CN114710742A - Indoor positioning method for constructing fingerprint map based on multi-chain interpolation - Google Patents

Indoor positioning method for constructing fingerprint map based on multi-chain interpolation Download PDF

Info

Publication number
CN114710742A
CN114710742A CN202210188290.1A CN202210188290A CN114710742A CN 114710742 A CN114710742 A CN 114710742A CN 202210188290 A CN202210188290 A CN 202210188290A CN 114710742 A CN114710742 A CN 114710742A
Authority
CN
China
Prior art keywords
point
sampling
chain
fingerprint map
data set
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202210188290.1A
Other languages
Chinese (zh)
Inventor
杨浩
朱立才
赵泳浩
季衍辉
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Yancheng Teachers University
Original Assignee
Yancheng Teachers University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Yancheng Teachers University filed Critical Yancheng Teachers University
Priority to CN202210188290.1A priority Critical patent/CN114710742A/en
Publication of CN114710742A publication Critical patent/CN114710742A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
    • H04W4/02—Services making use of location information
    • H04W4/023—Services making use of location information using mutual or relative location information between multiple location based services [LBS] targets or of distance thresholds
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B17/00—Monitoring; Testing
    • H04B17/30—Monitoring; Testing of propagation channels
    • H04B17/309—Measuring or estimating channel quality parameters
    • H04B17/318—Received signal strength
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
    • H04W4/02—Services making use of location information
    • H04W4/021—Services related to particular areas, e.g. point of interest [POI] services, venue services or geofences
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
    • H04W4/30—Services specially adapted for particular environments, situations or purposes
    • H04W4/33—Services specially adapted for particular environments, situations or purposes for indoor environments, e.g. buildings
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W64/00—Locating users or terminals or network equipment for network management purposes, e.g. mobility management
    • H04W64/006—Locating users or terminals or network equipment for network management purposes, e.g. mobility management with additional information processing, e.g. for direction or speed determination

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Quality & Reliability (AREA)
  • Physics & Mathematics (AREA)
  • Electromagnetism (AREA)
  • Position Fixing By Use Of Radio Waves (AREA)

Abstract

The invention discloses an indoor positioning method for constructing a fingerprint map based on multi-chain interpolation, belonging to the technical field of wireless equipment positioning method research, and comprising the following steps: sampling a sample target according to a sampling rule to obtain a sampling point data set, wherein the sampling point data set comprises position coordinate information of a sampling point and an RSS value of the sampling point; inputting the sampling point data set into the system for constructing the fingerprint map based on the multi-chain interpolation; carrying out internal processing by a system for constructing a fingerprint map based on multi-chain interpolation to obtain a complete fingerprint map, wherein the complete fingerprint map comprises a sampling point data set and a to-be-inserted point data set, and the to-be-inserted point data set comprises position coordinate information of the to-be-inserted point and an RSS value of the to-be-inserted point; the position coordinate information of the point to be detected is obtained through the complete fingerprint map and the RSS value of the point to be detected.

Description

Indoor positioning method for constructing fingerprint map based on multi-chain interpolation
Technical Field
The invention belongs to the technical field of wireless equipment positioning method research, and particularly relates to an indoor positioning method for constructing a fingerprint map based on multi-chain interpolation.
Background
Location-based services are becoming an indispensable service in daily life, and are generally classified into outdoor positioning and indoor positioning according to implementation scenarios. The traditional positioning method comprises a method for positioning by using a satellite positioning system, the method for positioning by using the satellite positioning system comprises the deployment of the satellite positioning system (such as Beidou, GPS and the like), in an outdoor positioning scene, the deployment of the satellite positioning system (such as Beidou, GPS and the like) enables outdoor positioning and navigation to be well applied, the positioning accuracy is very high, and the sub-meter level can be achieved. In an indoor positioning scene, due to interference of various indoor obstacles on satellite signals, the satellite cannot realize indoor accurate positioning.
The widespread deployment of wireless devices in indoor environments makes indoor location-based services a research hotspot. For the research of indoor positioning methods, researchers usually use wireless signal fingerprints to perform matching positioning methods. The basic idea is to use the spatial difference of Access Points (APs) at different positions and use the wireless signal characteristics of the reference points as the characteristics of the physical positions to determine the positions. The wireless Signal characteristic is called Fingerprint (fingerprintt) characteristic, wherein Received Signal Strength (RSS) is a typical Fingerprint characteristic. Indoor positioning based on received signal strength is a key for providing accurate location service, and the method firstly needs to construct a fingerprint map (Site Survey) for a sensing area, needs to acquire signal data from a large number of reference points of the sensing area, and is extremely time-consuming and labor-consuming. In order to solve the problem, researchers collect the Signal Received Strength (RSS) of part of reference points and use the RSS to interpolate surrounding non-sampling points to construct a map of the whole area. In order to meet the positioning requirements of large-scale scenes, such as large commercial and trade centers, parking lots, conference administration centers and the like, huge measurement cost is needed, and meanwhile, in order to improve positioning accuracy, signal data of more reference points are required to be sampled as far as possible. Meanwhile, due to the inherent characteristics of the wireless signals, the signals of the reference points can change along with time, so that fingerprints at the same position are inconsistent in a sampling stage and a testing stage, and cannot be matched. For this reason, the fingerprint repository needs to be maintained and updated at intervals, which further increases the cost of location. Therefore, it is crucial how to reduce the workload of the fingerprint database establishment and update process while ensuring the positioning accuracy.
For the fingerprint positioning method, a lot of time and labor cost are needed for creating or updating the map. In actual sampling, when the positioning scene is large, it is obviously not practical to densely sample the perceptual environment. To solve this problem, researchers have constructed maps in a way that the fingerprint maps are virtualized. Specifically, when the map is created, the sensing environment is not comprehensively sampled, only partial point signals are collected, and then the values are used for simulating the signals of the non-sampled positions to obtain the complete fingerprint map through expansion, so that the sampling amount can be greatly reduced, and the sampling cost is saved. When the area of the positioning scene is larger, resources saved by fingerprint map virtualization are more, and the importance of the resources can be reflected more. This way, the workload of the map construction and updating process is reduced, and the positioning precision is ensured to a certain extent. In the fingerprint virtualization construction, fingerprint map construction facing WiFi and radio frequency is a representative mode.
In WiFi fingerprint mapping, the mapping method based on crowd sensing is still the most common method to alleviate this problem at present. For example, more users are encouraged to participate in the crowdsourcing process in an incentive way, so that the users actively collect information; the data is sampled by a user under unconscious participation in an implicit or transparent mode, and the aim of collecting the data explicitly is fulfilled without exciting the user; by sampling a small amount of fingerprint data with identification (namely a data set containing RSS and position information) and a large amount of fingerprint data without identification (namely a data set containing only RSS information but no position information), the effects of reducing the workload and making up for the positioning accuracy are achieved. On the other hand, many scholars build maps based on models that fill the entire sampling area by fitting signals on other, non-sampled reference points from the model, primarily through a small amount of sampled data. For example, an algorithm of region segmentation and curve fitting is provided through a polynomial model, and a fitting function is obtained on each sub-region based on a small number of reference points so as to obtain an unsampled point signal; estimating the position of an Access Point (AP) through the weighted centroid of the sampling Point fingerprint, and evaluating the signal transmission power and the parameters of a path loss model by using a Bayesian method. Fitting the existing small amount of sampling data through Gaussian process regression to generate a mean value curved surface and a standard deviation curved surface of a local area, and generating particles by combining a particle filter to assign weights; and in order to better update the fingerprint map, Gaussian process regression is combined with K-means.
In the construction of a radio frequency fingerprint map, the LANDMARC algorithm is used for assisting in updating the fingerprint database on line by adding landmarks based on an RFID positioning technology, and the influence of a dynamic environment can be responded without acquiring the map in advance. Based on this, many scholars have proposed many directions for improvement. The VIRE algorithm is a virtual tag positioning algorithm based on the LANDMARC algorithm, positioning accuracy is improved by 13-73% in different indoor environments, virtual reference tags are introduced, virtual tags are added among fixed reference tags, the virtual tags achieve the same effect as the fixed reference tags, and excessive number of virtual tags does not cause signal interaction. Therefore, the use of spatial interpolation algorithms to reduce the amount of sampling effort during the training phase is increasingly accepted by more researchers. In order to increase the density of the fingerprints, a virtual sampling fingerprint is generated by using a linear interpolation method, and an accurate position is estimated by using a Bayes algorithm, so that compared with the traditional Bayes algorithm, the time complexity of O (4n/5) is reduced, and the positioning precision is improved by 6%; the multi-path error is eliminated by introducing a variable coefficient into the traditional Leide criterion, and then virtual reference marking is carried out by using natural neighborhood interpolation, so that the stability and the accuracy of virtual label similarity measurement are improved, and 80% of errors of the test labels are accumulated within 4cm and 90% of errors are accumulated within 10 cm; the influence of the kriging interpolation algorithm and several deformation algorithms on the map estimation accuracy is also provided and compared; when a fingerprint database is constructed, wall penetration loss and distance power loss are estimated, then RSS values are inserted based on the loss models, an interpolation method of spatial interpolation is applied to perfect a fingerprint map, and the average positioning precision is improved by about 25%; in the traditional RFID virtual tag positioning, dynamic particles are used for replacing a traditional static reference tag, a particle swarm optimization algorithm is used for updating a Monte Carlo sample particle swarm, different weights are given based on the signal intensity difference between the sampling particles and an undetermined tag, and therefore the interpolation effect of the unknown tag is achieved; also for the construction of virtual maps, it proposes a region-based repair method, mainly by limiting the path loss exponent for constructing the virtual reference point and limiting the region to minimize the estimation error. The methods relate to the idea of interpolation and virtual points more or less, and the workload of constructing the fingerprint map is optimized to a certain extent. However, the above methods have certain limitations, and are difficult to be directly applied to WiFi fingerprint positioning.
Therefore, a new fingerprint map construction method is needed to realize indoor positioning.
Disclosure of Invention
The purpose of the invention is as follows: in order to overcome the defects that the prior art cannot simultaneously improve the precision and the efficiency and the like, the invention provides an indoor positioning method for constructing a fingerprint map based on multi-chain interpolation.
The technical scheme is as follows: in order to achieve the above object, the present invention provides an indoor positioning method for constructing a fingerprint map based on multi-chain interpolation, which comprises the following steps: the method provides a system for constructing a fingerprint map based on multi-chain interpolation, and the method comprises the following steps:
s1, sampling the sample target according to the sampling rule, and sampling to obtain a sampling point data set, wherein the sampling point data set comprises position coordinate information of a sampling point and an RSS value of the sampling point;
s2, inputting the sample point data set into the system for constructing the fingerprint map based on the multi-chain interpolation;
s3, internally processing by a system for constructing a fingerprint map based on multi-chain interpolation to obtain a complete fingerprint map, wherein the complete fingerprint map comprises a sampling point data set and a to-be-inserted point data set, and the to-be-inserted point data set comprises position coordinate information of the to-be-inserted point and an RSS value of the to-be-inserted point;
s4 obtaining the position coordinate information of the point to be detected through the complete fingerprint map and the RSS value of the point to be detected
Further, the system for constructing the fingerprint map based on the multi-chain interpolation comprises:
an input module: inputting a sampling point data set into the system for constructing the fingerprint map based on the multi-chain interpolation;
an RSS value estimation module of a point to be inserted: the RSS value is used for calculating the RSS value of the point to be inserted;
a position coordinate calculation module of the point to be inserted: the device is used for calculating the position coordinates of a point to be inserted;
data set union module: taking the RSS value of the point to be inserted and the position coordinate information of the point to be inserted as a point data set to be inserted, and combining the point data set to be inserted and the sampling point data set to be used as a complete fingerprint map;
an output module: for outputting a complete fingerprint map.
Further, the RSS value estimation module to be inserted is configured to calculate an RSS value of the point to be inserted, and the RSS value estimation module to be inserted specifically includes:
a direction valuation module: determining an AP signal propagation path corresponding to a point to be inserted, and performing interpolation calculation on the point to be inserted in each AP signal propagation path direction to obtain an RSS value of the point to be inserted in each direction chain;
the strength estimation module: obtaining weights of the directional chains through the RSS values of the to-be-inserted point on the directional chains by using an inverse distance weighting method, and calculating an estimated RSS value of the to-be-inserted point through the weights of the directional chains;
the weight optimization module: fitting the adjacent sampling point data set of the point to be inserted on each directional chain to a signal attenuation model on each directional chain, calculating the fitting RSS value of the adjacent sampling point of the point to be inserted on each directional chain through the signal attenuation model, calculating the error of the signal attenuation model on each directional chain through the fitting RSS value of the adjacent sampling point, and recalculating the weight on each directional chain through the error of the signal attenuation model on each directional chain;
and an adjustment valuation module: recalculating the RSS value of the point to be inserted according to the RSS value of the point to be inserted on each directional chain and the weight of the point to be inserted on each directional chain;
an iteration expansion module: and repeating the module steps to obtain the RSS values of all points to be inserted.
Further, the position coordinate calculation module of the point to be inserted calculates the position coordinate of the point to be inserted through sensing environment pre-definition.
Further, the sampling rule adopts a top angle sampling rule or a top angle center full sampling rule.
Furthermore, the number of AP signal propagation paths corresponding to the point to be inserted is 4.
Furthermore, the adjacent sampling points of the point to be inserted on each directional chain at least comprise 2.
Has the advantages that: compared with the prior art, the invention has the advantages that:
1. according to the method, the signal of the insertion point is formed by overlapping a plurality of direction signals according to the propagation characteristics of the wireless signal, the relevance of the adjacent points at the periphery cannot be considered, and the insertion point is expanded facing different directions.
2. Compared with the traditional sampling mode, the method designs a Multi-Chain Interpolation fingerprint map construction Method (MCI), and estimates each point to be inserted by utilizing a plurality of direction chains according to the signal propagation characteristics in the actual environment. Specifically, interpolation calculation is performed in different directions under a given sampling rule, and a pre-estimated value of a point to be inserted is obtained by using inverse distance weighting. Using this value, a corresponding signal attenuation model is fitted in each direction and the error is calculated. And acquiring the directional weight again and obtaining an estimation value of the point to be inserted. Finally, a fingerprint database consisting of real points and virtual points is formed, and the maintenance and updating cost of the fingerprint database is effectively reduced. According to experimental verification, the MCI method improves the positioning accuracy while reducing the measurement workload.
3. The method of the invention uses two Sampling modes (the 'Corner Sampling' and the 'Corner & Centre Sampling'), and only a small amount of fingerprint point signals are required to be collected to construct a complete fingerprint map of a sensing area. The sampling amount of the two modes is respectively 25% and 50% of that of the full sampling, namely, the workload of the fingerprint map construction process is respectively reduced by 75% and 50%, and the workload of the map construction and updating process is greatly reduced.
4. According to large-scale experimental verification, the map construction method MCI based on multi-chain interpolation improves the positioning precision and has good stability. Compared with the full sampling mode, the positioning accuracy of the two MCI sampling modes is respectively improved by 13.58% and 4.74%, and compared with the classical interpolation mode, the MCI method has better stability and more obvious advantages especially when the sampling quantity is lower. When the sampling amount is only 25% of the full sampling amount, the average accuracy is improved by 18.50% compared with the classical interpolation mode, and the method has good expansibility.
5. The method provided by the invention aims at map construction of WiFi fingerprint positioning, reduces cost consumption, fully considers signal propagation characteristics, evaluates signal superposition conditions of different links, constructs an accurate fingerprint map, and improves positioning accuracy and stability.
Drawings
FIG. 1 is a flow chart of the method steps of the present invention.
FIG. 2 is a diagram of a system for constructing a fingerprint map based on multi-chain interpolation according to the present invention.
Fig. 3 is a structural diagram of an RSS value estimator of the present invention.
Fig. 4 is a schematic diagram of the present invention when the propagation path of the AP signal corresponding to the point to be inserted is 4.
Fig. 5 is a schematic diagram of the present invention employing corner sampling rules.
Fig. 6 is a schematic diagram of the inventive full sampling rule using the center of the vertex angle.
Detailed Description
The present invention will be further described with reference to the following examples and the accompanying drawings.
Example 1:
the indoor positioning method for constructing the fingerprint map based on the multi-chain interpolation is different from the traditional interpolation method, the traditional interpolation method is based on the signal propagation theory, namely the closer the physical positions of two sampling points are, the more similar the intensity values of the two sampling points are, therefore, the traditional interpolation method utilizes the principle to estimate the intensity of the substitute insertion point, correspondingly, the designed interpolation mode takes the neighbors as reference, obviously, the interpolation mode only carries out interpolation calculation from the angle of numerical similarity, and the signal propagation condition in the actual environment is not considered.
According to actual sampling, the distance from a propagation source to an insertion point to be inserted on a path to be propagated is close to the distance from the propagation source to a neighbor point, and therefore, the relevance between the insertion point to be inserted and the neighbor point is based on the propagation direction of a signal rather than the neighbor surrounding the insertion point to be inserted.
The steps of the method of the embodiment are shown in fig. 1, and the method specifically comprises the following steps:
firstly, a system for constructing a fingerprint map based on multi-chain interpolation is provided, the system is used for constructing a complete fingerprint map, and the multi-chain is based on multi-propagation-oriented links during signal interpolation.
And then sampling the sample target according to a sampling rule to obtain a sampling point data set, wherein the sampling point data set comprises position coordinate information of the sampling point and an RSS value of the sampling point, the sampling point in the sample target is removed to leave a to-be-inserted point, and the obtained sampling point data set is not only used as a reference data set for solving the data set of the to-be-inserted point, but also forms a complete fingerprint database, namely a fingerprint map, together with the data set of the to-be-inserted point.
Then inputting the sampling point data set into the system for constructing the fingerprint map based on the multi-chain interpolation;
and then, carrying out internal processing by a system for constructing a fingerprint map based on multi-chain interpolation to obtain a to-be-inserted point data set so as to obtain a complete fingerprint map, wherein the complete fingerprint map comprises a sampling point data set and a to-be-inserted point data set, the to-be-inserted point data set comprises position coordinate information of a to-be-inserted point and an RSS value of the to-be-inserted point, and the sampling point data set comprises position coordinate information of a sampling point and the RSS value of the sampling point.
And finally, obtaining the position coordinate information of all points to be detected in the testing stage through the complete fingerprint map obtained in the steps and the RSS value of the point to be detected which needs to be tested in the testing stage. Specifically, the complete fingerprint map obtained through the above steps includes position coordinate information of all points and RSS values of the points, and the position coordinate information of the points to be detected can be obtained through comparison according to the RSS values of the points to be detected.
Example 2:
in this embodiment, based on embodiment 1 and referring to fig. 2, the indoor positioning method for constructing a fingerprint map based on multi-chain interpolation includes:
an input module: inputting a sampling point data set into the system for constructing the fingerprint map based on the multi-chain interpolation;
an RSS value estimation module of a point to be inserted: the RSS value is used for calculating the RSS value of the point to be inserted;
a position coordinate calculation module of the point to be inserted: the device is used for calculating the position coordinates of a point to be inserted;
data set union module: taking the RSS value of the point to be inserted and the position coordinate information of the point to be inserted as a point data set to be inserted, and combining the point data set to be inserted and the sampling point data set to be used as a complete fingerprint map;
an output module: the fingerprint map is used for outputting a complete fingerprint map;
the RSS value estimation module to be inserted is configured to calculate an RSS value of the to-be-inserted point, and with reference to fig. 3, the RSS value estimation module to be inserted specifically includes:
a direction valuation module: determining an AP signal propagation path corresponding to an to-be-inserted point, where generally, no less than 2 APs corresponding to the to-be-inserted point are determined, that is, no less than 2 AP signal propagation paths corresponding to the to-be-inserted point are determined, and performing multi-link interpolation calculation on the to-be-inserted point in each AP signal propagation path direction, where the multi-link interpolation calculation means performing interpolation calculation based on the AP signal propagation paths corresponding to the to-be-inserted points to obtain RSS values of the to-be-inserted point in each directional chain, where one directional chain refers to one AP signal propagation path;
the strength estimation module: obtaining weights of the directional chains through the RSS values of the to-be-inserted point on the directional chains by using an inverse distance weighting method, and calculating an estimated RSS value of the to-be-inserted point through the weights of the directional chains;
a weight optimization module: fitting a signal attenuation model on each directional chain by using a sampling point data set adjacent to the point to be inserted on each directional chain, calculating a fitting RSS value of the adjacent sampling point of the point to be inserted on each directional chain through the signal attenuation model, calculating an error of the signal attenuation model on each directional chain through the fitting RSS value of the adjacent sampling point, and recalculating the weight on each directional chain through the error of the signal attenuation model on each directional chain;
an adjustment valuation module: recalculating the RSS value of the point to be inserted according to the RSS value of the point to be inserted on each directional chain and the weight of the point to be inserted on each directional chain; in the method of this embodiment, it is considered that the to-be-inserted point receives wireless signals in different directions, that is, the AP signal propagation path corresponding to the to-be-inserted point includes multiple paths, and the signal strength of the to-be-inserted point is affected by the superposition of multiple signals.
An iteration expansion module: and repeating the module steps to obtain the RSS values of all points to be inserted.
And calculating the RSS value of the point to be inserted through the RSS value estimation module of the point to be inserted, and further obtaining the data set of the point to be inserted.
Example 3:
the indoor positioning method for constructing the fingerprint map based on the multi-chain interpolation is based on the embodiment 2, and the RSS value estimation module of the point to be inserted is used for calculating the RSS value of the point to be inserted and comprises a direction estimation module, a strength estimation module, a weight optimization module, an adjustment estimation module and an iteration expansion module;
the method for solving the RSS value of the point to be inserted on each direction chain through the direction estimation module specifically comprises the following steps:
if the point to be inserted is P, the position coordinate of the point to be inserted P is P (x, y), and the directional link of the point to be inserted P (x, y), that is, the directional link is determined, where the point to be inserted P (x, y) has n directional links, and the n directional links are marked as (r)1,r2,…,rn) Setting the insertion point P (x, y) at rkThe number of the neighbor link layers on the directional link is l (k) (k is more than or equal to 1 and less than or equal to n), and the point P (x, y) to be inserted is at rkNeighbor set marking on directional links
Figure BDA0003523569620000081
Figure BDA0003523569620000082
Is the point P (x, y) to be inserted at rkThe nearest neighbor point on the directional link,
Figure BDA0003523569620000083
is the point of insertion
Figure BDA0003523569620000084
At rkThe nearest neighbor point on the directional link,
Figure BDA0003523569620000085
is that
Figure BDA0003523569620000086
At rkNearest neighbor points on the directional link, and so on;
point P (x, y) to be inserted is at rkRSS values on directional links
Figure BDA0003523569620000087
Calculated by equation (1):
Figure BDA0003523569620000088
in the above-mentioned formula (1),
Figure BDA0003523569620000089
is a point
Figure BDA00035235696200000810
The RSS value of (a) of (b),
Figure BDA00035235696200000811
is a point
Figure BDA00035235696200000812
The RSS value of (a) of (b),
Figure BDA00035235696200000813
is the point P (x, y) to be inserted and its value at rkNearest neighbor point on directional link
Figure BDA00035235696200000814
The distance between the two or more of the two or more,
Figure BDA00035235696200000815
is a point
Figure BDA00035235696200000816
And point
Figure BDA00035235696200000817
Distance between, point
Figure BDA00035235696200000818
RSS value of
Figure BDA00035235696200000819
Dot
Figure BDA00035235696200000820
RSS value of
Figure BDA00035235696200000821
Can be obtained by the step of sampling,
Figure BDA00035235696200000822
Figure BDA00035235696200000823
and
Figure BDA00035235696200000824
all can be obtained by calculating the position coordinates of the points;
and calculating the RSS value of the point to be inserted in each direction chain through the steps.
Through the estimated RSS value of the point to be inserted into the strength estimation module, firstly, the weight of each directional chain is calculated through the steps, and specifically, the method comprises the following steps:
setting the insertion point P (x, y) at rkThe weight on the directional link is w (r)k) Then w (r)k) Calculated according to equation (2):
Figure BDA00035235696200000825
in the above-mentioned formula (2),
Figure BDA00035235696200000826
is a point
Figure BDA00035235696200000827
The RSS value of (a) of (b),
Figure BDA00035235696200000828
is a point
Figure BDA00035235696200000829
The RSS value of (a) of (b),
Figure BDA00035235696200000830
is a point
Figure BDA00035235696200000831
And point
Figure BDA00035235696200000832
Distance between, point
Figure BDA00035235696200000833
RSS value of
Figure BDA00035235696200000834
Dot
Figure BDA00035235696200000835
RSS value of
Figure BDA00035235696200000836
Can be obtained by the sampling step, and can be obtained by the sampling step,
Figure BDA00035235696200000837
all can be obtained by calculating position coordinates;
obtaining the point P (x, y) to be inserted at r through the stepskThe weight on the directional link is w (r)k)。
Calculating RSS predicted value RSS of point P (x, y) to be inserted according to formula (3)P；
Figure BDA0003523569620000091
In the above-mentioned formula (3),
Figure BDA0003523569620000092
calculated by the formula (1), w (r)k) The expression is calculated by formula (2), n is a natural number, and k is more than or equal to 1 and less than or equal to n;
the RSS estimated value RSS of the point P (x, y) to be inserted is obtained through the stepsP。
The weight on each directional chain is recalculated through the weight optimization module, which specifically comprises the following steps:
firstly, a signal attenuation model on each direction chain is fitted by using a data set of adjacent sampling points of the point to be inserted on each direction chain, and the point to be inserted P (x, y) is set at rkThe adjacent sampling points on the direction chain are marked as
Figure BDA0003523569620000093
RSS value tagging of neighboring sample points
Figure BDA0003523569620000094
m is the number of adjacent sampling points, the position coordinate information of the adjacent sampling points is obtained through the sampling step, and the parameters of the signal attenuation model formula (4) are obtained according to the signal attenuation model formula (4) and the adjacent sampling point data set in a fitting mode
Figure BDA0003523569620000095
And the value of the parameter σ:
Figure BDA0003523569620000096
the point P (x, y) to be inserted is at rkAnd (3) substituting the data set of the adjacent sampling points on the direction chain into the formula (4), wherein the data set of the adjacent sampling points comprises the RSS values and the position coordinate information of the adjacent sampling points to obtain
Figure BDA0003523569620000097
In the above-mentioned formula,
Figure BDA0003523569620000098
is a point
Figure BDA0003523569620000099
The distance to the point P (x, y) to be inserted,
Figure BDA00035235696200000910
is a point
Figure BDA00035235696200000911
The distance to the point P (x, y) to be inserted,
Figure BDA00035235696200000912
Figure BDA00035235696200000913
is a point
Figure BDA00035235696200000914
The distance to the point P (x, y) to be inserted, and so on. Calculating the parameters by the above formula
Figure BDA00035235696200000915
And the value of the parameter σ, to obtain rkFormula of signal attenuation model on direction chain
Figure BDA00035235696200000916
Figure BDA00035235696200000917
Then obtaining a signal attenuation model formula through fitting and RSS values of adjacent sampling points of points to be inserted on each directional chain
Figure BDA00035235696200000918
Calculating fitted RSS values of sampling points adjacent to the point to be inserted on each directional chain, wherein the fitted RSS values of the sampling points adjacent to the point to be inserted are marked as
Figure BDA0003523569620000101
The fitted RSS value of the adjacent sampling point of the point to be inserted on each directional chain is calculated by the following formula:
Figure BDA0003523569620000102
in the above formula, the RSS values of adjacent sampling points
Figure BDA0003523569620000103
Obtained by a sampling step
Figure BDA0003523569620000104
And the value of the parameter sigma is calculated by formula (4),
Figure BDA0003523569620000105
is a point
Figure BDA0003523569620000106
The distance to the point P (x, y) to be inserted,
Figure BDA0003523569620000107
is a point
Figure BDA0003523569620000108
The distance to the point P (x, y) to be inserted,
Figure BDA0003523569620000109
is a point
Figure BDA00035235696200001010
The distance to the point P (x, y) to be inserted, and so on.
Then, r is calculated according to the formula (5)kError of the signal attenuation model on the directional link:
Figure BDA00035235696200001011
in the above formula (5), α is based on
Figure BDA00035235696200001012
Definition of the spacing to the point to be inserted, beta being based on
Figure BDA00035235696200001013
Definition of the distance to the point to be inserted, γ being based on the point
Figure BDA00035235696200001014
Definition of the distance to the point to be inserted, m is the point to be inserted P (x, y) at rkThe number of adjacent sampling points on the direction chain is not less than 2;
and finally, recalculating the point P (x, y) to be inserted at r according to the formula (6)kOptimized weights w' (r) on the directional linkk)：
Figure BDA00035235696200001015
In the above formula (6), D (r)k) The method is obtained by calculation according to a formula (5), wherein n is a natural number, and k is more than or equal to 1 and less than or equal to n;
recalculating the optimized RSS value RSS' P of the point to be inserted P (x, y) according to formula (7):
Figure BDA00035235696200001016
in the above-mentioned formula (7),
Figure BDA00035235696200001017
calculated by the formula (1), w' (r)k) The method is obtained by calculation according to a formula (6), wherein n is a natural number, and k is more than or equal to 1 and less than or equal to n;
the obtained optimized RSS value RSS 'of the point P (x, y) to be inserted'PAnd as the final RSS value of the point to be inserted P (x, y), obtaining the final RSS values of all the points to be inserted in the sample target through all the steps.
Example 4:
in this embodiment, based on embodiment 3, in the direction estimation module, referring to fig. 4, it is shown in fig. 4 that, in the direction estimation module, AP signal propagation paths corresponding to an access point to be inserted are 4, that is, the access point to be inserted receives 4 AP direction signals, and the access point to be inserted receives an AP signal to be insertedThe number of the signals is superposition of 4 AP direction signals, and the 4 APs are respectively APs1、AP2、AP3And AP4The signal to be inserted is AP1、AP2、AP3And AP4And (3) superposing the direction signals, wherein the point to be inserted P (x, y) comprises RSS values on 4 direction links, the point to be inserted P (x, y) comprises weights on the 4 direction links, and an RSS estimated value RSS of the point to be inserted P (x, y) is obtained by calculating the RSS values on the 4 direction links and the weights on the 4 direction linksPFurther, an RSS value RSS 'after optimization of the insertion point P (x, y) is obtained'P。
Example 5:
in this embodiment, based on embodiment 3, in the weight optimization module, the sampling point data sets adjacent to the to-be-inserted point on each directional chain are used to fit the signal attenuation model on each directional chain, the fitted RSS values of the sampling points adjacent to the to-be-inserted point on each directional chain are calculated through the signal attenuation model, the error of the signal attenuation model on each directional chain is calculated through the fitted RSS values of the adjacent sampling points, and the weights on each directional chain are recalculated through the error of the signal attenuation model on each directional chainkThe adjacent sampling points on the direction chain are marked as
Figure BDA0003523569620000111
Fitting to obtain rkSignal attenuation model formula on direction chain
Figure BDA0003523569620000112
Then calculating fitted RSS values of sampling points adjacent to the point to be inserted on each directional chain, wherein the fitted RSS values of the sampling points adjacent to the point to be inserted are marked as
Figure BDA0003523569620000113
Then, r is calculated according to the formula (5)kError of the signal attenuation model on the directional link:
Figure BDA0003523569620000114
in the above formula (5), α is based on
Figure BDA0003523569620000115
Definition of the distance to the point to be inserted, β being based on the point
Figure BDA0003523569620000116
Definition of the distance to the point to be inserted, γ being based on the point
Figure BDA0003523569620000117
Definition of the distance to the point to be inserted, m is the point to be inserted P (x, y) at rkThe number of its neighboring samples on the directional chain, m equals 2, then equation (5) transforms to:
Figure BDA0003523569620000118
in the above formula, α is based on
Figure BDA0003523569620000119
Definition of the distance to the point to be inserted, β being based on the point
Figure BDA00035235696200001110
The spacing to the point to be inserted is defined.
Obtaining D (r) by the above transformation formulak) And finally, recalculating the point to be inserted P (x, y) at r according to formula (6)kOptimized weights w' (r) on the directional linkk)：
Figure BDA00035235696200001111
And finally, recalculating the optimized RSS value RSS 'of the point to be inserted P (x, y) according to a formula (7)'P：
Figure BDA0003523569620000121
The obtained optimized RSS value RSS 'of the point P (x, y) to be inserted'PAnd as the final RSS value of the point to be inserted P (x, y), obtaining the final RSS values of all the points to be inserted in the sample target through all the steps.
Example 6
In the indoor positioning method for building the fingerprint map based on the multi-chain interpolation of the embodiment, based on the embodiment 3, the position coordinate calculation module of the point to be inserted calculates the position coordinate of the point to be inserted through the sensing environment predefining, that is, when a sample target is sampled according to the sampling rule, sampling is performed to obtain a sampling point data set, and the position coordinate of the point to be inserted is calculated through the sensing environment predefining, so that the point to be inserted data set is obtained.
Example 7
In the method for indoor positioning based on a multi-chain interpolation to construct a fingerprint map according to the embodiment 6, in Sampling a sample target according to a Sampling rule, the Sampling rule employs a vertex angle Sampling rule, and when the vertex angle Sampling rule is employed, a corresponding Sampling map is divided into a plurality of grids, fingerprints are uniformly acquired at intervals on the grids at a certain distance, referring to fig. 5, the specific rule is that the map is divided into a plurality of "3 × 3" squares, four corners (four vertices) of each square are sampled, and finally, the whole Sampling space is filled up, and the Sampling mode is referred to as a "vertex angle Sampling mode (Corner Sampling)" according to the specific rule, as shown in the drawing, 4 vertices in each "3 × 3" square are real Sampling points, and the remaining 5 points in the "3 × 3" square are to be inserted. To ensure the interval sampling, each time the "3 × 3" grid slides 2 sampling points, the 10m × 10m map in fig. 5 is sampled by using this pattern, and the real sampling points only account for 25/100 of all points, which is 25%.
Example 8
In the indoor positioning method for constructing a fingerprint map based on multi-chain interpolation according to the embodiment 6, in Sampling a sample object according to a Sampling rule, the Sampling rule employs a vertex angle center full Sampling rule, it is known that the more Sampling points are, the better the constructed fingerprint map effect is, and the more similar the fingerprints of the more adjacent points are, based on the above practical situation, on the basis of a "vertex angle Sampling mode (Corner Sampling) mode", to further improve the positioning accuracy and stability, the Sampling points increase the grid center points, specifically, when the vertex angle center full Sampling rule is employed, in addition to acquiring four vertex angles of a "3 × 3" grid, the center points are also included, the Sampling form is shown in fig. 6, the Sampling rule is called "vertex angle + center Sampling mode (Corner & center Sampling)", as shown in the figure, wherein 4 vertex points and 1 center point in each "3 × 3" grid are real Sampling points, the remaining 4 points in the "3 x 3" square are the points to be inserted. Similar to the Corner Sampling mode, the grid slides 2 samples at a time, and in this Sampling mode, for the 10m × 10m map of fig. 6, the real sample points account for 50% of all points ((25+ 25))/100%.
Example 9
In the indoor positioning method for constructing the fingerprint map based on the multi-chain interpolation, based on embodiment 7, the performance of the MCI method using the Corner Sampling mode (Corner Sampling) is verified through experiments in the embodiment, and experiments are performed in a large-scale scene to verify the performance of the MCI method. The experimental scene is an indoor environment with 3000 square meters, and 50 equal-height iBeacons are placed at a height of 2 meters from the ground to serve as APs. In order to verify the stability of the performance of MCI in different devices, this embodiment uses 6 smartphones for sampling, which are HUAWEI mate7, HUAWEI mate20, HUAWEI horor, VIVO x20, VIVO x23 and YAAO, respectively. The sampling points are spaced by 1 meter, each sampling point collects signals for 1 minute, the whole sampling process takes about 250 hours, and the following table 1 is obtained through experimental comparison:
TABLE 1
Figure BDA0003523569620000131
Table 1 shows the results of comparing the MCI method with other methods when Sampling in the Corner Sampling mode (Corner Sampling), and it can be seen from table 1 that the MCI method has an average positioning accuracy 4.74% higher than that of the full Sampling mode (i.e. Sampling all points in the sample object).
Example 10
In the indoor positioning method for constructing the fingerprint map based on the multi-chain interpolation, based on embodiment 8, the performance of the MCI method using the Corner + center Sampling mode (Corner & center Sampling) is verified through experiments in the embodiment, and experiments are performed in a large-scale scene to verify the performance of the MCI method. The experimental scenario is 3000 square meters's indoor environment, has placed 50 equal height iBeacon as the AP in 2 meters high from the ground, for the stability of verifying MCI performance in different equipment, this embodiment has used 6 kinds of smart mobile phones to sample, is HUAWEImate 7, HUAWEImate 20, HUAWEI honor, VIVO x20, VIVO x23 and YAAO respectively. The sampling points are spaced at 1 meter intervals, and each sampling point collects a signal for 1 minute, which takes about 250 hours for the entire sampling process. Table 2 was obtained by experimental comparison:
TABLE 2
Figure BDA0003523569620000141
Table 2 shows the results of comparing the MCI method with other methods when Sampling in "Corner + center Sampling mode" (inner & center Sampling), the MCI method has an average positioning accuracy 13.58% higher than that of the full Sampling method (i.e. Sampling all points in the sample object).
According to experimental verification, under the two sampling modes, when the sampling points are reduced by 50% and 75% compared with the full sampling mode, the positioning accuracy of the MCI is improved by 13.58% and 4.74% respectively. Compared with a classical interpolation mode, the MCI method has better stability, and when the sampling quantity is less, the advantage of positioning precision is more obvious. When the sampling amount is only 25% of the full sampling amount, the average accuracy rate is improved by 18.50% compared with the classical interpolation mode. Therefore, the MCI method can better ensure the positioning accuracy and greatly reduce the workload of the map construction process.
The above description is only of the preferred embodiments of the present invention, and it should be noted that: it will be apparent to those skilled in the art that various modifications and adaptations can be made without departing from the principles of the invention and these are intended to be within the scope of the invention.

Claims (7)

1. An indoor positioning method for constructing a fingerprint map based on multi-chain interpolation is characterized by comprising the following steps: the method provides a system for constructing a fingerprint map based on multi-chain interpolation, and the method comprises the following steps:
s1, sampling the sample target according to the sampling rule, and sampling to obtain a sampling point data set, wherein the sampling point data set comprises position coordinate information of a sampling point and an RSS value of the sampling point;
s2, inputting the sample point data set into the system for constructing the fingerprint map based on the multi-chain interpolation;
s3, internally processing by a system for constructing a fingerprint map based on multi-chain interpolation to obtain a complete fingerprint map, wherein the complete fingerprint map comprises a sampling point data set and a to-be-inserted point data set, and the to-be-inserted point data set comprises position coordinate information of the to-be-inserted point and an RSS value of the to-be-inserted point;
and S4, obtaining the position coordinate information of the point to be detected through the complete fingerprint map and the RSS value of the point to be detected.
2. The indoor positioning method for constructing the fingerprint map based on the multi-chain interpolation as claimed in claim 1, wherein: the system for constructing the fingerprint map based on the multi-chain interpolation comprises the following steps:
an input module: inputting a sampling point data set into the system for constructing the fingerprint map based on the multi-chain interpolation;
an RSS value estimation module of a point to be inserted: the RSS value used for calculating the point to be inserted;
a position coordinate calculation module of the point to be inserted: the device is used for calculating the position coordinates of a point to be inserted;
data set union module: taking the RSS value of the point to be inserted and the position coordinate information of the point to be inserted as a point data set to be inserted, and combining the point data set to be inserted and the sampling point data set to be used as a complete fingerprint map;
an output module: for outputting a complete fingerprint map.
3. The indoor positioning method for constructing the fingerprint map based on the multi-chain interpolation as claimed in claim 2, wherein: the RSS value estimation module for the point to be inserted is used for calculating the RSS value of the point to be inserted, and specifically comprises:
a direction valuation module: determining an AP signal propagation path corresponding to the point to be inserted, and performing interpolation calculation on the point to be inserted in the direction of each AP signal propagation path to obtain an RSS value of the point to be inserted in each direction chain;
the strength estimation module: obtaining weights of the directional chains through the RSS values of the to-be-inserted point on the directional chains by using an inverse distance weighting method, and calculating an estimated RSS value of the to-be-inserted point through the weights of the directional chains;
the weight optimization module: fitting the adjacent sampling point data set of the point to be inserted on each directional chain to a signal attenuation model on each directional chain, calculating the fitting RSS value of the adjacent sampling point of the point to be inserted on each directional chain through the signal attenuation model, calculating the error of the signal attenuation model on each directional chain through the fitting RSS value of the adjacent sampling point, and recalculating the weight on each directional chain through the error of the signal attenuation model on each directional chain;
and an adjustment valuation module: recalculating the RSS value of the point to be inserted according to the RSS value of the point to be inserted on each directional chain and the weight of the point to be inserted on each directional chain;
an iteration expansion module: and repeating the module steps to obtain the RSS values of all points to be inserted.
4. The indoor positioning method for constructing the fingerprint map based on the multi-chain interpolation as claimed in claim 3, wherein: and the position coordinate calculation module of the point to be inserted calculates the position coordinate of the point to be inserted through sensing environment pre-definition.
5. The indoor positioning method for constructing the fingerprint map based on the multi-chain interpolation as claimed in claim 4, wherein: the sampling rule adopts a vertex angle sampling rule or a vertex angle center full-sampling rule.
6. The indoor positioning method for constructing fingerprint map based on multi-chain interpolation as claimed in claim 5, wherein: and 4 AP signal propagation paths corresponding to the point to be inserted.
7. The indoor positioning method for constructing fingerprint map based on multi-chain interpolation as claimed in claim 6, wherein: the adjacent sampling points of the point to be inserted on each directional chain at least comprise 2.
CN202210188290.1A 2022-02-28 2022-02-28 Indoor positioning method for constructing fingerprint map based on multi-chain interpolation Pending CN114710742A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210188290.1A CN114710742A (en) 2022-02-28 2022-02-28 Indoor positioning method for constructing fingerprint map based on multi-chain interpolation

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210188290.1A CN114710742A (en) 2022-02-28 2022-02-28 Indoor positioning method for constructing fingerprint map based on multi-chain interpolation

Publications (1)

Publication Number Publication Date
CN114710742A true CN114710742A (en) 2022-07-05

Family

ID=82166784

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210188290.1A Pending CN114710742A (en) 2022-02-28 2022-02-28 Indoor positioning method for constructing fingerprint map based on multi-chain interpolation

Country Status (1)

Country Link
CN (1) CN114710742A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115905787A (en) * 2022-10-21 2023-04-04 盐城师范学院 High-precision indoor positioning method based on fuzzy migration learning model

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115905787A (en) * 2022-10-21 2023-04-04 盐城师范学院 High-precision indoor positioning method based on fuzzy migration learning model
CN115905787B (en) * 2022-10-21 2023-09-29 盐城师范学院 High-precision indoor positioning method based on fuzzy migration learning model

Similar Documents

Publication Publication Date Title
CN109743683B (en) Method for determining position of mobile phone user by adopting deep learning fusion network model
CN104038901B (en) Indoor positioning method for reducing fingerprint data acquisition workload
CN103984981B (en) Building environmental sensor measuring point optimization method based on Gaussian process model
Du et al. CRCLoc: A crowdsourcing-based radio map construction method for WiFi fingerprinting localization
CN103220777A (en) Mobile device positioning system
CN107607122B (en) The building of location fingerprint library and dynamic updating method towards indoor positioning
CN104540221A (en) WLAN indoor positioning method based on semi-supervised SDE algorithm
CN112218330A (en) Positioning method and communication device
CN106961659A (en) The bluetooth localization method of one species fingerprint
CN108834041B (en) Indoor position fingerprint positioning Radio Map establishing method based on tensor reconstruction
CN105652235A (en) Linear regression algorithm-based WLAN indoor positioning multi-user RSS (Received Signal Strength) fusion method
CN107850656A (en) The determination of model parameter for positioning purposes
CN115100376A (en) Electromagnetic spectrum map implementation method based on improved inverse distance interpolation method
CN106658538B (en) Mobile phone base station signal coverage area simulation method based on Thiessen polygon
CN109033181B (en) Wind field geographic numerical simulation method for complex terrain area
CN114710742A (en) Indoor positioning method for constructing fingerprint map based on multi-chain interpolation
Dong et al. A wifi fingerprint augmentation method for 3-d crowdsourced indoor positioning systems
CN105187139B (en) A kind of outdoor radio signal reception strength map constructing method based on intelligent perception
Arai et al. Color radiomap interpolation for efficient fingerprint wifi-based indoor location estimation
CN108632752B (en) Indoor position fingerprint positioning Radio Map establishing method based on tensor filling
Hosseini et al. NSGA-II based optimal Wi-Fi access point placement for indoor positioning: A BIM-based RSS prediction
CN114187524B (en) Household air conditioner identification method, device, equipment, storage medium and product
Assayag et al. Indoor positioning system using synthetic training and data fusion
CN113645565B (en) Indoor positioning method based on hexagonal closest packing structure
CN115942231A (en) RSS-based 5G outdoor positioning method

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication