal frequency, is the number of penetrations through the wall, is the amount of penetration through the floor.

[0067] A fixed threshold is set to divide the overall distribution area of ​​the WiFi signal into different areas according to the continuity of the wireless signal strength; the continuous part of the WiFi signal in the indoor space is segmented to obtain several sub-areas.

[0068] For different regions obtained by threshold segmentation, Gaussian process regression models with different kernel functions are used to perform calculation experiments on each sub-region, verify the applicable kernel functions of different regions, and construct an improved Gaussian process regression model. Kernel functions include exponential kernel function, square exponential kernel function, and Matern kernel function.

[0069] The Gaussian process regression model specifically includes defining the mapping :

[0070] In the Gaussian process regression model, the RSS training data in the training phase includes and .in, is the reference point position, For a certain signal access point, express , is the corresponding WiFi signal strength in a specific location. is the number of reference points in the positioning space. The positions and signal strengths of the reference points conform to the multivariate Gaussian distribution. The multivariate Gaussian distribution has been described in Example 1 and will not be repeated here. The multivariate joint Gaussian distribution is described by the following process:

[0071] set up represents the measurement Gaussian noise with independent and identical distribution, is the variance, , then the position of the reference point and the corresponding WiFi signal strength The relationship between can be determined by the following observation model:

[0072] Among them, covariance is represented by various kernel functions, which are also called covariance functions and are used to describe the correlation between different data points. Different kernel functions can be used to model different types of data characteristics, so multiple kernel functions are the core components of improving the Gaussian process regression model.

[0073] Use the terminal device to collect the RSS signal of WiFi in the target indoor environment, input it into the partitioned Gaussian regression model, and fit the indoor WiFi RSS signal distribution map in real time. The obtained WiFi signal distribution simulation map is as follows Figure 3 The model fitting diagram is shown in Figure 4 shown.

[0074] The WiFi fingerprint database is constructed through the RSS signal distribution map; the WiFi fingerprint database specifically includes: In the database Expressed as The indoor space coordinate information of the reference point corresponds to the current indoor The RSS signal sequence of an AP is ,When all reference point information is collected, the database is constructed.

[0075] The pre-processed real-time RSS fingerprint is similarly calculated with the WiFi fingerprint database to obtain the current location estimate, specifically including: the algorithm used for similarity calculation between the RSS fingerprint and the WiFi fingerprint database is the K-weighted neighbor method.

[0076] In the target space Reference point position coordinates , the signal information of the reference point is expressed as ,in, , indicating the reference point Received The signal strength value of the WiFi access point. The received WiFi fingerprint information can be expressed as . WKNN algorithm estimates the test point With each reference point The distance between , then sort the reference points from small to large distance, and select the closest The calculation method of the distance between the reference points as the nearest neighbor point is Euclidean distance. The calculation formula of Euclidean distance is as follows:

[0077] Should The set of neighboring points , the corresponding distance is , the weight of the neighboring points Set to the inverse of the distance, the formula is as follows:

[0078] The position coordinates of the point to be measured are calculated by weighted average of the coordinates of the adjacent points:

[0079]

[0080] The cumulative distribution diagram of the error after positioning matching is as follows: Figure 5 As shown, DGPR is an improved Gaussian process regression model of the present invention. Compared with GPR and BGPR, the accuracy of DGPR is improved, which further proves the superiority of the present invention.

[0081] Example 3 This embodiment provides a WiFi fingerprint positioning system based on Gaussian process regression, which is used to implement the WiFi fingerprint positioning method based on Gaussian process regression in the above-mentioned embodiment 1 and embodiment 2. The system includes: The WiFi distribution simulation module is used to simulate the overall distribution of the target indoor WiFi signal in a known indoor physical environment of the target; The area division module is used to divide the overall distribution area of ​​the indoor WiFi signal into several sub-areas according to the set threshold; A Gaussian process regression model construction module obtains corresponding kernel functions according to the plurality of sub-regions, and constructs a Gaussian process regression model based on the kernel functions corresponding to the plurality of sub-regions; A location fingerprint database acquisition module is used to acquire the RSS signal of the WiFi in the target indoor environment, input the RSS signal into the Gaussian process regression model, and fit the RSS signal distribution curve of the WiFi in the target room; and construct a WiFi location fingerprint database according to the RSS signal distribution curve; The WiFi fingerprint positioning module calculates the similarity between the real-time RSS fingerprint and the WiFi location fingerprint database to obtain the current location positioning estimation result.

[0082] Example 4 In another embodiment of the present invention, the present invention also provides a storage medium, specifically a computer-readable storage medium (Memory), which is a memory device in a terminal device for storing programs and data. It is understandable that the computer-readable storage medium here can include both the built-in storage medium in the terminal device and the extended storage medium supported by the terminal device, and can be any tangible medium containing or storing a program, which can be used by an instruction execution system, device or device or used in combination with it. The computer-readable storage medium provides a storage space, which stores the operating system of the terminal. In addition, one or more instructions suitable for being loaded and executed by the processor are also stored in the storage space, and these instructions can be one or more computer programs (including program codes). It should be noted that more specific examples (non-exhaustive list) of the computer-readable storage medium here include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0083] Computer readable storage media also include data signals propagated in baseband or as part of a carrier wave, which carry readable program codes. Such propagated data signals can take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The readable storage medium can also be any readable medium other than a readable storage medium, which can send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, device, or device. The program code contained on the readable storage medium can be transmitted using any appropriate medium, including but not limited to wireless, wired, optical cable, RF, etc., or any suitable combination of the above.

[0084] Program code for performing the operations of the present invention may be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java, C++, etc., and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user device, as a separate software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server. In the case of a remote computing device, the remote computing device may be connected to the user computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computing device (e.g., through the Internet using an Internet service provider).

[0085] The processor may load and execute one or more instructions stored in the computer-readable storage medium to implement the corresponding steps of the WiFi fingerprint positioning method based on Gaussian process regression in the above embodiment; the processor may load and execute the following steps of one or more instructions in the computer-readable storage medium: In a known indoor physical environment, simulate the overall distribution of the target's indoor WiFi signal; According to the set threshold, the overall distribution area of ​​the indoor WiFi signal is divided into several sub-areas; Obtain corresponding kernel functions according to the plurality of sub-regions, respectively, and construct a Gaussian process regression model based on the kernel functions corresponding to the plurality of sub-regions; Obtain the RSS signal of the WiFi in the target indoor environment, input the RSS signal into the Gaussian process regression model, and fit the RSS signal distribution curve of the WiFi in the target room; Constructing a WiFi location fingerprint database according to the RSS signal distribution curve; The real-time RSS fingerprint is similarly calculated with the WiFi location fingerprint database to obtain the current location positioning estimation result.

[0086] Example 5 See also Figure 6 , the terminal device is a computer device, and the computer device 60 of this embodiment includes: a processor 61, a memory 62, and a computer program 63 stored in the memory 62 and executable on the processor 61. When the computer program 63 is executed by the processor 61, the WiFi fingerprint positioning method based on Gaussian process regression in the embodiment is implemented. To avoid repetition, it is not described one by one here. Alternatively, when the computer program 63 is executed by the processor 61, the functions of each model / unit in the computing system composed of embodiment 1 are implemented. To avoid repetition, it is not described one by one here.

[0087] The computer device 60 may be a computing device such as a desktop computer, a notebook, a PDA, or a cloud server. The computer device 60 may include, but is not limited to, a processor 61 and a memory 62. Those skilled in the art will appreciate that Figure 6 This is only an example of the computer device 60 and does not constitute a limitation of the computer device 60. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the computer device may also include input and output devices, network access devices, buses, etc.

[0088] The processor 61 may be a central processing unit (CPU), or other general-purpose processors, central processing units, graphics processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, data processing logic based on quantum computing, discrete hardware components, etc. A general-purpose processor may be a microprocessor or the processor may be any conventional processor, etc.

[0089] The memory 62 may be an internal storage unit of the computer device 60, such as a hard disk or memory of the computer device 60. The memory 62 may also be an external storage device of the computer device 60, such as a plug-in hard disk, a smart memory card (SmartMediaCard, SMC), a secure digital (SecureDigital, SD) card, a flash card (FlashCard), etc. equipped on the computer device 60.

[0090] Furthermore, the memory 62 may include both an internal storage unit of the computer device 60 and an external storage device. The memory 62 is used to store computer programs and other programs and data required by the computer device. The memory 62 may also be used to temporarily store data that has been output or is to be output.

[0091] Any reference to a memory, database or other medium used in the embodiments provided in this application may include at least one of a non-volatile and a volatile memory. Non-volatile memory may include a read-only memory (ROM), a magnetic tape, a floppy disk, a flash memory, an optical memory, a high-density embedded non-volatile memory, a resistive random access memory (ReRAM), a magnetic random access memory (MRAM), a ferroelectric random access memory (FRAM), a phase change memory (PCM), a graphene memory, etc. Volatile memory may include a random access memory (RAM) or an external cache memory, etc. As an illustration and not limitation, RAM may be in various forms, such as a static random access memory (SRAM) or a dynamic random access memory (DRAM), etc.

[0092] The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. The non-relational database may include a distributed database based on blockchain, etc., but is not limited thereto. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but is not limited thereto.

Claims

1. A WiFi fingerprint positioning method based on Gaussian process regression, characterized in that: include: In a known indoor physical environment, simulate the overall distribution of the target's indoor WiFi signal; According to the set threshold, the overall distribution area of ​​the indoor WiFi signal is divided into several sub-areas; Obtain corresponding kernel functions according to the plurality of sub-regions, respectively, and construct a Gaussian process regression model based on the kernel functions corresponding to the plurality of sub-regions; Obtain the RSS signal of the WiFi in the target indoor environment, input the RSS signal into the Gaussian process regression model, and fit the RSS signal distribution curve of the WiFi in the target room; Constructing a WiFi location fingerprint database according to the RSS signal distribution curve; The real-time RSS fingerprint is similarly calculated with the WiFi location fingerprint database to obtain the current location positioning estimation result.

2. The WiFi fingerprint positioning method based on Gaussian process regression according to claim 1, characterized in that: In an indoor physical environment where the target is known, simulate the overall distribution of indoor WiFi signals, including: The overall distribution of indoor WiFi signals is simulated using the WiFi signal path loss model and the indoor wireless signal propagation model. The signal propagation model adopts the Motley-Kennan model, which includes: The formula expression of the path loss model is: In the formula, is the propagation distance between the transmitting source and the receiving end, is the distance from the emission source The path loss at is the reference distance The path loss at is the path loss exponent, is the number of obstacles, It is The attenuation factor of the obstacle is is the propagation distance, is the signal frequency, is the number of penetrations through the wall, is the amount of penetration through the floor.

3. The WiFi fingerprint positioning method based on Gaussian process regression according to claim 2 is characterized in that: According to the set threshold, the overall distribution area of ​​the indoor WiFi signal is divided into: The fixed threshold is set by the contour method, and the overall distribution area of ​​the indoor WiFi signal is divided according to the strength continuity of the indoor WiFi signal; The area where the indoor WiFi signal is continuous is divided into several sub-areas.

4. The WiFi fingerprint positioning method based on Gaussian process regression according to claim 1, characterized in that: Corresponding kernel functions are obtained respectively according to the plurality of sub-regions, specifically including: the kernel function includes at least one of an exponential kernel function, a square exponential kernel function, and a Matern kernel function.

5. The WiFi fingerprint positioning method based on Gaussian process regression according to claim 4 is characterized in that: A Gaussian process regression model is constructed based on the kernel functions corresponding to several sub-regions, including: Defining the Mapping : Where S is the reference point position, satisfying ; y is the corresponding WiFi signal strength at a specific location, expressed as ; N is the number of reference points in the positioning space; the reference point positions and WiFi signal strengths conform to the multivariate Gaussian distribution: In the formula, represents the mean, express The covariance matrix of the RSS values ​​between the reference points.

6. The WiFi fingerprint positioning method based on Gaussian process regression according to claim 1 is characterized in that: Building a WiFi location fingerprint database according to the RSS signal distribution curve specifically includes: The WiFi location fingerprint database includes multiple reference point information; the reference point information includes , Expressed as The indoor space coordinate information of the reference point corresponds to the current indoor The RSS signal sequence of the access point is: ,When all reference point information is collected, the location fingerprint database is constructed.

7. The WiFi fingerprint positioning method based on Gaussian process regression according to claim 1 is characterized in that: The real-time RSS fingerprint is similarly calculated with the WiFi location fingerprint database. The similarity calculation algorithm adopts the K-weighted proximity method, which specifically includes: Obtain all the test points in the target indoor environment, and use the WKNN algorithm to estimate the Euclidean distance between the test point and each reference point; Sort the reference points according to the distance, select the K closest reference points as neighboring points, and calculate the weights of the neighboring points; According to the weights of the neighboring points, the position coordinates of the point to be measured are obtained by performing weighted average calculation on the coordinates of the neighboring points.

8. A WiFi fingerprint positioning system based on Gaussian process regression, used to implement the WiFi fingerprint positioning method based on Gaussian process regression according to any one of claims 1 to 7, characterized in that: include: The WiFi distribution simulation module is used to simulate the overall distribution of the target indoor WiFi signal in a known indoor physical environment of the target; The area division module is used to divide the overall distribution area of ​​the indoor WiFi signal into several sub-areas according to the set threshold; A Gaussian process regression model construction module obtains corresponding kernel functions according to the plurality of sub-regions, and constructs a Gaussian process regression model based on the kernel functions corresponding to the plurality of sub-regions; A location fingerprint database acquisition module is used to acquire the RSS signal of the WiFi in the target indoor environment, input the RSS signal into the Gaussian process regression model, and fit the RSS signal distribution curve of the WiFi in the target room; and construct a WiFi location fingerprint database according to the RSS signal distribution curve; The WiFi fingerprint positioning module calculates the similarity between the real-time RSS fingerprint and the WiFi location fingerprint database to obtain the current location positioning estimation result.

9. A computer-readable storage medium storing one or more programs, characterized in that: The one or more programs include instructions, which, when executed by a computing device, enable the computing device to execute the WiFi fingerprint positioning method based on Gaussian process regression as described in any one of claims 1 to 7.

10. A computing device, characterized in that include: One or more processors, a memory and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs include steps for executing the WiFi fingerprint positioning method based on Gaussian process regression as described in any one of claims 1 to 7.

Images