CN112911704B

CN112911704B - CSI indoor fingerprint positioning method and device based on dynamic fusion characteristics

Info

Publication number: CN112911704B
Application number: CN202110080046.9A
Authority: CN
Inventors: 邓中亮; 许允飞; 韩可; 付加伟
Original assignee: Beijing University of Posts and Telecommunications
Current assignee: Beijing University of Posts and Telecommunications
Priority date: 2021-01-21
Filing date: 2021-01-21
Publication date: 2022-04-05
Anticipated expiration: 2041-01-21
Also published as: CN112911704A

Abstract

The embodiment of the invention provides a CSI indoor fingerprint positioning method and device based on dynamic fusion characteristics, which are used for configuring parameters with environmental adaptability, namely dynamically determining fusion weights and a matching threshold, reducing the problem of feature fuzzification caused by the introduction of random noise information in the fingerprint forming process of a fingerprint positioning system, improving the feature resolution of the fingerprint positioning method, improving the indoor positioning precision and simultaneously providing theoretical reference for the improvement of the positioning precision of the indoor positioning method.

Description

CSI indoor fingerprint positioning method and device based on dynamic fusion characteristics

Technical Field

The invention relates to the technical field of indoor positioning, in particular to a CSI indoor fingerprint positioning method and device based on dynamic fusion characteristics.

Background

With the development of communication technology, location services are becoming more and more important as part of communication services. Fingerprint positioning technology is a common technology in location services. Currently, a Global Navigation Satellite System (GNSS) can provide high-precision positioning outdoors, but GNSS signals affect indoor positioning under the condition of shielding of buildings, so that indoor positioning can be performed by using a fingerprint positioning technology. In the indoor positioning method, a Wireless Fidelity (WiFi for short) positioning technology is widely used due to the convenience of infrastructure deployment.

In the positioning process by using the WiFi positioning technology, Received Signal Strength (RSS) is used as basic location information for indoor positioning. Changes in the indoor environment, such as the addition of a wall or the increase in traffic, affect the RSS information. That is, the RSS signal is a superimposed value of a multipath signal, and for the information receiving end, the fluctuation degree of the RSS signal is greatly affected by the multipath effect, so that a certain deviation is generated on the positioning result, and the positioning accuracy is reduced.

Disclosure of Invention

The embodiment of the invention aims to provide a CSI indoor fingerprint positioning method and device based on dynamic fusion characteristics, which can improve the indoor positioning precision. The specific technical scheme is as follows:

in a first aspect, an embodiment of the present invention provides a CSI indoor fingerprint positioning method based on dynamic fusion features, including:

acquiring Channel State Information (CSI) online data of different antennas for acquiring to-be-positioned points; wherein the CSI-on-line data provides amplitude data and phase data collected on a plurality of subcarriers;

analyzing the correlation between the amplitude data and the phase data collected on the multiple subcarriers to obtain a data maximum characteristic value, wherein the data maximum characteristic value comprises: maximum eigenvalue of amplitude data and maximum eigenvalue of phase data; the data maximum characteristic value is used for measuring the correlation between amplitude data and phase data acquired by different antennas;

respectively calculating the maximum eigenvalue of the amplitude data and the maximum eigenvalue of the phase data, and obtaining the fusion weight of the amplitude data and the fusion weight of the phase data according to the proportion of the maximum eigenvalue of the data;

weighting and fusing the amplitude data and the phase data of the to-be-positioned points by adopting the fusion weight of the amplitude data and the fusion weight of the phase data to obtain fusion fingerprint data serving as an online CSI position fingerprint of the to-be-positioned points;

matching the on-line CSI position fingerprint of the to-be-positioned point with the off-line CSI position fingerprint in the database to obtain an off-line CSI position fingerprint which is most matched with the on-line CSI position fingerprint so as to determine the position of the to-be-positioned point; the off-line CSI position fingerprints in the database are obtained by preprocessing and correlation analysis of CSI measurement data of each reference point acquired by different antennas to obtain fusion weights of amplitude data and phase data, and performing weighted fusion on the preprocessed amplitude data and preprocessed phase data of each reference point to obtain off-line CSI position fingerprints of each reference point; the offline CSI location fingerprint that best matches the online CSI location fingerprint is most similar to the online CSI location fingerprint.

Further, the matching the online CSI location fingerprint of the to-be-located point with the offline CSI location fingerprint in the database to obtain the offline CSI location fingerprint that is most matched with the online CSI location fingerprint includes:

and calculating the similarity between the on-line CSI position fingerprint and the off-line CSI position fingerprint in the database by using the EDR to obtain the off-line CSI position fingerprint with the maximum similarity to the on-line fingerprint, and using the off-line CSI position fingerprint as the off-line CSI position fingerprint which is most matched with the on-line CSI position fingerprint.

Further, the calculating, by using EDR, a similarity between the online CSI location fingerprint and the offline CSI location fingerprint in the database to obtain an offline CSI location fingerprint having a maximum similarity with the online fingerprint, as an offline CSI location fingerprint that is most matched with the online CSI location fingerprint, includes:

calculating the similarity between the online CSI position fingerprint and the offline CSI position fingerprint in the database by using EDR to obtain the similarity between the online CSI position fingerprint and the offline CSI position fingerprint;

acquiring a matching threshold;

taking the off-line CSI position fingerprint with the similarity reaching the matching threshold as the off-line CSI position fingerprint which is most matched with the on-line CSI position fingerprint;

the method comprises the following steps of:

calculating the variance of the amplitude data and the phase data collected on each path of subcarrier to obtain variance data;

averaging the variance data to obtain an average variance;

according to the Laplace criterion, the following formula is adopted:

obtaining the matching threshold;

wherein, epsilon is a matching threshold value,

the average variance collected over all subcarriers for merging fingerprint data, as a whole,

n is the total number of subcarriers contained in the fingerprint data, M is the number of acquired data packets, L_ijI is the data packet sequence number, j is the subcarrier sequence number, sigma is the summation sign,

is the average value of all data packets on the jth sub-carrier.

Further, the following steps are adopted to determine the offline CSI location fingerprint in the database:

acquiring a plurality of Channel State Information (CSI) measurement data of different antenna acquisition reference points;

respectively preprocessing a plurality of groups of amplitude data and a plurality of groups of phase data collected on each path of subcarrier to obtain preprocessed amplitude data and preprocessed phase data of the reference point;

performing correlation analysis on the preprocessed amplitude data and the preprocessed phase data to obtain a data maximum characteristic value, wherein the data maximum characteristic value comprises: maximum eigenvalue of amplitude data and maximum eigenvalue of phase data; the data maximum characteristic value is used for measuring the correlation between amplitude data and phase data acquired by different antennas;

respectively calculating the maximum eigenvalue of the amplitude data and the maximum eigenvalue of the phase data, determining the proportion of the maximum eigenvalue of the data, and determining the fusion weight of the amplitude data and the fusion weight of the phase data;

and performing weighted fusion on the preprocessed amplitude data and the preprocessed phase data of each reference point to obtain the offline CSI position fingerprint of each reference point.

Further, the obtaining of the preprocessed amplitude data and the preprocessed phase data for the sets of amplitude data and sets of phase data collected on each path of subcarriers respectively includes:

dividing a plurality of groups of amplitude data collected on each path of subcarrier into a plurality of data clusters, and extracting the maximum data cluster with the largest number of samples; wherein the sample is amplitude data in each data cluster;

carrying out mean processing on all samples in the maximum data cluster to obtain amplitude data after preprocessing;

performing unwrapping processing on a plurality of groups of phase data collected on each path of subcarrier, and comparing whether the absolute value of the difference between two adjacent phase data is greater than pi; if the absolute value of the difference between two adjacent phase data is larger than pi, increasing or decreasing 2 pi to all the numerical values behind the two adjacent phase data until the absolute value of the difference between the two adjacent phase data is smaller than pi, and obtaining unwrapped phase data;

and performing linear transformation on the unwrapped phase data to obtain preprocessed phase data.

In a second aspect, an embodiment of the present invention provides a CSI indoor fingerprint positioning apparatus based on dynamic fusion features, including:

the acquisition module is used for acquiring the Channel State Information (CSI) online data of the to-be-positioned points acquired by different antennas; wherein the CSI-on-line data provides amplitude data and phase data collected on a plurality of subcarriers;

an analysis module, configured to analyze a correlation between amplitude data and phase data collected on multiple subcarriers, to obtain a maximum characteristic value of the data, where the maximum characteristic value of the data includes: maximum eigenvalue of amplitude data and maximum eigenvalue of phase data; the data maximum characteristic value is used for measuring the correlation between amplitude data and phase data acquired by different antennas;

the first processing module is used for respectively calculating the maximum eigenvalue of the amplitude data and the maximum eigenvalue of the phase data, and obtaining the fusion weight of the amplitude data and the fusion weight of the phase data according to the proportion of the maximum eigenvalue of the data;

the weighted fusion module is used for weighting and fusing the amplitude data and the phase data of the to-be-positioned point by adopting the fusion weight of the amplitude data and the fusion weight of the phase data to obtain fusion fingerprint data which is used as an online CSI position fingerprint of the to-be-positioned point;

the matching positioning module is used for matching the on-line CSI position fingerprint of the to-be-positioned point with the off-line CSI position fingerprint in the database to obtain an off-line CSI position fingerprint which is most matched with the on-line CSI position fingerprint so as to determine the position of the to-be-positioned point; the off-line CSI position fingerprints in the database are obtained by preprocessing and correlation analysis of CSI measurement data of each reference point acquired by different antennas to obtain fusion weights of amplitude data and phase data, and performing weighted fusion on the preprocessed amplitude data and preprocessed phase data of each reference point to obtain off-line CSI position fingerprints of each reference point; the offline CSI location fingerprint that best matches the online CSI location fingerprint is most similar to the online CSI location fingerprint.

Further, the matching location module is configured to:

acquiring a matching threshold;

the method comprises the following steps of:

averaging the variance data to obtain an average variance;

according to the Laplace criterion, the following formula is adopted:

obtaining the matching threshold；

Wherein, epsilon is a matching threshold value,

n is the total number of subcarriers contained in the fingerprint data, M is the number of acquired data packets, L_ijTo the size of the jth subcarrier data on the ith packet,_iis the packet sequence number, j is the subcarrier sequence number, Σ is the summation symbol,

is the average value of all data packets on the jth sub-carrier.

Further, the apparatus further comprises: an establishing module, configured to determine an offline CSI location fingerprint in a database by using the following steps:

In a third aspect, an embodiment of the present invention provides an electronic device, including a processor, a communication interface, a memory, and a communication bus, where the processor and the communication interface complete communication between the memory and the processor through the communication bus;

a memory for storing a computer program;

a processor for implementing the steps of the method of any one of the first aspect when executing a program stored in the memory.

In a fourth aspect, the present invention provides a computer-readable storage medium, which stores instructions that, when executed on a computer, cause the computer to perform the method of any one of the above first aspects.

The embodiment of the invention has the following beneficial effects:

according to the CSI indoor fingerprint positioning method and device based on the dynamic fusion characteristics, the dynamic fusion weights are obtained by acquiring phase data and amplitude data of positions to be positioned of different antennas, namely, correlation analysis is carried out on the amplitude data and the phase data to obtain the maximum characteristic value of the data, the maximum characteristic value of the amplitude data and the maximum characteristic value of the phase data are respectively calculated, and the fusion weights of the amplitude data and the fusion weights of the phase data are obtained according to the proportion of the maximum characteristic values of the data; then, weighting and fusing the amplitude data and the phase data by using fusion weight to obtain novel fingerprint data DFF, namely the online CSI position fingerprint of the to-be-positioned point; and matching the on-line CSI position fingerprint of the to-be-positioned point with the off-line CSI position fingerprint in the database to obtain the off-line CSI position fingerprint which is most matched with the on-line CSI position fingerprint so as to determine the position of the to-be-positioned point.

Through the technical scheme, the embodiment of the invention configures the parameters with environmental adaptability, namely dynamically determines the fusion weight, reduces the problem of feature fuzzification caused by the introduction of random noise information in the fingerprint forming process of the fingerprint positioning system, improves the feature resolution of the fingerprint positioning method and further improves the indoor positioning precision. Meanwhile, theoretical reference is provided for improving the positioning accuracy of the indoor positioning method.

Of course, not all of the advantages described above need to be achieved at the same time in the practice of any one product or method of the invention.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.

Fig. 1 is an overall flowchart of a CSI indoor fingerprint positioning method based on dynamic fusion features according to an embodiment of the present invention;

fig. 2 is a schematic basic flow chart of a CSI indoor fingerprint location method based on dynamic fusion features according to an embodiment of the present invention;

fig. 3 is a partial schematic flow chart of a CSI indoor fingerprint location method based on dynamic fusion features according to an embodiment of the present invention;

fig. 4 is a schematic structural diagram of a CSI indoor fingerprint positioning apparatus based on dynamic fusion features according to an embodiment of the present invention;

fig. 5 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

First, a CSI indoor fingerprint positioning method and apparatus based on dynamic fusion features provided in the embodiments of the present invention are described in the following.

In the related art, when positioning is performed by using the WiFi positioning technology, Received Signal Strength (RSS) is used as basic position information for indoor positioning. Typical WiFi positioning techniques will use a fingerprint library. The establishment process of the fingerprint database is to sample signal values in advance in a positioning place before positioning to obtain fingerprint data and establish the fingerprint database. Changes in the indoor environment, such as the addition of a wall or the increase in traffic, affect the RSS information. That is, the RSS signal is used as a superimposed value of the multipath signal, and for the information receiving end, the fluctuation degree of the RSS signal is greatly affected by the multipath effect, so that a certain deviation is generated on the fingerprint database and the positioning result, thereby reducing the positioning accuracy.

The inventors have determined the positioning accuracy of the system taking into account that the feature resolution of the fingerprint data represents the probability of successful feature matching. Generally, the higher the resolution, the smaller the range to be confused, and the higher the positioning accuracy. However, due to a complex indoor environment, when similarity matching is performed, fingerprint data at an online stage may contain uncertain noise data, thereby causing fluctuation of the online fingerprint data. Thereby affecting more or less the effect of fingerprint matching and the performance of the positioning method.

Meanwhile, compared with RSS, Channel State Information (CSI for short) can provide phase data and amplitude data collected on multiple subcarriers, and has a higher dimensional position characteristic. The positioning is completed by dynamic fusion of position features with higher dimensionality, namely, the weighted fusion of the phase data and the amplitude data, so that the accuracy of the positioning is improved by improving the feature resolution of a fingerprint positioning system.

Therefore, the embodiment of the invention provides a CSI indoor fingerprint positioning method and device based on dynamic fusion characteristics, wherein dynamic fusion weights are obtained by acquiring phase data and amplitude data of positions to be positioned of different antennas, namely, correlation analysis is performed on the amplitude data and the phase data to obtain maximum characteristic values of the data, the maximum characteristic values of the amplitude data and the maximum characteristic values of the phase data are respectively calculated, and the fusion weights of the amplitude data and the fusion weights of the phase data are obtained by the ratio of the maximum characteristic values of the data; then, weighting and fusing the amplitude data and the phase data by using fusion weight to obtain novel fingerprint data DFF, namely the online CSI position fingerprint of the to-be-positioned point; and matching the on-line CSI position fingerprint of the to-be-positioned point with the off-line CSI position fingerprint in the database to obtain the off-line CSI position fingerprint which is most matched with the on-line CSI position fingerprint so as to determine the position of the to-be-positioned point.

Through the technical scheme, the embodiment of the invention configures the parameters with environmental adaptability, namely dynamically determines the fusion weight, reduces the problem of feature fuzzification caused by the introduction of random noise information in the fingerprint forming process of the fingerprint positioning system, improves the feature resolution of the fingerprint positioning method, improves the indoor positioning precision, and provides theoretical reference for improving the positioning precision of the indoor positioning method.

The following provides a CSI indoor fingerprint location method based on dynamic fusion features according to an embodiment of the present invention.

The embodiment of the invention provides a CSI indoor fingerprint positioning method based on dynamic fusion characteristics, which is applied to

In particular, the method can be applied to electronic equipment, and the electronic equipment can be: intelligent mobile terminal and vehicle mounted terminal etc.. Without limitation, any electronic device that can implement the embodiments of the present invention is within the scope of the present invention.

As shown in fig. 1, a CSI indoor fingerprint location method based on dynamic fusion features provided in an embodiment of the present invention may include an offline data collection stage (hereinafter, may be referred to as an offline stage) and an online fingerprint matching stage (hereinafter, may be referred to as an online matching stage). In an off-line stage, each reference point collects CSI off-line data, and off-line CSI position fingerprints of each reference point are generated by preprocessing the CSI off-line data collected by each reference point and integrating the preprocessed CSI off-line data. And storing the offline CSI location fingerprints that ultimately form each reference point into an offline CSI location fingerprint database (i.e., fingerprint database). In the on-line matching stage, the CSI on-line data of the to-be-located point is integrated to generate an on-line CSI position fingerprint of the to-be-located point, and the on-line CSI position fingerprint is subjected to feature matching with the off-line CSI position fingerprint of each reference point in the fingerprint database to determine the position of the to-be-located point. The reference points refer to points whose locations are known and whose offline CSI location fingerprints are known. The detailed description is as follows.

As shown in fig. 2, the method of the embodiment of the present invention may include the following steps:

step 110, acquiring Channel State Information (CSI) online data of different antennas for acquiring to-be-positioned points; wherein the CSI-on-line data provides amplitude data and phase data collected on multiple subcarriers.

In this step 110, the point to be located is a position that needs to be located, and the embodiment of the present invention finally needs to locate the position coordinate of the point to be located, where the position coordinate includes longitude, latitude and altitude. The point to be located is located in a location area, which may be an outdoor area, such as a mountain tunnel; this location area may also be an outdoor area, such as a building area. Illustratively, the building area is a mall, an office building, an apartment building, a teaching building, a residential building, and the like. Any point to be located that can use the embodiments of the present invention belongs to the scope of the embodiments of the present invention.

The CSI collection of The invention depends on The Institute of Electrical and Electronics Engineers (IEEE) EEE 802.11n protocol, and realizes The sending and receiving of CSI data through commercial WiFi equipment and IWL5300 wireless network card. After positioning is completed, each to-be-positioned point generates a file containing P data packets, each data packet contains a QxR matrix, and Q and R are the number of the transmitting-receiving antenna pairs and the number of subcarriers in CSI data respectively.

In order to realize the on-line matching stage, namely real-time positioning or test positioning, in the subsequent positioning process, the embodiment of the invention acquires the CSI data of the to-be-positioned points acquired by different antennas in real time in the preset area through establishing the fingerprint database and determines the CSI data as the CSI on-line data. The different antennas are also within the predetermined area. Therefore, the CSI online data of the to-be-positioned point can be obtained in the preset area, and the subsequent positioning of the to-be-positioned point can be carried out.

Step 120, analyzing the correlation between the amplitude data and the phase data collected on the multiple subcarriers to obtain a maximum characteristic value of the data, wherein the maximum characteristic value of the data includes: maximum eigenvalue of amplitude data and maximum eigenvalue of phase data; and the data maximum characteristic value is used for measuring the correlation between the amplitude data and the phase data acquired by different antennas.

The size concentration degree of the characteristic values can reflect the correlation between the amplitude data and the phase data, and the fusion weight of the amplitude data and the fusion weight of the phase data are respectively related to the correlation between the amplitude data and the phase data acquired by different antennas. Therefore, the maximum characteristic value of the data is obtained by analyzing the correlation between the amplitude data and the phase data collected on the multiple subcarriers in step 120, and the fusion weight of the amplitude data and the fusion weight of the phase data in step 130 are obtained subsequently.

And step 130, calculating the maximum eigenvalue of the amplitude data and the maximum eigenvalue of the phase data respectively, and obtaining the fusion weight of the amplitude data and the fusion weight of the phase data according to the proportion of the maximum eigenvalue of the data.

In the embodiment of the invention, although in the off-line stage, when the fingerprint database is established, the fusion weight determined off-line is determined based on the unchanged physical position of the antenna. However, in the online matching stage, since the positions of the different transmitting ends (i.e., the positions to be located) are not determined, the transmitting and receiving directions of the antennas are changed, and therefore, the embodiment of the present invention still needs to determine the fusion weight again in the online positioning stage. In an off-line stage, fusion weight is determined so as to enable the acquired off-line CSI position fingerprint to be more accurate; when online, the fusion weight is determined again, so that the acquired online CSI location fingerprint is more accurate.

In order to implement the dynamic configuration of the fusion weight in the line matching stage, the embodiment of the present invention needs to dynamically determine the fusion weight of the amplitude data and the fusion weight of the phase data. Considering that the correlation between the data and the weight value are negative correlation, in step 130, calculating the ratio of the maximum eigenvalue of the amplitude data to the maximum eigenvalue of the data to obtain the fusion weight of the amplitude data; and calculating the proportion of the phase data maximum characteristic value in the data maximum characteristic value to obtain the fusion weight of the phase data.

And 140, weighting and fusing the amplitude data and the phase data of the to-be-positioned point by adopting the fusion weight of the amplitude data and the fusion weight of the phase data to obtain fusion fingerprint data serving as the on-line CSI position fingerprint of the to-be-positioned point.

In order to use the multi-dimensional characteristics of the CSI online data, namely amplitude data and phase data, in the online matching stage, the characteristics of the amplitude data, the phase data and the like of the point to be located are formed into one characteristic by adopting weighted fusion, and the online CSI position fingerprint of the point to be located can be obtained.

Step 150, matching the on-line CSI position fingerprint of the to-be-positioned point with the off-line CSI position fingerprint in the database to obtain an off-line CSI position fingerprint which is most matched with the on-line CSI position fingerprint so as to determine the position of the to-be-positioned point; the off-line CSI position fingerprints in the database are obtained by preprocessing and correlation analysis of CSI measurement data of each reference point acquired by different antennas to obtain fusion weights of amplitude data and phase data, and performing weighted fusion on the preprocessed amplitude data and preprocessed phase data of each reference point to obtain off-line CSI position fingerprints of each reference point; the offline CSI location fingerprint that best matches the online CSI location fingerprint is most similar to the online CSI location fingerprint. The preprocessing is used for reducing the noise of the amplitude data and eliminating the error of the phase data.

In the step 150, the offline CSI position fingerprint most matched with the online CSI position fingerprint is determined by calculating the similarity between the online CSI position fingerprint of the to-be-located point and the offline CSI position fingerprint in the database; and averaging the determined off-line CSI position fingerprints which are most matched with the on-line CSI position fingerprints to obtain an average value of the matched off-line CSI position fingerprints, and using the average value as the position of the to-be-positioned point. In this way, the position of the point to be located can be determined by averaging the offline CSI location fingerprints that most closely match the online CSI location fingerprints.

In the embodiment of the invention, dynamic fusion weight is obtained by carrying out correlation analysis on the preprocessed phase data and the preprocessed amplitude data, namely, the amplitude data and the phase data are subjected to correlation analysis to obtain a data maximum characteristic value, and the fusion weight of the amplitude data is determined according to the proportion of the amplitude data maximum characteristic value in the data maximum characteristic value; determining the fusion weight of the phase data according to the proportion of the maximum characteristic value of the phase data in the maximum characteristic value of the data; then, weighting and fusing the preprocessed amplitude data and the preprocessed phase data by using fusion weight to obtain novel fingerprint data DFF, namely the online CSI position fingerprint of the point to be located; and then matching the on-line CSI position fingerprint of the to-be-positioned point with the off-line CSI position fingerprint in the database to obtain the off-line CSI position fingerprint which is most matched with the on-line CSI position fingerprint so as to determine the position of the to-be-positioned point. By configuring parameters with environmental adaptability, namely dynamically determining fusion weight, the problem of feature fuzzification caused by the introduction of random noise information in the fingerprint forming process of a fingerprint positioning system is solved, the feature resolution of the fingerprint positioning method is improved, the indoor positioning precision is also improved, and meanwhile, theoretical reference is provided for the improvement of the positioning precision of the indoor positioning method.

The following is by the first part: preprocessing data; a second part: forming a new offline CSI location fingerprint DFF; and a third part: the similarity measure index IEDR is divided into three parts. The steps 110 to 115 are detailed to illustrate the process of completing the positioning of the to-be-positioned point:

in order to better reflect the characteristics of each reference point and reduce the influence of random noise data, the CSI provides amplitude data and phase data collected on multiple subcarriers, and often in a data collection stage in an offline stage, multiple pieces of CSI data are collected on one reference point, and the amplitude data and the phase data of the multiple pieces of CSI data are extracted. Since the offline CSI location fingerprints in the fingerprint database are stored as a set of data, the sets of amplitude data and the sets of phase data collected on each subcarrier are integrated into a set of data by preprocessing, so as to prepare for forming the offline CSI location fingerprints in the fingerprint database. Such a path of subcarriers corresponds to a set of data.

The method comprises the following steps of determining an offline CSI position fingerprint in a database:

the method comprises the first step of obtaining a plurality of Channel State Information (CSI) measurement data of different antenna acquisition reference points.

And secondly, respectively preprocessing a plurality of groups of amplitude data and a plurality of groups of phase data collected on each path of subcarrier to obtain preprocessed amplitude data and preprocessed phase data of the reference point.

Wherein the second step further comprises:

step 1, dividing a plurality of groups of amplitude data collected on each path of subcarrier into a plurality of data clusters, and extracting a maximum data cluster with the largest number of samples; wherein the sample is one amplitude data in each data cluster.

For amplitude data, as the distribution of the CSI amplitude data is concentrated, a better effect can be achieved by using a density-based clustering algorithm. The embodiment of the invention adopts a Density Clustering algorithm of Density-Based Noise application space Clustering (DBSCAN for short), divides a plurality of groups of amplitude data collected on each subcarrier into a plurality of data clusters, and forms a set C (C) from the plurality of data clusters₁,C₂,...,C_m}. By utilizing the characteristic of concentrated distribution of amplitude data, the embodiment of the invention extracts the data cluster with the largest number of samples (namely the largest data cluster) from a plurality of data clusters, and adopts the following formula:

dividing the multiple groups of amplitude data collected on each path of subcarrier into a plurality of data clusters, and extracting the data cluster with the largest data quantity;

wherein, C_maxThe data cluster with the largest data size is selected, max is the maximum value, mu is the number of samples contained in the ith data cluster, {. the sum of the data clusters is set, num (. -) is the number of data in the data cluster, and C is the number of samples contained in the ith data cluster_iThe ith data cluster i is the serial number of the data cluster;

and 2, carrying out mean processing on all samples in the maximum data cluster to obtain preprocessed amplitude data.

Assuming that 30 subcarriers are provided, for all 30 subcarriers, data (i.e., amplitude data and phase data) of each subcarrier is formed into a data cluster by using DBSCAN, and then the following formula is adopted:

carrying out mean processing on all samples in the data cluster with the largest number of samples to obtain preprocessed amplitude data;

wherein, Amp_pre(f_k) For pre-processing the amplitude data on the k-th sub-carrier, Amp_preFor the preprocessed amplitude data, pre is an expression mode of preprocessed data symbols, k is a serial number collected on a subcarrier, and Amp_j(f_k) For the jth amplitude data, Amp, in the largest data cluster collected on the kth subcarrier_jJ is the j th amplitude data in the data cluster and is the serial number of the amplitude data, N_fkNumber of samples contained for the largest data cluster among different sub-carriers, f_kIs the k path sub-carrier;

step 3, performing unwrapping processing on a plurality of groups of phase data collected on each path of subcarrier, and comparing whether the absolute value of the difference between two adjacent phase data is greater than pi; and if the absolute value of the difference between the two adjacent phase data is greater than pi, increasing or decreasing by 2 pi for all the numerical values after the two adjacent phase data until the absolute value of the difference between the two adjacent phase data is less than pi, and obtaining the unwrapped phase data.

Wherein the unrolling is to ensure the continuity of the collected signals, the range of the initial phase signal (i.e. the phase data) is maintained between [ -pi, pi ]. So that in the initial signal the signal jumps to pi if the actual value of the signal drops below-pi.

And 4, performing linear transformation on the unwrapped phase data to obtain preprocessed phase data.

For phase data, the directly measured phase data can generate a folding phenomenon due to the self-circulation characteristic, so that the phase is maintained in an area of [ -pi, pi ]. Therefore, further unwinding processing is required for the phase data.

In addition, since the phase data also has Carrier Frequency Offset (CFO) generated by incomplete synchronization of the center frequencies of the transmitter and the receiver and Sampling Frequency Offset (SFO) generated by clock asynchronization, the phase data needs to be further processed for use. Firstly, determining the measured phase data of each sub-carrier

Then the comprehensive consideration is carried out,

the components δ and β contained in the initial phase in the calculation formula (2) are difficult to measure, so the embodiment of the invention adopts a linear transformation mode to eliminate errors. And calculating the slope and offset values of the linear calibration of the phase data to obtain the preprocessed phase data. The concrete expression is as follows.

The above-mentioned 4 th step adopts the following formula:

performing linear transformation on the unwrapped phase data to obtain preprocessed phase data;

wherein,

the preprocessed phase data, a is the slope,

phase data for last sub-carrier

For phase data K of the first sub-carrier_LIndex value K of last subcarrier₁The index value L for the first subcarrier is the number of subcarriers

For test phase data psi on k-th sub-carrier_kFor true phase values, pi is 180 degrees radian, K_kIs the index value collected on the k path sub-carrier, N is the size of fast Fourier transform, delta is the time lag caused by sampling frequency shift, beta is the unknown phase shift caused by carrier frequency shift, Z is the measurement noise, b is the shift amount,

k is the number of the subcarrier

Is the preprocessed phase data.

The sequence from step 1 to step 4 is not limited, and the above is merely for convenience of illustration, and it is needless to say that step 1 and step 2 may be performed after step 3 and step 4 are performed.

Performing correlation analysis on the preprocessed amplitude data and the preprocessed phase data to obtain a data maximum characteristic value, wherein the data maximum characteristic value comprises: maximum eigenvalue of amplitude data and maximum eigenvalue of phase data; and the data maximum characteristic value is used for measuring the correlation between the amplitude data and the phase data acquired by different antennas.

And step four, respectively calculating the maximum eigenvalue of the amplitude data and the maximum eigenvalue of the phase data, and determining the fusion weight of the amplitude data and the fusion weight of the phase data according to the ratio of the maximum eigenvalue of the data.

With the following formula being used,

determining a fusion weight of the amplitude data and a fusion weight of the phase data;

wherein λ is_ampIn order to obtain the maximum characteristic value of the amplitude data through the correlation analysis of the amplitude data and the phase data, amp is the expression mode of the data symbols related to the amplitude, and pha is the expression mode lambda of the data symbols related to the phase_phaThe maximum characteristic value of the phase data is obtained through the correlation analysis of the amplitude data and the phase data. Lambda [ alpha ]_max＝max(λ₁,λ₂,λ₃)，λ_maxIs a covariance matrix Cov_η1,η2,η3The maximum eigenvalue of the eigenvalue of (c),

Cov(η_a,η_b) Is a vector η_aAnd η_bA is the expression of the data symbols associated with the data type a, 1<a<3, b is the expression of the data symbol associated with data type a, 1<b<3，η₁For normalized vector data 1 eta₂Is normalized vector data 2, eta₃For normalized vector data 3, Cov is the variable sign of the covariance matrix, [.]For the contents of the covariance matrix, λ₁Is the 1 st eigenvalue, λ, of the covariance matrix₂Is the 2 nd eigenvalue, λ, of the covariance matrix₃Is the 3 rd eigenvalue of the covariance matrix. The covariance matrix is obtained by normalizing the vector data D_mVector data D obtained by covariance calculation_mFor vector data corresponding to m antennas, D_m＝{D_m,1,D_m,2,...D_m,n,...,D_m,30}，D_m,nAmplitude data or phase data collected on the nth subcarrier for the mth antenna. It is clear that the more concentrated the eigenvalue magnitude distribution, λ_maxThe larger the value will be.

The process of determining the fusion weight of the amplitude data and the fusion weight of the phase data is specifically illustrated as follows.

In the embodiment of the invention, the A-type data is phase data, and the B-type data is amplitude data; the class a data is amplitude data and the class B data is phase data. The hardware device used in the embodiment of the present invention can collect 3 sets of amplitude data and phase data at each position through 3 antennas, respectively, and the positions of the 3 antennas do not change during data collection. Then for both the amplitude data and the phase data, if the correlation between 3 sets of a-type data collected by 3 antennas at the position of a reference point is smaller than that of the B-type data, which indicates that the position distinction degree of the a-type data at the position of the reference point is stronger than that of the B-type data, it can be considered that the weight of the B-type data is lower than that of the a-type data when the fingerprint data of the position is established. Through the above analysis, the embodiment of the present invention considers the correlation between the data as the reference of the fusion weight. One type of data (i.e., amplitude data or phase data) collected at the location of one reference point may be represented after preprocessing as follows:

D_m＝{D_m,1,D_m,2,...D_m,n,...,D_m,30}

vector data for 3 antennas correspondences D₁,D₂,D₃Before correlation analysis, the embodiment of the invention standardizes the vector to obtain standardized vector data { eta [. eta. ]₁,η₂,η₃}. Then, the covariance calculation is performed on the 3 sets of normalized vector data, and finally the covariance matrix is formed as follows:

due to the covariance matrix

Is symmetrical and the values on the diagonal are all 1 because of the process of prior normalization. Then, the covariance matrix is subjected to eigenvalue calculation to obtain an eigenvalueλ₁,λ₂,λ₃. By analyzing the characteristics of the covariance matrix and the vector data, the embodiment of the invention knows that the eigenvalue of the covariance matrix reflects the category distribution of the multi-dimensional data characteristics, and if the correlation between the data is stronger, the category distribution is more concentrated, otherwise, the category distribution is more dispersed. Since the sum of the element values of the main diagonal of the matrix is constantly 3, the sum of the eigenvalue values is also constantly 3. Under the same condition, the concentration degree of the numerical values of the characteristic values can be used as vector data eta₁,η₂,η₃The correlation between the two is reflected. After obtaining the eigenvalue of the covariance matrix, the embodiment of the invention obtains the maximum eigenvalue through the following formula:

λ_max＝max(λ₁,λ₂,λ₃)

wherein, by the above analysis,. lambda._maxCan be used as a parameter for measuring the correlation between the data collected by the three antennas. Considering that the correlation between the data and the weight value are negative correlation, the calculation of the dynamic weight value may be:

and fifthly, carrying out weighted fusion on the preprocessed amplitude data and the preprocessed phase data of each reference point to obtain the offline CSI position fingerprint of each reference point.

After the CSI data of one reference point is preprocessed, the amplitude data preprocessed on the kth path of subcarriers is assumed to be Amp_pre(f_k) The preprocessed phase data is Pha_pre(f_k) Then the following formula is adopted:

DFF(f_k)＝α*Amp_pre(f_k)+β*Pha_pre(f_k) And performing weighted fusion on the preprocessed amplitude data and the preprocessed phase data of each reference point to obtain the offline CSI position fingerprint of each reference point.

Wherein, DFF (f)_k) For the offline CSI position fingerprint of each reference point, α is the amplitudeAnd the fusion weight of the data and the phase data represents that the proportion beta of the adopted amplitude data in the fusion characteristic data is the fusion weight of the phase data and represents the proportion of the phase data in the fusion characteristic data. α and β also numerically satisfy the following constraints:

the new off-line CSI location fingerprint DFF in the embodiment of the present invention is mainly characterized in that α and β obtained by weighted fusion both have location specificity.

Through the data correlation analysis and the calculation of the dynamic weight described above, the corresponding weight values of the amplitude data and the phase data can be obtained at the position of each reference point, and finally, the fused data is formed. Because the weight value is dynamic and has position specificity, the characteristics and robustness of the position fingerprint data are improved to a certain extent, and the accuracy of positioning is improved. The embodiment of the invention finally sorts and summarizes the data of the three antennas to form new CSI position fingerprint data DFF:

wherein,

and representing the fused fingerprint data on the jth subcarrier collected by the ith antenna. Since it is assumed that three sets of amplitude data and phase data are collected, the DFF_csiIs a 1 x 90 vector data representing CSI fingerprint information at a fixed location.

In the process of determining the offline CSI location fingerprint in the database, the following steps are adopted to determine the matching threshold:

averaging the variance data to obtain an average variance;

according to the Laplace criterion, the following formula is adopted:

obtaining the matching threshold;

in the embodiment of the present invention, the above-mentioned rale criterion is mainly used as a determination method of a dynamic matching threshold, and according to a formula in EDR content, a distance between two points to be matched (i.e. an online CSI location fingerprint and an offline CSI location fingerprint) is defined to be less than three times of a standard deviation, and it can be determined that the two points to be matched are successfully matched.

Wherein, epsilon is a matching threshold value,

n is the total number of subcarriers contained in the fingerprint data, M is the number of acquired data packets, L_ijI is the data packet sequence number, j is the subcarrier sequence number sigma which is the summation symbol,

is the average value of all data packets on the jth sub-carrier.

In order to complete the matching between the on-line CSI location fingerprint of the to-be-located point and the off-line CSI location fingerprint in the database at the on-line matching stage, a similarity calculation method or a clustering method may be generally adopted to match the on-line CSI location fingerprint of the to-be-located point with the off-line CSI location fingerprint in the database, so as to obtain the off-line CSI location fingerprint most matched with the on-line CSI location fingerprint.

When the similarity calculation method or the clustering method is adopted, in combination with the above embodiment, the matching threshold that can be adopted is preset, and the offline CSI location fingerprint whose similarity reaches the matching threshold is used as the offline CSI location fingerprint that is most matched with the online CSI location fingerprint. The similarity calculation method includes, but is not limited to: and (4) an EDR algorithm. The matching threshold value can be a fixed matching threshold value set according to the needs of the user. Therefore, one or more offline CSI position fingerprints with the similarity reaching the matching threshold can be obtained, and the positioning result of the to-be-positioned point is obtained through subsequent averaging.

In order to better reduce the influence caused by noise, the embodiment of the invention constructs an EDR-based similarity measurement method IEDR. By the idea of editing Distance (ED for short), on one hand, a one-to-many data mapping mode is adopted and abnormal data points are allowed to be thrown out, and on the other hand, the Distance between elements is quantized to 0 and 1, so that the superposition of abnormal values is avoided, and the influence of abnormal data can be reduced. The IEDR is then formed by changing the fixed matching threshold in EDR to a dynamic parameter that is influenced by the environment, i.e. (the matching threshold described below), further improving the environmental suitability of the fingerprint matching phase.

In order to dynamically set a matching threshold in an EDR algorithm, a measurement index IEDR with environmental adaptability is established, and the position of a to-be-positioned point is obtained by comparing and matching. Therefore, as shown in fig. 3, in step 150 of the embodiment of the present invention, the following possible implementation manners may be adopted to match the online CSI location fingerprint of the to-be-located point with the offline CSI location fingerprint in the database, so as to obtain the offline CSI location fingerprint that is most matched with the online CSI location fingerprint, and specifically include:

For the above steps, referring to fig. 2, in step 151, the similarity between the on-line CSI location fingerprint and the off-line CSI location fingerprint in the database may be obtained by calculating the similarity between the on-line CSI location fingerprint and the off-line CSI location fingerprint by using EDR. Step 152, a matching threshold is obtained. And step 153, taking the offline CSI location fingerprint with the similarity reaching the matching threshold as the offline CSI location fingerprint which is most matched with the online CSI location fingerprint.

Specifically, the step 151 further includes, but is not limited to:

determining an online CSI location fingerprint and an offline CSI location fingerprint;

the determined on-line CSI position fingerprint and the determined off-line CSI position fingerprint refer to the on-line CSI position fingerprint and the off-line CSI position fingerprint which participate in matching measurement.

Performing EDR calculation on the online CSI position fingerprint and the offline CSI position fingerprint to obtain an EDR distance between the online CSI position fingerprint and the offline CSI position fingerprint;

and taking the reciprocal of the EDR distance obtained by calculation as the similarity between the on-line CSI position fingerprint and the off-line CSI position fingerprint based on the inverse relation between the EDR distance of the fused fingerprint data and the similarity between the fingerprints.

In step 152, the following steps are adopted to determine the matching threshold:

averaging the variance data to obtain an average variance;

according to the Laplace criterion, the following formula is adopted:

obtaining the matching threshold;

in the embodiment of the present invention, the above-mentioned rale criterion is mainly used as a determination method of a dynamic matching threshold, and according to a formula in EDR content, a distance between two points to be matched (i.e. an online CSI location fingerprint and an offline CSI location fingerprint) is defined to be less than three times of a standard deviation, and it can be determined that matching is successful.

Wherein, epsilon is a matching threshold value,

is the average value of all data packets on the jth sub-carrier.

In the embodiment of the invention, in the aspect of fingerprint data, the relationship between amplitude data and phase data collected by different antennas is considered, and the amplitude data and the phase data are fused by weighting. In the aspect of similarity measurement indexes, similarity measurement is carried out by using an EDR algorithm, a matching threshold value in the EDR algorithm is dynamically set, a measurement index IEDR with environmental adaptability is established, and the position of a to-be-positioned point is obtained by comparing and matching.

In the embodiment of the invention, by configuring the parameter with environmental adaptability, namely the matching threshold, the problem of feature fuzzification caused by the introduction of random noise information in the fingerprint matching stage of the fingerprint positioning system is reduced, the feature resolution of the fingerprint positioning method is improved, the indoor positioning precision is also improved, and meanwhile, theoretical reference is provided for the improvement of the positioning precision of the indoor positioning method.

EDR is an improved computational metric based on ED, which is used to measure the similarity between two sequences and is widely used in bioinformatics and speech recognition. ED (A, B) is a measure of the number of operations (insertions, productions and replacements) required to convert string A to B. Since matching of sequences of traces is actually required rather than strings, in the definition of EDR, it is assumed that r is_iAnd s_jA pair of trajectory elements above the trajectory vectors R and S, respectively, and defining a matching threshold epsilon, the elements of the trajectory vectors R and S satisfying the following equation:

wherein, R and S represent two track vector data participating in the EDR distance measurement, and are used for describing the calculation process of the EDR distance. match () is used to determine whether a match criterion, d (r), is met between two elements_i,s_j) I.e. the euclidean distance between two element points. On the basis of the above formula, the EDR between trajectory vectors R and S is defined as follows:

wherein R is_nRepresenting a trajectory vector consisting of the first n elements of the trajectory vector R, S_mRepresents the track vector consisting of m elements before the track vector S, in match (r)_n,s_m) When true, ρ is 0, otherwise ρ is 1. And (3) obtaining an EDR value EDR (R, S) between two groups of complete track vectors R and S by accumulating and calculating the EDR distance, and finishing the measurement of the similarity of one group of track vectors.

After collecting and processing the online data of unknown locations, the EDR is used to perform similarity calculations between the online data fingerprints and the offline location fingerprint database. Since the distance value obtained by EDR represents the shortest edit distance of two pieces of CSI-fused fingerprint data, it is in inverse relation to the similarity between fingerprints. By calculating the EDR of two CSI-fused fingerprints, the similarity between the two fingerprints is then derived, as shown in the following formula:

where γ (P, Q) represents the similarity degree calculated by the EDR method of the fingerprint data P and Q. Theta denotes a number infinitely close to zero and greater than zero in order to avoid the occurrence of a denominator of zero. By using the similarity measurement, the fingerprint data can be matched in the CSI fingerprint positioning method of the embodiment of the present invention to confirm the position. In the embodiment of the invention, the online stage fingerprint is compared with the offline stage fingerprint by using the formula, so that the matching degree is measured. P, Q show in this formula are two sets of fingerprint data that participate in the matching metric.

In the EDR calculation process, the matching threshold epsilon is a criterion for judging whether a pair of data points on two sets of trajectory vectors belong to "matching success". Therefore, in the present invention, when the EDR is used to calculate the similarity, the static matching threshold is obviously not completely adaptive to the complex indoor environment. Therefore, when the IEDR measurement index based on the EDR is established, the selection standard of the dynamic matching threshold epsilon is established, and the matching accuracy can be further improved compared with the fixed parameter.

Considering that a dynamic matching threshold needs to adapt to the signal fluctuation degree of the online phase data, the method reflects the signal fluctuation degree by acquiring the variance data on each path of subcarrier of the online phase data, and then acquires the average variance of the online data on one position by utilizing an averaging mode. The EDR formula above has a parameter ρ that is 0 when the matching threshold is met, and does not meet 1. The following

And epsilon is a measure of the dynamic match threshold. They are integral. The specific formula is as follows:

where M denotes the number of packets collected at the on-line stage, N denotes the total number of subcarriers included in the fingerprint data, and L_ijIndicating the size of the jth sub-carrier data on the ith packet,

represents the average value of all data packets on the jth sub-carrier,

representing the average variance of the on-line measurement data over all subcarriers. By the above calculation, becauseThe number of the data packets collected in the online stage is enough, the difference value of the distances between the online data and the fingerprint data on each path of subcarrier at the same position approximately conforms to the variance of

A gaussian distribution with a mean value of 0.

Determining a matching threshold value by adopting the following steps: calculating the variance of the amplitude data and the phase data collected on each path of subcarrier to obtain variance data; averaging the variance data to obtain an average variance; according to the rule of ladida (the Pauta criterion) (or the 3 σ criterion), the following formula is adopted:

obtaining the matching threshold epsilon;

wherein, epsilon is a matching threshold value,

the average variance collected over all subcarriers for the amplitude data and phase data, as a whole,

is the average value of all data packets on the jth sub-carrier.

In summary, by generating a dynamic matching threshold epsilon, a similarity measure index IEDR based on EDR distance is established, and the method can better analyze and match the characteristics of the signals in a complex environment.

The following description is continued on a CSI indoor fingerprint positioning apparatus based on dynamic fusion features provided in an embodiment of the present invention.

Referring to fig. 4, fig. 4 is a schematic structural diagram of a CSI indoor fingerprint location apparatus based on dynamic fusion features according to an embodiment of the present invention. The CSI indoor fingerprint positioning device based on the dynamic fusion characteristics provided by the embodiment of the invention can comprise the following modules:

the acquisition module 21 is configured to acquire channel state information CSI online data of different antennas acquiring to-be-located points; wherein the CSI-on-line data provides amplitude data and phase data collected on a plurality of subcarriers;

an analysis module 22, configured to analyze a correlation between amplitude data and phase data collected on multiple subcarriers to obtain a data maximum characteristic value, where the data maximum characteristic value includes: maximum eigenvalue of amplitude data and maximum eigenvalue of phase data; the data maximum characteristic value is used for measuring the correlation between amplitude data and phase data acquired by different antennas;

the first processing module 23 is configured to calculate a ratio of a maximum eigenvalue of the amplitude data to a maximum eigenvalue of the phase data in the maximum eigenvalue of the data, and obtain a fusion weight of the amplitude data and a fusion weight of the phase data;

the weighted fusion module 24 is configured to perform weighted fusion on the amplitude data and the phase data of the to-be-located point by using a fusion weight of the amplitude data and a fusion weight of the phase data to obtain fusion fingerprint data, which is used as an online CSI position fingerprint of the to-be-located point;

the matching positioning module 25 is configured to match the online CSI location fingerprint of the to-be-positioned point with the offline CSI location fingerprint in the database to obtain an offline CSI location fingerprint that is most matched with the online CSI location fingerprint, so as to determine the position of the to-be-positioned point; the off-line CSI position fingerprints in the database are obtained by preprocessing and correlation analysis of CSI measurement data of each reference point acquired by different antennas to obtain fusion weights of amplitude data and phase data, and performing weighted fusion on the preprocessed amplitude data and preprocessed phase data of each reference point to obtain off-line CSI position fingerprints of each reference point; the offline CSI location fingerprint that most matches the online CSI location fingerprint is most similar to the online CSI location fingerprint;

in one possible implementation manner, the matching location module is configured to:

In one possible implementation, the apparatus further includes:

a second processing module for determining a matching threshold using the steps of:

averaging the variance data to obtain an average variance;

according to the Laplace criterion, the following formula is adopted:

obtaining the matching threshold;

wherein, epsilon is a matching threshold value,

is the average value of all data packets on the jth sub-carrier.

In one possible implementation, the apparatus further includes: the database building module is used for determining the offline CSI position fingerprint in the database by adopting the following steps:

In a possible implementation manner, the library building module is configured to:

In one possible implementation, the apparatus further includes:

a third processing module for employing the following formula:

wherein, C_maxTaking the maximum value mu as the number of samples contained in the ith data cluster, wherein { } is the set of data clusters, and num (·) is the number C of data in the data clusters_iThe ith data cluster i is the serial number of the data cluster;

the following formula is adopted:

wherein, Amp_pre(f_k) For pre-processing the amplitude data on the k-th sub-carrier, Amp_preFor the preprocessed amplitude data, pre is an expression mode of preprocessed data symbols, k is a serial number collected on a subcarrier, and Amp_j(f_k) For the largest data cluster collected on the k-th sub-carrierJth amplitude data, Amp_jJ is the j th amplitude data in the data cluster and is the serial number of the amplitude data, N_fkNumber of samples contained for the largest data cluster among different sub-carriers, f_kIs the k path sub-carrier;

the following formula is adopted:

wherein,

the preprocessed phase data, a is the slope,

is the phase data of the last sub-carrier,

is the phase data of the first sub-carrier, K_LIndex value K of last subcarrier₁Is the index value of the first subcarrier, L is the number of subcarriers,

for test phase data on the k-th sub-carrier, psi_kFor true phase values, pi is 180 degrees radian, K_kIs the index value collected on the k path sub-carrier, N is the size of fast Fourier transform, delta is the time lag caused by sampling frequency shift, beta is the unknown phase shift caused by carrier frequency shift, Z is the measurement noise, b is the shift amount,

k is the number of the subcarrier

Is the preprocessed phase data.

The following continues to describe the electronic device provided by the embodiment of the present invention.

Referring to fig. 5, fig. 5 is a schematic structural diagram of an electronic device according to an embodiment of the present invention. The embodiment of the present invention further provides an electronic device, which includes a processor 31, a communication interface 32, a memory 33 and a communication bus 34, wherein the processor 31, the communication interface 32 and the memory 33 complete mutual communication through the communication bus 34,

a memory 33 for storing a computer program;

the processor 31 is configured to implement the steps of the CSI indoor fingerprint location method based on the dynamic fusion feature when executing the program stored in the memory 33, and in a possible implementation manner of the present invention, the following steps may be implemented:

The communication bus mentioned in the electronic device may be a PCI (Peripheral Component Interconnect) bus, an EISA (Extended Industry Standard Architecture) bus, or the like. The communication bus may be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, only one thick line is shown, but this does not mean that there is only one bus or one type of bus.

The communication interface is used for communication between the electronic equipment and other equipment.

The Memory may include a RAM (Random Access Memory) or an NVM (Non-Volatile Memory), such as at least one disk Memory. Optionally, the memory may also be at least one memory device located remotely from the processor.

The Processor may be a general-purpose Processor, including a Central Processing Unit (CPU), a Network Processor (NP), and the like; but also a DSP (Digital Signal Processing), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array) or other Programmable logic device, discrete Gate or transistor logic device, discrete hardware component.

An embodiment of the present invention provides a computer-readable storage medium, in which a computer program is stored, and the computer program, when executed by a processor, implements the steps of the above CSI indoor fingerprint positioning method based on dynamic fusion features.

Embodiments of the present invention provide a computer program product comprising instructions which, when run on a computer, cause the computer to perform the steps of the above-described CSI indoor fingerprint location method based on dynamic fusion features.

Embodiments of the present invention provide a computer program, which when running on a computer, causes the computer to perform the steps of the above-mentioned CSI indoor fingerprint location method based on dynamic fusion features.

It is noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.

All the embodiments in the present specification are described in a related manner, and the same and similar parts among the embodiments may be referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, for the apparatus/electronic device/storage medium/computer program product/computer program embodiment comprising instructions, the description is relatively simple as it is substantially similar to the method embodiment, and reference may be made to some descriptions of the method embodiment for relevant points.

The above description is only for the preferred embodiment of the present invention, and is not intended to limit the scope of the present invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention shall fall within the protection scope of the present invention.

Claims

1. A CSI indoor fingerprint positioning method based on dynamic fusion characteristics is characterized by comprising the following steps:

matching the on-line CSI position fingerprint of the to-be-positioned point with the off-line CSI position fingerprint in the database to obtain an off-line CSI position fingerprint which is most matched with the on-line CSI position fingerprint so as to determine the position of the to-be-positioned point; the off-line CSI position fingerprints in the database are obtained by preprocessing and correlation analysis of CSI measurement data of each reference point acquired by different antennas to obtain fusion weights of amplitude data and phase data, and performing weighted fusion on the preprocessed amplitude data and preprocessed phase data of each reference point to obtain off-line CSI position fingerprints of each reference point; the offline CSI location fingerprint that most matches the online CSI location fingerprint is most similar to the online CSI location fingerprint;

the matching of the on-line CSI position fingerprint of the to-be-positioned point and the off-line CSI position fingerprint in the database to obtain the off-line CSI position fingerprint most matched with the on-line CSI position fingerprint comprises:

calculating the similarity between the on-line CSI position fingerprint and the off-line CSI position fingerprint in the database by using EDR to obtain the off-line CSI position fingerprint with the maximum similarity to the on-line fingerprint, and using the off-line CSI position fingerprint as the off-line CSI position fingerprint which is most matched with the on-line CSI position fingerprint;

the calculating, by using EDR, a similarity between the online CSI location fingerprint and the offline CSI location fingerprint in the database to obtain an offline CSI location fingerprint having a maximum similarity with the online fingerprint, as an offline CSI location fingerprint that is most matched with the online CSI location fingerprint, includes:

acquiring a matching threshold;

the method comprises the following steps of:

averaging the variance data to obtain an average variance;

according to the Laplace criterion, the following formula is adopted:

obtaining the matching threshold;

wherein, epsilon is a matching threshold value,

is the average value of all data packets on the jth sub-carrier.

2. The method of claim 1, wherein the off-line CSI location fingerprint in the database is determined by:

3. The method of claim 2, wherein the calculating amplitude data maximum eigenvalues and phase data maximum eigenvalues, respectively, the fraction among the data maximum eigenvalues, and the determining the fusion weights for the amplitude data and the fusion weights for the phase data, comprises:

4. A CSI indoor fingerprint positioning apparatus based on dynamic fusion features, comprising:

the matching positioning module is used for matching the on-line CSI position fingerprint of the to-be-positioned point with the off-line CSI position fingerprint in the database to obtain an off-line CSI position fingerprint which is most matched with the on-line CSI position fingerprint so as to determine the position of the to-be-positioned point; the off-line CSI position fingerprints in the database are obtained by preprocessing and correlation analysis of CSI measurement data of each reference point acquired by different antennas to obtain fusion weights of amplitude data and phase data, and performing weighted fusion on the preprocessed amplitude data and preprocessed phase data of each reference point to obtain off-line CSI position fingerprints of each reference point; the offline CSI location fingerprint that most matches the online CSI location fingerprint is most similar to the online CSI location fingerprint;

the matching positioning module is used for:

acquiring a matching threshold;

the method comprises the following steps of:

averaging the variance data to obtain an average variance;

according to the Laplace criterion, the following formula is adopted:

obtaining the matching threshold;

wherein, epsilon is a matching threshold value,

is the average value of all data packets on the jth sub-carrier.

5. The apparatus of claim 4, wherein the apparatus further comprises: an establishing module, configured to determine an offline CSI location fingerprint in a database by using the following steps:

6. An electronic device, comprising a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus;

the memory is used for storing a computer program;

the processor, when executing the program stored in the memory, implementing the method steps of any of claims 1-3.