CN114895111B - Method and system for constructing electromagnetic map based on weight distribution - Google Patents

Method and system for constructing electromagnetic map based on weight distribution Download PDF

Info

Publication number
CN114895111B
CN114895111B CN202210785057.1A CN202210785057A CN114895111B CN 114895111 B CN114895111 B CN 114895111B CN 202210785057 A CN202210785057 A CN 202210785057A CN 114895111 B CN114895111 B CN 114895111B
Authority
CN
China
Prior art keywords
electromagnetic
data
representing
geographic
matrix
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.)
Active
Application number
CN202210785057.1A
Other languages
Chinese (zh)
Other versions
CN114895111A (en
Inventor
王红军
林栋明
刘金帆
沈哲贤
李歆昊
王军
安永旺
孟祥豪
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.)
National University of Defense Technology
Original Assignee
National University of Defense Technology
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 National University of Defense Technology filed Critical National University of Defense Technology
Priority to CN202210785057.1A priority Critical patent/CN114895111B/en
Publication of CN114895111A publication Critical patent/CN114895111A/en
Application granted granted Critical
Publication of CN114895111B publication Critical patent/CN114895111B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R29/00—Arrangements for measuring or indicating electric quantities not covered by groups G01R19/00 - G01R27/00
    • G01R29/08—Measuring electromagnetic field characteristics
    • G01R29/0864—Measuring electromagnetic field characteristics characterised by constructional or functional features
    • G01R29/0892—Details related to signal analysis or treatment; presenting results, e.g. displays; measuring specific signal features other than field strength, e.g. polarisation, field modes, phase, envelope, maximum value
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G—PHYSICS
    • G09—EDUCATION; CRYPTOGRAPHY; DISPLAY; ADVERTISING; SEALS
    • G09B—EDUCATIONAL OR DEMONSTRATION APPLIANCES; APPLIANCES FOR TEACHING, OR COMMUNICATING WITH, THE BLIND, DEAF OR MUTE; MODELS; PLANETARIA; GLOBES; MAPS; DIAGRAMS
    • G09B29/00—Maps; Plans; Charts; Diagrams, e.g. route diagram
    • G09B29/003—Maps
    • G09B29/005—Map projections or methods associated specifically therewith

Abstract

The invention discloses a method and a system for constructing an electromagnetic map based on weight distribution, and belongs to the technical field of image data processing. The method comprises the following steps: deploying a plurality of sensing nodes in a target area to obtain a sensing matrix, acquiring electromagnetic observation data of the target area through sampling, and constructing electromagnetic related data of an electromagnetic map based on the sensing matrix and the electromagnetic observation data; acquiring geographic sampling data of a target area through sampling based on a plurality of deployed sensing nodes, and predicting geographic related data of an un-sampled position according to geographic correlation among the geographic sampling data so as to obtain the geographic related data of the electromagnetic map; weights are assigned to the electromagnetic-related data of the electromagnetic map and the geographic-related data of the electromagnetic map, and a complete electromagnetic map of the target area is constructed by fusing the weighted electromagnetic-related data and the weighted geographic-related data. The electromagnetic data distribution obtained by the scheme disclosed by the invention is closer to real electromagnetic data.

Description

Method and system for constructing electromagnetic map based on weight distribution
Technical Field
The invention belongs to the technical field of image data processing, and particularly relates to a method and a system for constructing an electromagnetic map based on weight distribution.
Background
In order to solve the contradiction between the limited spectrum resources and the increasing spectrum resource requirements, the dynamic allocation and use of the spectrum resources are imperative; to realize dynamic allocation of spectrum resources, the use condition of the current spectrum resources, i.e. electromagnetic data, is obtained first. The electromagnetic map covers various electromagnetic signal radiation sources, can reflect images of electromagnetic signal related information in a complex geographic environment, and can provide accurate electromagnetic data for use evaluation of spectrum resources.
When the electromagnetic map is drawn, the traditional data acquisition method mainly adopts manual drive test, and a large amount of manpower and material resources are consumed in the mode, particularly in a large-range area, and meanwhile, the result is greatly influenced by the subjectivity of people; the existing method mostly takes a sensing node as a main point, namely, the automatic acquisition of data is realized by utilizing an intelligent and miniaturized technology. However, the cost and effectiveness of this approach is proportional to the number of nodes used. Although complete and accurate results can be obtained by excessive deployment, the number of consumed nodes is large, and the cost is high; a small number of deployments may only get a portion of the data, but the cost is lower than an excessive number of deployments. In addition, in a strict sense, only the data of the position of the node in the data sensed by the node is accurate, and other data are all estimated according to the radio wave propagation equation, and the data has a large error when the environment condition of the target area is unknown.
At present, an inverse distance weighting interpolation method and a kriging interpolation method are mostly adopted when the electromagnetic map is constructed; the latter also applies the nearest neighbor method to construct electromagnetic maps. The method comprises the following steps that a kriging interpolation method calculates corresponding lag distance and variation values according to sampling data; using the lag distance and the corresponding variation value to fit the selected kriging variation function; determining a correlation coefficient of the variation function; and predicting data at the non-sampling point by using the variation value between the non-sampling point and the sampling point position and the sampling data. The method comprises the following steps of calculating the distances between an unsampled point and all sampling points by an inverse distance weighted interpolation method; calculating a corresponding weight value by using the calculated distance; the data at the non-sampled points can be predicted by summing the products of the corresponding weights and the sample data values. In addition, other interpolation methods such as nearest neighbor, modified schilder interpolation, local polynomial, etc. can be used to predictively reconstruct the null data, and these methods are applicable to less than kriging interpolation and inverse distance weighted interpolation.
The kriging interpolation method is sensitive to abnormal values, the precision is greatly influenced by the abnormal values, and the abnormal values can inevitably exist in practical application; the computation complexity of the inverse distance weighting interpolation method is low, but a severe bullseye phenomenon exists, and the accuracy of the numerical value edge is difficult to ensure; the improved Sheberd interpolation method is an improvement of an inverse distance weighting interpolation method, and can effectively relieve the bulls-eye phenomenon, but the setting of related parameters is difficult, so that the optimal solution is difficult to obtain; the nearest neighbor method is generally suitable for predicting uniformly spaced data, but the result has the problem of discontinuous change, so that the precision is influenced; the local polynomial method is suitable for the data prediction problem of short-range change, but the action effect is easily influenced by the neighborhood distance. However, in the method based on the propagation model, once the prior information including the signal propagation environment characteristics in the target region cannot be accurately grasped, the result precision is very low.
Disclosure of Invention
In view of the above technical problems, the present invention provides a scheme for constructing an electromagnetic map based on weight distribution, where the scheme includes a method for constructing an electromagnetic map based on weight distribution, a corresponding electronic device, and a computer-readable storage medium.
The invention discloses a method for constructing an electromagnetic map based on weight distribution in a first aspect. The electromagnetic map comprises electromagnetic related data and geographic related data; the method comprises the following steps: s1, deploying a plurality of sensing nodes in a target area to obtain a sensing matrix, obtaining electromagnetic observation data of the target area through sampling, and constructing the electromagnetic related data of the electromagnetic map based on the sensing matrix and the electromagnetic observation data; s2, acquiring geographic sampling data of the target area through sampling based on the deployed sensing nodes, and predicting geographic related data of an unsampled position according to geographic correlation among the geographic sampling data to obtain the geographic related data of the electromagnetic map; s3, distributing weights for the electromagnetic relevant data of the electromagnetic map and the geographic relevant data of the electromagnetic map, and constructing the complete electromagnetic map of the target area by fusing the electromagnetic relevant data and the geographic relevant data distributed with the weights.
According to the method of the first aspect of the present invention, the step S1 specifically includes the following procedures.
Deploying a plurality of sensing nodes in the target area to acquire a sensing matrix, specifically comprising: equally dividing the target area into a plurality of sub-areas and numbering, wherein the number is
Figure 660697DEST_PATH_IMAGE001
The same number of sensing nodes are deployed in each sub-area, and the deployment mode is random deployment.
The electromagnetic correlation data is characterized by equation (1):
Figure 558859DEST_PATH_IMAGE003
（1）
wherein the content of the first and second substances,
Figure 294734DEST_PATH_IMAGE004
representing said electromagnetic correlation data to be constructed,
Figure 900158DEST_PATH_IMAGE005
，
Figure 37879DEST_PATH_IMAGE006
representing the total amount of electromagnetic data within the target region,
Figure 437767DEST_PATH_IMAGE007
representing the electromagnetic observation data as a function of time,
Figure 660938DEST_PATH_IMAGE008
，
Figure 804475DEST_PATH_IMAGE007
in which comprises
Figure 528192DEST_PATH_IMAGE009
The data of the bar is transmitted to the mobile terminal,
Figure 161299DEST_PATH_IMAGE010
representing an observation matrix.
Based on the electromagnetic correlation data
Figure 747132DEST_PATH_IMAGE011
In a sparse matrix
Figure 818994DEST_PATH_IMAGE012
Projection onto
Figure 541093DEST_PATH_IMAGE013
Will solve for the electromagnetic correlation data
Figure 79522DEST_PATH_IMAGE004
Conversion to solve formula (2):
Figure 339602DEST_PATH_IMAGE014
（2）
wherein the content of the first and second substances,
Figure 759694DEST_PATH_IMAGE015
representing calculation matrices
Figure 195355DEST_PATH_IMAGE016
The norm of the number of the first-order-of-arrival,
Figure 170264DEST_PATH_IMAGE013
representing said electromagnetic related data
Figure 324165DEST_PATH_IMAGE017
In a sparse matrix
Figure 144353DEST_PATH_IMAGE018
The projection of the image onto the image plane is performed,
Figure 434520DEST_PATH_IMAGE019
-representing said perceptual matrix by means of a perceptual matrix,
Figure 314752DEST_PATH_IMAGE020
，
Figure 424790DEST_PATH_IMAGE018
representing the sparse matrix, characterized by equation (3):
Figure 272440DEST_PATH_IMAGE021
（3）
wherein the content of the first and second substances,
Figure 948272DEST_PATH_IMAGE022
。
solving equation (2) specifically includes performing the following process for each sub-region.
Initializing each parameter: residual error
Figure 264984DEST_PATH_IMAGE023
Index matrix
Figure 862318DEST_PATH_IMAGE024
Number of elements
Figure 555468DEST_PATH_IMAGE025
，
Figure 554648DEST_PATH_IMAGE026
Representing step size, the size of which is the number of the sub-regions and the iteration number
Figure 917627DEST_PATH_IMAGE027
，
Figure 530486DEST_PATH_IMAGE028
Representing the perception matrix
Figure 89644DEST_PATH_IMAGE019
To (1) a
Figure 349855DEST_PATH_IMAGE029
Column, phase index
Figure 8369DEST_PATH_IMAGE030
Said projection
Figure 111455DEST_PATH_IMAGE013
Is estimated value of
Figure 756194DEST_PATH_IMAGE031
。
Computing
Figure 261124DEST_PATH_IMAGE032
Before selecting the element values according to the calculated values
Figure 822031DEST_PATH_IMAGE033
A value of an element, and comparing the value
Figure 412412DEST_PATH_IMAGE033
Value of each element in the perception matrix
Figure 860842DEST_PATH_IMAGE034
The set of corresponding column sequence numbers in
Figure 954700DEST_PATH_IMAGE035
。
Calculating out
Figure 220597DEST_PATH_IMAGE036
，
Figure 908061DEST_PATH_IMAGE037
(ii) a Computing a least squares solution
Figure 612712DEST_PATH_IMAGE038
(ii) a According to the least square solution
Figure 823726DEST_PATH_IMAGE039
The absolute value of each element item is selected from the values
Figure 870310DEST_PATH_IMAGE033
The element items are extracted
Figure 763180DEST_PATH_IMAGE033
Element item in the sensing matrix
Figure 553413DEST_PATH_IMAGE034
Composition of corresponding column in (1)
Figure 621863DEST_PATH_IMAGE040
Corresponding column number constitution
Figure 698403DEST_PATH_IMAGE041
。
A new residual value is calculated and,
Figure 485094DEST_PATH_IMAGE042
，
Figure 935142DEST_PATH_IMAGE043
. Judgment of
Figure 858099DEST_PATH_IMAGE044
Whether the result is true or not; if so, then
Figure 246486DEST_PATH_IMAGE045
Update
Figure 520472DEST_PATH_IMAGE046
Will be
Figure 42721DEST_PATH_IMAGE046
In the corresponding position of
Figure 961129DEST_PATH_IMAGE047
Is set to
Figure 642122DEST_PATH_IMAGE048
(ii) a If not, further judgment is made
Figure 137825DEST_PATH_IMAGE049
Whether the result is true or not; if so, then
Figure 339130DEST_PATH_IMAGE050
，
Figure 236679DEST_PATH_IMAGE051
，
Figure 91503DEST_PATH_IMAGE052
And recalculating said new residual valuesr tc (ii) a If not, then
Figure 340082DEST_PATH_IMAGE045
，
Figure 810990DEST_PATH_IMAGE053
，
Figure 563045DEST_PATH_IMAGE052
And recalculating said new residual valuesr tc . Thereby solving the electromagnetic correlation data in the sub-area
Figure 588770DEST_PATH_IMAGE054
Fusing electromagnetic correlation data of the respective sub-regions
Figure 527907DEST_PATH_IMAGE055
，
Figure 930069DEST_PATH_IMAGE056
The electromagnetic relevant data of each sub-area after fusion is passed through a median filter to obtain the electromagnetic relevant data in the electromagnetic map in the target area
Figure 536631DEST_PATH_IMAGE057
Wherein, in the step (A),
Figure 670941DEST_PATH_IMAGE058
the median filtering process is indicated.
According to the method of the first aspect of the present invention, said step S2 comprises the following steps.
The geographical sampling data of any sampling point position is expressed by the reference signal receiving power of the sampling point position:
Figure 32127DEST_PATH_IMAGE059
（4）
wherein the content of the first and second substances,
Figure 378926DEST_PATH_IMAGE060
is representative of the reference signal received power,
Figure 699049DEST_PATH_IMAGE061
represents the geographical location of any of the sample points,
Figure 4259DEST_PATH_IMAGE062
a function representing a prediction model of the model,
Figure 855672DEST_PATH_IMAGE063
，
Figure 495511DEST_PATH_IMAGE064
a parameter representing the prediction model is determined,
Figure 342244DEST_PATH_IMAGE065
the number of functions of the prediction model and the number of parameters of the prediction model are both N,
Figure 146252DEST_PATH_IMAGE066
representing a random process.
Then the equation (5) holds:
Figure 953802DEST_PATH_IMAGE067
（5）
wherein the content of the first and second substances,
Figure 829355DEST_PATH_IMAGE068
representing the random process
Figure 405961DEST_PATH_IMAGE066
In the expectation that the position of the target is not changed,
Figure 112361DEST_PATH_IMAGE069
indicating the geographical location of the two sample points,
Figure 531841DEST_PATH_IMAGE070
a model of the correlation is represented by,
Figure 86450DEST_PATH_IMAGE071
a parameter representative of the correlation model is determined,
Figure 48721DEST_PATH_IMAGE072
representing the random process
Figure 460111DEST_PATH_IMAGE066
The variance of (c).
Setting the sample point of the geographic sampling data as
Figure 177007DEST_PATH_IMAGE073
The corresponding sample value is
Figure 4148DEST_PATH_IMAGE074
To yield formula (6):
Figure 352084DEST_PATH_IMAGE075
（6）
wherein the content of the first and second substances,
Figure 403217DEST_PATH_IMAGE076
to represent
Figure 407076DEST_PATH_IMAGE064
The predicted value of (a) is obtained,
Figure 503820DEST_PATH_IMAGE077
to represent
Figure 565317DEST_PATH_IMAGE078
The predicted value of (a) is obtained,
Figure 725034DEST_PATH_IMAGE079
is composed of
Figure 747348DEST_PATH_IMAGE080
The matrix of the composition is composed of a plurality of matrixes,
Figure 851046DEST_PATH_IMAGE081
，
Figure 235891DEST_PATH_IMAGE082
to represent
Figure 566509DEST_PATH_IMAGE083
A correlation matrix of type sample points, the constituent elements of which are
Figure 341698DEST_PATH_IMAGE084
。
Defining a correlation function:
Figure 517596DEST_PATH_IMAGE085
（7）
wherein the content of the first and second substances,
Figure 426121DEST_PATH_IMAGE086
is shown and
Figure 927641DEST_PATH_IMAGE061
the geographical location of a different one of the other sampling points,
Figure 190126DEST_PATH_IMAGE087
is the dimension of the sample point; and has the following formula (8):
Figure 169715DEST_PATH_IMAGE088
（8）
wherein the content of the first and second substances,
Figure 57381DEST_PATH_IMAGE089
representation solving
Figure 870747DEST_PATH_IMAGE082
The determinant of (2) adopts a spherical model.
Then there are:
Figure 354949DEST_PATH_IMAGE091
（9）
wherein, the first and the second end of the pipe are connected with each other,
Figure 262862DEST_PATH_IMAGE092
then, there is formula (10):
Figure 552505DEST_PATH_IMAGE093
（10）
wherein the content of the first and second substances,
Figure 989302DEST_PATH_IMAGE094
to represent
Figure 960800DEST_PATH_IMAGE095
The predicted value of (2).
Substituting the non-sampled position into formula (10) to obtain the geographic related data of the non-sampled position, and finally obtaining the geographic related data of the electromagnetic map
Figure 751033DEST_PATH_IMAGE096
。
According to the method of the first aspect of the present invention, said step S3 comprises the following steps.
The weight distribution result is:
Figure 85062DEST_PATH_IMAGE097
（11）
constructing the fused complete electromagnetic map is characterized by equation (12):
Figure 854215DEST_PATH_IMAGE098
（12）
wherein the content of the first and second substances,min J( )it is shown that the minimum value is found,
Figure 313009DEST_PATH_IMAGE099
a 2-norm representing the computation matrix,
Figure 765987DEST_PATH_IMAGE100
pair of representations
Figure 298731DEST_PATH_IMAGE101
The constraint of the whole variable is carried out,
Figure 215346DEST_PATH_IMAGE102
the parameters of the constraint are represented by a representation,
Figure 161437DEST_PATH_IMAGE101
and representing complete electromagnetic data obtained by fusing the electromagnetic related data and the geographic related data, wherein the electromagnetic map is obtained by visualizing the complete electromagnetic data.
Solving equation (12) specifically includes:
initializing intermediate parameters of the solution process
Figure 90210DEST_PATH_IMAGE103
，
Figure 317273DEST_PATH_IMAGE104
。
The loop is iterated through the loop,
Figure 204458DEST_PATH_IMAGE105
denotes the first
Figure 372265DEST_PATH_IMAGE106
The number of sub-iterations is,
Figure 836220DEST_PATH_IMAGE107
representing the total number of iterations, then:
Figure 671452DEST_PATH_IMAGE109
（13）
Figure 339325DEST_PATH_IMAGE110
（14）
Figure 257078DEST_PATH_IMAGE111
（15）
wherein the content of the first and second substances,
Figure 386708DEST_PATH_IMAGE112
denotes the first
Figure 545288DEST_PATH_IMAGE113
Intermediate parameters after sub-iteration
Figure 508695DEST_PATH_IMAGE114
The value of (a) is,
Figure 651095DEST_PATH_IMAGE115
are all intermediate parameters, and
Figure 784749DEST_PATH_IMAGE116
，
Figure 532256DEST_PATH_IMAGE117
obtained after the last iteration of the process,
Figure 994461DEST_PATH_IMAGE118
represents a pair of numbers comprising
Figure 952053DEST_PATH_IMAGE119
，
Figure 33273DEST_PATH_IMAGE114
Has a dimension of
Figure 494341DEST_PATH_IMAGE120
，
Figure 858939DEST_PATH_IMAGE121
OfDegree of
Figure 569406DEST_PATH_IMAGE122
。
Equation (16) holds:
Figure 188737DEST_PATH_IMAGE123
（16）
wherein, in formula (13)
Figure 35470DEST_PATH_IMAGE124
Operation satisfies
Figure 980424DEST_PATH_IMAGE125
，
Figure 647028DEST_PATH_IMAGE126
，
Figure 480771DEST_PATH_IMAGE127
，
Figure 916432DEST_PATH_IMAGE128
And is provided with
Figure 766707DEST_PATH_IMAGE129
，
Figure 186187DEST_PATH_IMAGE130
，
Figure 271955DEST_PATH_IMAGE131
，
Figure 562122DEST_PATH_IMAGE132
，
Figure 314790DEST_PATH_IMAGE133
Is shown in
Figure 955987DEST_PATH_IMAGE134
The quadrature operation is performed.
The invention discloses a system for constructing an electromagnetic map based on weight distribution in a second aspect. The electromagnetic map comprises electromagnetic related data and geographic related data; the system comprises: a first processing unit configured to: deploying a plurality of sensing nodes in a target area to acquire a sensing matrix, acquiring electromagnetic observation data of the target area through sampling, and constructing the electromagnetic relevant data of the electromagnetic map based on the sensing matrix and the electromagnetic observation data; a second processing unit configured to: acquiring geographic sampling data of the target area through sampling based on the deployed sensing nodes, and predicting geographic related data of non-sampling positions according to geographic correlation among the geographic sampling data so as to obtain the geographic related data of the electromagnetic map; a third processing unit configured to: and distributing weights to the electromagnetic relevant data of the electromagnetic map and the geographic relevant data of the electromagnetic map, and constructing a complete electromagnetic map of the target area by fusing the electromagnetic relevant data and the geographic relevant data distributed with the weights.
According to the system of the second aspect of the invention, the first processing unit is specifically configured to perform the following procedure.
Deploying a plurality of sensing nodes in the target area to acquire a sensing matrix, specifically comprising: equally dividing the target area into a plurality of sub-areas and numbering the sub-areas, wherein the numbering is
Figure 579866DEST_PATH_IMAGE136
The same number of sensing nodes are deployed in each sub-area, and the deployment mode is random deployment.
The electromagnetic correlation data is characterized by equation (1):
Figure 990119DEST_PATH_IMAGE138
（1）
wherein the content of the first and second substances,
Figure 306831DEST_PATH_IMAGE004
representing said electromagnetic correlation data to be constructed,
Figure 904165DEST_PATH_IMAGE139
，
Figure 66156DEST_PATH_IMAGE006
representing the total amount of electromagnetic data within the target region,
Figure 330916DEST_PATH_IMAGE007
representing the electromagnetic observation data in a form of a plurality of electromagnetic observations,
Figure 550020DEST_PATH_IMAGE140
，
Figure 41175DEST_PATH_IMAGE007
in which comprises
Figure 6857DEST_PATH_IMAGE141
The data of the bar is transmitted to the mobile terminal,
Figure 126123DEST_PATH_IMAGE010
representing an observation matrix.
Based on the electromagnetic correlation data
Figure 315796DEST_PATH_IMAGE011
In a sparse matrix
Figure 825406DEST_PATH_IMAGE012
Projection onto
Figure 732794DEST_PATH_IMAGE013
Will solve for the electromagnetic correlation data
Figure 503304DEST_PATH_IMAGE004
Conversion to solve formula (2):
Figure 4824DEST_PATH_IMAGE014
（2）
wherein the content of the first and second substances,
Figure 1730DEST_PATH_IMAGE142
representing calculation matrices
Figure 978389DEST_PATH_IMAGE143
The number of the norm is calculated,
Figure 744350DEST_PATH_IMAGE144
representing said electromagnetic related data
Figure 416771DEST_PATH_IMAGE017
In a sparse matrix
Figure 635394DEST_PATH_IMAGE018
The projection of the image onto the optical system,
Figure 277728DEST_PATH_IMAGE019
to represent the said perception matrix or matrices,
Figure 895267DEST_PATH_IMAGE145
，
Figure 800906DEST_PATH_IMAGE018
representing the sparse matrix, characterized by equation (3):
Figure 100300DEST_PATH_IMAGE146
（3）
wherein the content of the first and second substances,
Figure 156112DEST_PATH_IMAGE147
。
solving equation (2) specifically includes performing the following process for each sub-region.
Initializing each parameter: residual error
Figure 490141DEST_PATH_IMAGE023
Index matrix
Figure 970277DEST_PATH_IMAGE024
Number of elements
Figure 491388DEST_PATH_IMAGE148
，
Figure 616470DEST_PATH_IMAGE026
Representing step size, the size of which is the number of sub-regions and the number of iterations
Figure 805006DEST_PATH_IMAGE027
，
Figure 583606DEST_PATH_IMAGE149
Representing the perception matrix
Figure 326434DEST_PATH_IMAGE019
To (1) a
Figure 455539DEST_PATH_IMAGE029
Column, phase index
Figure 233003DEST_PATH_IMAGE150
Said projection
Figure 182504DEST_PATH_IMAGE013
Is estimated by
Figure 84732DEST_PATH_IMAGE151
。
Computing
Figure 145092DEST_PATH_IMAGE152
Before selecting the element values according to the calculated values
Figure 449166DEST_PATH_IMAGE033
A value of an element, and comparing the value of the element
Figure 590076DEST_PATH_IMAGE033
Value of each element in the perception matrix
Figure 573076DEST_PATH_IMAGE034
The set of corresponding column sequence numbers in
Figure 578072DEST_PATH_IMAGE035
。
Computing
Figure 595706DEST_PATH_IMAGE036
，
Figure 27956DEST_PATH_IMAGE037
(ii) a Computing a least squares solution
Figure 232672DEST_PATH_IMAGE038
(ii) a According to the least square solution
Figure 772851DEST_PATH_IMAGE153
The absolute value of each element item is selected from the values
Figure 644992DEST_PATH_IMAGE033
Individual element terms and extracting the
Figure 841618DEST_PATH_IMAGE033
Element item in the sensing matrix
Figure 533630DEST_PATH_IMAGE034
Of (1) corresponding column composition
Figure 473904DEST_PATH_IMAGE040
Corresponding column number constitution
Figure 934973DEST_PATH_IMAGE154
。
A new value of the residual error is calculated,
Figure 302500DEST_PATH_IMAGE042
，
Figure 619824DEST_PATH_IMAGE155
(ii) a Judgment of
Figure 98210DEST_PATH_IMAGE156
Whether the result is true; if so, then
Figure 679364DEST_PATH_IMAGE045
Update, update
Figure 748952DEST_PATH_IMAGE046
Will be
Figure 556502DEST_PATH_IMAGE046
In the corresponding position of
Figure 369737DEST_PATH_IMAGE047
Is set to
Figure 208992DEST_PATH_IMAGE157
(ii) a If not, further judgment is made
Figure 856006DEST_PATH_IMAGE049
Whether the result is true or not; if so, then
Figure 947589DEST_PATH_IMAGE050
，
Figure 439882DEST_PATH_IMAGE051
，
Figure 992698DEST_PATH_IMAGE052
And recalculating the new residual valuesr tc (ii) a If not, then
Figure 545034DEST_PATH_IMAGE045
，
Figure 858334DEST_PATH_IMAGE158
，
Figure 747793DEST_PATH_IMAGE052
And recalculating said new residual valuesr tc 。
Thereby solving the electromagnetic correlation data in the sub-area
Figure 298991DEST_PATH_IMAGE159
Fusing electromagnetic correlation data of the respective sub-regions
Figure 612773DEST_PATH_IMAGE160
，
Figure 741266DEST_PATH_IMAGE161
The electromagnetic related data of each sub-area after fusion is passed through a median filter to obtain the electromagnetic related data in the electromagnetic map in the target area
Figure 575361DEST_PATH_IMAGE057
Wherein, in the process,
Figure 840120DEST_PATH_IMAGE058
the median filtering process is indicated.
According to the system of the second aspect of the invention, the second processing unit is specifically configured to perform the following procedure.
The geographical sampling data of any sampling point position is represented by the reference signal received power of the sampling point position:
Figure 62154DEST_PATH_IMAGE059
（4）
wherein the content of the first and second substances,
Figure 350047DEST_PATH_IMAGE060
is representative of the reference signal received power,
Figure 250483DEST_PATH_IMAGE162
represents the geographical location of any of the sample points,
Figure 510694DEST_PATH_IMAGE062
a function representing a prediction model of the target,
Figure 169208DEST_PATH_IMAGE163
，
Figure 678818DEST_PATH_IMAGE064
a parameter representing the prediction model is determined,
Figure 117365DEST_PATH_IMAGE164
the number of functions of the prediction model and the number of parameters of the prediction model are both N,
Figure 294400DEST_PATH_IMAGE066
representing a random process.
Then the equation (5) holds:
Figure 389395DEST_PATH_IMAGE165
（5）
wherein the content of the first and second substances,
Figure 120722DEST_PATH_IMAGE068
representing the random process
Figure 365889DEST_PATH_IMAGE066
In the expectation of the above-mentioned method,
Figure 152359DEST_PATH_IMAGE166
indicating the geographic location of the two sample points,
Figure 824780DEST_PATH_IMAGE167
a correlation model is represented that is representative of,
Figure 636878DEST_PATH_IMAGE071
a parameter representative of the correlation model is determined,
Figure 92261DEST_PATH_IMAGE078
representing the random process
Figure 303275DEST_PATH_IMAGE066
The variance of (c).
Setting the sample point of the geographic sampling data as
Figure 474494DEST_PATH_IMAGE168
The corresponding sample value is
Figure 649254DEST_PATH_IMAGE074
To yield formula (6):
Figure 236224DEST_PATH_IMAGE075
（6）
wherein, the first and the second end of the pipe are connected with each other,
Figure 835833DEST_PATH_IMAGE169
to represent
Figure 847127DEST_PATH_IMAGE064
The predicted value of (a) is determined,
Figure 305921DEST_PATH_IMAGE170
represent
Figure 837528DEST_PATH_IMAGE078
The predicted value of (a) is determined,
Figure 291643DEST_PATH_IMAGE171
is composed of
Figure 473838DEST_PATH_IMAGE080
The matrix of the composition is composed of a plurality of matrixes,
Figure 419928DEST_PATH_IMAGE081
，
Figure 83122DEST_PATH_IMAGE082
to represent
Figure 532689DEST_PATH_IMAGE172
A correlation matrix of type sample points, the constituent elements of which are
Figure 482190DEST_PATH_IMAGE084
。
Defining a correlation function:
Figure 381489DEST_PATH_IMAGE085
（7）
wherein, the first and the second end of the pipe are connected with each other,
Figure 176270DEST_PATH_IMAGE173
is shown and
Figure 745922DEST_PATH_IMAGE061
the geographical location of a different one of the other sampling points,
Figure 538429DEST_PATH_IMAGE174
is the dimension of the sample point; and has the following formula (8):
Figure 190603DEST_PATH_IMAGE175
（8）
wherein the content of the first and second substances,
Figure 789074DEST_PATH_IMAGE089
representation solution
Figure 275550DEST_PATH_IMAGE176
The determinant of (1) adopts a spherical model, and comprises the following components:
Figure 973379DEST_PATH_IMAGE178
（9）
wherein, the first and the second end of the pipe are connected with each other,
Figure 443675DEST_PATH_IMAGE092
then, there is formula (10):
Figure 517941DEST_PATH_IMAGE179
（10）
wherein, the first and the second end of the pipe are connected with each other,
Figure 59256DEST_PATH_IMAGE094
represent
Figure 131249DEST_PATH_IMAGE060
The predicted value of (2).
Substituting the non-sampled position into formula (10) to obtain the geographic related data of the non-sampled position, and finally obtaining the geographic related data of the electromagnetic map
Figure 760944DEST_PATH_IMAGE180
。
According to the system of the second aspect of the present invention, the third processing unit is specifically configured to:
the weight distribution result is:
Figure 373322DEST_PATH_IMAGE097
（11）
constructing the fused complete electromagnetic map is characterized by equation (12):
Figure 97040DEST_PATH_IMAGE098
（12）
wherein the content of the first and second substances,min J( )it is shown that the minimum value is found,
Figure 136671DEST_PATH_IMAGE099
a 2-norm representing the computation matrix,
Figure 847138DEST_PATH_IMAGE100
presentation pair
Figure 325524DEST_PATH_IMAGE101
The constraint of the whole variable is carried out,
Figure 47624DEST_PATH_IMAGE181
the parameters of the constraint are represented by a representation,
Figure 851632DEST_PATH_IMAGE101
representing the complete electromagnetic data obtained by fusing the electromagnetic-related data and the geographic-related data, the complete electromagnetic data being visualizedAnd obtaining the electromagnetic map.
Solving equation (12) specifically includes:
initializing intermediate parameters of the solution process
Figure 210848DEST_PATH_IMAGE182
，
Figure 492925DEST_PATH_IMAGE104
。
The loop is iterated through the loop,
Figure 866269DEST_PATH_IMAGE105
denotes the first
Figure 513282DEST_PATH_IMAGE113
The number of sub-iterations is,
Figure 667183DEST_PATH_IMAGE183
representing the total number of iterations, then:
Figure 422124DEST_PATH_IMAGE185
（13）
Figure 853237DEST_PATH_IMAGE186
（14）
Figure 999047DEST_PATH_IMAGE111
（15）
wherein, the first and the second end of the pipe are connected with each other,
Figure 312348DEST_PATH_IMAGE112
is shown as
Figure 936228DEST_PATH_IMAGE106
Intermediate parameters after sub-iteration
Figure 281234DEST_PATH_IMAGE114
The value of (a) is set to (b),
Figure 597946DEST_PATH_IMAGE187
are all intermediate parameters, and
Figure 601805DEST_PATH_IMAGE116
，
Figure 29375DEST_PATH_IMAGE117
obtained after the last iteration, and then the final iteration is carried out,
Figure 294134DEST_PATH_IMAGE134
representing pairs of numbers, including
Figure 450922DEST_PATH_IMAGE188
，
Figure 738815DEST_PATH_IMAGE114
Has a dimension of
Figure 970076DEST_PATH_IMAGE189
，
Figure 27025DEST_PATH_IMAGE121
Has a dimension of
Figure 560905DEST_PATH_IMAGE122
。
Equation (16) holds:
Figure 926640DEST_PATH_IMAGE190
（16）
wherein, in formula (13)
Figure 571379DEST_PATH_IMAGE124
Operation satisfaction
Figure 748414DEST_PATH_IMAGE191
，
Figure 843409DEST_PATH_IMAGE126
，
Figure 105894DEST_PATH_IMAGE192
，
Figure 348132DEST_PATH_IMAGE128
And is provided with
Figure 317356DEST_PATH_IMAGE129
，
Figure 989777DEST_PATH_IMAGE130
，
Figure 473979DEST_PATH_IMAGE131
，
Figure 519908DEST_PATH_IMAGE193
，
Figure 405955DEST_PATH_IMAGE133
Is shown in
Figure 577174DEST_PATH_IMAGE134
The quadrature operation is performed.
A third aspect of the invention discloses an electronic device. The electronic device comprises a memory and a processor, the memory stores a computer program, and the processor, when executing the computer program, implements the steps of the method for constructing an electromagnetic map based on weight assignment according to any one of the first aspect of the present disclosure.
A fourth aspect of the invention discloses a computer-readable storage medium. The computer-readable storage medium has stored thereon a computer program which, when executed by a processor, implements the steps in a method of constructing an electromagnetic map based on weight assignment according to any one of the first aspect of the present disclosure.
In conclusion, the technical scheme provided by the invention can effectively overcome the defect that the accuracy of the kriging interpolation method in the electromagnetic map construction is greatly influenced by abnormal values; the inverse distance weighting interpolation method has low accuracy and serious bulls eye phenomenon; the result change of the nearest neighbor method is discontinuous; the problems that parameter setting is not easy by the Sheberd interpolation method and a local polynomial method is sensitive to the neighborhood distance are solved. Meanwhile, sufficient prior information is not needed, and compared with a method based on a propagation model, the method has the advantages that the application range is wider; the obtained electromagnetic data distribution is closer to real electromagnetic data, and the method has the characteristics of high precision, good robustness and strong consistency.
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 embodiments or the description in the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
Fig. 1 is a flowchart of a method of constructing an electromagnetic map based on weight assignment according to an embodiment of the present invention.
Fig. 2 is a schematic diagram of a system for constructing an electromagnetic map based on weight assignment according to an embodiment of the present invention.
Fig. 3 is a block diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, 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 obtained by a person skilled in the art without making any creative effort based on the embodiments in the present invention, belong to the protection scope of the present invention.
The technical terms related to the invention include:
IDW: inverse Distance Weighting, inverse Distance weighted interpolation.
KGA: kriging Algorithm, crigy interpolation method.
MSM: modified Shepard's Method, modified schilde interpolation.
NN: nearest Neighbor method.
And (3) LP: local Polynomial, local Polynomial method.
Bulls eye: in the interpolation process, a circle phenomenon with an interpolation point as a circle center is formed due to the existence of small or large data.
And (4) RSRP: reference Signal Receiving Power, which represents an average value of Power of a Reference Signal carried by a symbol in a wireless communication network terminal.
RMSE: root Mean Square Error.
LTE:4G mobile communication network.
Sampling point ratio: the sensing node number used for sampling accounts for the proportion of the total electromagnetic data.
In order to solve the problem of acquiring electromagnetic data in a non-cooperative environment, the invention provides a scheme for constructing an electromagnetic map based on weight distribution. The scheme takes a wireless communication network with RSRP parameters as a research object, and the main architecture is as follows: the method comprises the steps of performing grid division on a target area according to actual conditions, determining the electromagnetic signal resolution of the area, namely only researching the electromagnetic signal intensity (electromagnetic data) at specific interval positions, taking the electromagnetic data at the center of each grid as a representation of all the electromagnetic data in the grid, and taking the geographic position of the center as the geographic position of corresponding data, so as to determine the total amount of the electromagnetic data, thereby controlling the data volume within the processing range of a data processing center and ensuring that the data acquisition is more regular. In practical application, the ratio of the resolution to the side length of the target area is controlled to be not more than 0.5%, so that the electromagnetic distribution condition of the area can be better described. If the resolution of the electromagnetic data in the 4000 × 4000 meter area is 20 meters, the total amount of the electromagnetic data is 40000. Randomly deploying a certain number of distributed sensing nodes in a target area to perform data sampling on electromagnetic signals, and then respectively reconstructing a group of strong geographically related electromagnetic data and a group of strong electromagnetic related electromagnetic data; and obtaining complete electromagnetic data which are both geographically related and electromagnetically related by utilizing weight distribution, then improving the precision of the data according to the data correlation by utilizing gradient mapping, further obtaining high-precision complete electromagnetic data of a target area, and finally drawing an electromagnetic map through an equal intensity line.
The method can effectively realize the construction of the electromagnetic map, is suitable for the practical application scene without prior knowledge of electromagnetic propagation environment information in a target area, can still ensure higher precision when the proportion of sampling points to the total amount of electromagnetic data is 2 percent, the root mean square error of the obtained result is less than 1.5, and the data distribution of the obtained result is closer to the distribution of real data, and can be widely applied to the field of wireless communication, such as wireless communication network optimization, spectrum resource use evaluation, electromagnetic spectrum management and control, electromagnetic situation perception, battlefield electromagnetic situation control and the like.
The invention discloses a method for constructing an electromagnetic map based on weight distribution in a first aspect. The electromagnetic map includes electromagnetic related data and geographic related data. FIG. 1 is a flow diagram of a method of constructing an electromagnetic map based on weight assignment, according to an embodiment of the present invention; as shown in fig. 1, the method includes: s1, deploying a plurality of sensing nodes in a target area to obtain a sensing matrix, obtaining electromagnetic observation data of the target area through sampling, and constructing the electromagnetic related data of the electromagnetic map based on the sensing matrix and the electromagnetic observation data; s2, acquiring geographic sampling data of the target area through sampling based on the deployed sensing nodes, and predicting geographic related data of an unsampled position according to geographic correlation among the geographic sampling data to obtain the geographic related data of the electromagnetic map; s3, distributing weights to the electromagnetic relevant data of the electromagnetic map and the geographic relevant data of the electromagnetic map, and constructing the complete electromagnetic map of the target area by fusing the electromagnetic relevant data and the geographic relevant data distributed with the weights.
In some embodiments, in said step S1: firstly, randomly deploying a small number of sensing nodes in a target region, self-organizing the sensing nodes into a distributed sensing network, and then carrying out data sampling on electromagnetic signals in the region.
Deploying a number of sensing nodes in the target area toAcquiring a perception matrix, specifically comprising: equally dividing the target area into a plurality of sub-areas and numbering the sub-areas, wherein the numbering is
Figure 876568DEST_PATH_IMAGE194
The same number of sensing nodes are deployed in each sub-area, and the deployment mode is random deployment. Specifically, dividing a target area into four sub-areas equally, and numbering; and randomly throwing the same number of sensing nodes in each sub-region, and respectively predicting and reconstructing the electromagnetic data of each sub-region.
In some embodiments, the electromagnetic correlation data is characterized by equation (1):
Figure 463538DEST_PATH_IMAGE137
（1）
wherein the content of the first and second substances,
Figure 201163DEST_PATH_IMAGE004
representing said electromagnetic correlation data to be constructed,
Figure 684228DEST_PATH_IMAGE005
，
Figure 64394DEST_PATH_IMAGE006
representing the total amount of electromagnetic data within the target region,
Figure 392738DEST_PATH_IMAGE007
representing the electromagnetic observation data as a function of time,
Figure 253378DEST_PATH_IMAGE008
，
Figure 271510DEST_PATH_IMAGE007
in which comprises
Figure 545497DEST_PATH_IMAGE141
The data of the bar is transmitted to the mobile terminal,
Figure 739849DEST_PATH_IMAGE010
representing an observation matrix.
Based on the electromagnetic correlation data
Figure 923837DEST_PATH_IMAGE011
In a sparse matrix
Figure 138917DEST_PATH_IMAGE012
Projection onto
Figure 710320DEST_PATH_IMAGE013
Will solve for the electromagnetic correlation data
Figure 114887DEST_PATH_IMAGE004
Conversion to solve formula (2):
Figure 543595DEST_PATH_IMAGE014
（2）
wherein, the first and the second end of the pipe are connected with each other,
Figure 739696DEST_PATH_IMAGE142
representing calculation matrices
Figure 66903DEST_PATH_IMAGE143
The number of the norm is calculated,
Figure 337479DEST_PATH_IMAGE144
representing said electromagnetic related data
Figure 289867DEST_PATH_IMAGE017
In a sparse matrix
Figure 220652DEST_PATH_IMAGE018
The projection of the image onto the image plane is performed,
Figure 363051DEST_PATH_IMAGE019
to represent the said perception matrix or matrices,
Figure 437317DEST_PATH_IMAGE145
，
Figure 309458DEST_PATH_IMAGE018
representing the sparse matrix, characterized by equation (3):
Figure 768734DEST_PATH_IMAGE146
（3）
wherein the content of the first and second substances,
Figure 601692DEST_PATH_IMAGE147
。
solving equation (2) specifically includes performing the following steps for each sub-region.
Initializing each parameter: residual error
Figure 807545DEST_PATH_IMAGE023
Indexing matrix
Figure 268614DEST_PATH_IMAGE024
Number of elements
Figure 42666DEST_PATH_IMAGE148
，
Figure 221974DEST_PATH_IMAGE026
Representing step size, the size of which is the number of sub-regions and the number of iterations
Figure 228589DEST_PATH_IMAGE027
，
Figure 950689DEST_PATH_IMAGE149
Representing the perception matrix
Figure 489117DEST_PATH_IMAGE019
To (1) a
Figure 686880DEST_PATH_IMAGE029
Column, phase index
Figure 844323DEST_PATH_IMAGE150
Said projection
Figure 545563DEST_PATH_IMAGE013
Is estimated value of
Figure 517543DEST_PATH_IMAGE151
。
Computing
Figure 77968DEST_PATH_IMAGE152
Before selecting the element values according to the calculated values
Figure 570261DEST_PATH_IMAGE033
A value of an element, and comparing the value
Figure 391586DEST_PATH_IMAGE033
Value of each element in the sensing matrix
Figure 678342DEST_PATH_IMAGE034
The set of corresponding column sequence numbers in
Figure 871205DEST_PATH_IMAGE035
。
Computing
Figure 308134DEST_PATH_IMAGE036
，
Figure 249545DEST_PATH_IMAGE037
(ii) a Computing a least squares solution
Figure 441623DEST_PATH_IMAGE038
. According to the least square solution
Figure 835695DEST_PATH_IMAGE153
The absolute value of each element item is selected from
Figure 994757DEST_PATH_IMAGE033
The element items are extracted
Figure 525095DEST_PATH_IMAGE033
Element item in the sensing matrix
Figure 419233DEST_PATH_IMAGE034
Of (1) corresponding column composition
Figure 35022DEST_PATH_IMAGE040
Corresponding column number constitution
Figure 876070DEST_PATH_IMAGE154
. A new value of the residual error is calculated,
Figure 260915DEST_PATH_IMAGE042
，
Figure 916500DEST_PATH_IMAGE155
。
judgment of
Figure 19585DEST_PATH_IMAGE195
Whether the result is true; if so, then
Figure 523379DEST_PATH_IMAGE045
Update
Figure 638097DEST_PATH_IMAGE046
Will be
Figure 201933DEST_PATH_IMAGE046
In the corresponding position of
Figure 792314DEST_PATH_IMAGE047
Is set to
Figure 834220DEST_PATH_IMAGE196
(ii) a If not, further judgment is made
Figure 190727DEST_PATH_IMAGE197
Whether the result is true; if so, then
Figure 925465DEST_PATH_IMAGE050
，
Figure 737563DEST_PATH_IMAGE051
，
Figure 255263DEST_PATH_IMAGE052
And recalculating the new residual valuesr tc (ii) a If not, then
Figure 734786DEST_PATH_IMAGE045
，
Figure 46950DEST_PATH_IMAGE158
，
Figure 343415DEST_PATH_IMAGE052
And recalculating said new residual valuesr tc . Thereby solving the electromagnetic relevant data in the subarea
Figure 992702DEST_PATH_IMAGE198
Fusing electromagnetic correlation data of the respective sub-regions
Figure 467677DEST_PATH_IMAGE199
，
Figure 809796DEST_PATH_IMAGE161
The electromagnetic related data of each sub-area after fusion is passed through a median filter to obtain the electromagnetic related data in the electromagnetic map in the target area
Figure 596487DEST_PATH_IMAGE057
Wherein, in the process,
Figure 924831DEST_PATH_IMAGE058
the median filtering process is indicated.
In some embodiments, in step S2, the data for predicting the location of the non-sampling point can also be implemented by the geographical correlation between the locations of the sampling points. Specifically, the geographical sampling data of any sampling point position is represented by the reference signal received power of the sampling point position:
Figure 376016DEST_PATH_IMAGE059
（4）
wherein, the first and the second end of the pipe are connected with each other,
Figure 29983DEST_PATH_IMAGE060
is representative of the reference signal received power,
Figure 569548DEST_PATH_IMAGE061
represents the geographical location of any of the sample points,
Figure 826217DEST_PATH_IMAGE200
a function representing a prediction model of the target,
Figure 275784DEST_PATH_IMAGE063
，
Figure 959707DEST_PATH_IMAGE064
a parameter representing the prediction model is determined,
Figure 593426DEST_PATH_IMAGE065
the number of functions of the prediction model and the number of parameters of the prediction model are both N,
Figure 60310DEST_PATH_IMAGE066
representing a random process.
Then the equation (5) holds:
Figure 629963DEST_PATH_IMAGE201
（5）
wherein, the first and the second end of the pipe are connected with each other,
Figure 15945DEST_PATH_IMAGE068
representing the random process
Figure 402540DEST_PATH_IMAGE066
In the expectation of the above-mentioned method,
Figure 1011DEST_PATH_IMAGE166
indicating the geographical location of the two sample points,
Figure 894012DEST_PATH_IMAGE167
a correlation model is represented that is representative of,
Figure 185316DEST_PATH_IMAGE202
a parameter representing the correlation model is determined,
Figure 62136DEST_PATH_IMAGE078
representing the random process
Figure 198720DEST_PATH_IMAGE066
The variance of (c).
Let the sample point of the geographic sampling data be
Figure 497893DEST_PATH_IMAGE203
Corresponding to the samples being
Figure 632202DEST_PATH_IMAGE074
To yield formula (6):
Figure 324215DEST_PATH_IMAGE075
（6）
wherein the content of the first and second substances,
Figure 795647DEST_PATH_IMAGE169
to represent
Figure 397661DEST_PATH_IMAGE064
The predicted value of (a) is determined,
Figure 293417DEST_PATH_IMAGE170
to represent
Figure 613671DEST_PATH_IMAGE078
The predicted value of (a) is obtained,
Figure 357636DEST_PATH_IMAGE171
to comprise
Figure 673211DEST_PATH_IMAGE080
In which
Figure 742798DEST_PATH_IMAGE204
The matrix of the type is such that,
Figure 143824DEST_PATH_IMAGE081
，
Figure 563916DEST_PATH_IMAGE082
to represent
Figure 265156DEST_PATH_IMAGE172
A correlation matrix of type sample points, the constituent elements of which are
Figure 443328DEST_PATH_IMAGE084
。
Defining a correlation function:
Figure 597229DEST_PATH_IMAGE085
（7）
wherein the content of the first and second substances,
Figure 417417DEST_PATH_IMAGE086
is shown and
Figure 707584DEST_PATH_IMAGE061
the geographical location of a different one of the other sampling points,
Figure 322236DEST_PATH_IMAGE087
is the dimension of the sample point; and has the following formula (8):
Figure 229012DEST_PATH_IMAGE088
（8）
wherein, the first and the second end of the pipe are connected with each other,
Figure 318804DEST_PATH_IMAGE089
representation solution
Figure 729056DEST_PATH_IMAGE176
The determinant of (2) adopts a spherical model. Then there are:
Figure 45768DEST_PATH_IMAGE205
（9）
wherein, the first and the second end of the pipe are connected with each other,
Figure 174261DEST_PATH_IMAGE092
then, there is formula (10):
Figure 273935DEST_PATH_IMAGE179
（10）
wherein the content of the first and second substances,
Figure 945219DEST_PATH_IMAGE094
to represent
Figure 898744DEST_PATH_IMAGE060
The predicted value of (2).
Substituting the non-sampled position into formula (10) to obtain the geographic related data of the non-sampled position, and finally obtaining the geographic related data of the electromagnetic map
Figure 921058DEST_PATH_IMAGE180
. Namely, the geographical position of the non-sampling point is brought into, so that the prediction of the non-sampling position data can be realized. From this a complete set of electromagnetic data of strong geographical relevance can be reconstructed.
In some embodiments, at step S3, the electromagnetic data is not fully correlated to geographic location, although the geographic environment can affect the electromagnetic data to a large extent. Therefore, the accuracy of the obtained data has room for improvement. The geographic correlation and the electromagnetic correlation are fused by using a weight distribution strategy, so that a group of electromagnetic data with stronger correlation can be obtained and used for improving the accuracy of the electromagnetic data.
In some embodiments, the step S3 specifically includes the following process.
The weight distribution result is:
Figure 824423DEST_PATH_IMAGE097
（11）
constructing the fused complete electromagnetic map is characterized by equation (12):
Figure 740426DEST_PATH_IMAGE098
（12）
wherein, the first and the second end of the pipe are connected with each other,min J( )it is shown that the minimum value is found,
Figure 274307DEST_PATH_IMAGE099
a 2-norm representing the computation matrix,
Figure 640042DEST_PATH_IMAGE100
pair of representations
Figure 81519DEST_PATH_IMAGE101
The constraint of the whole variable is carried out,
Figure 320870DEST_PATH_IMAGE181
the parameters of the constraint are represented by a representation,
Figure 291231DEST_PATH_IMAGE206
is composed of
Figure 22558DEST_PATH_IMAGE101
And
Figure 264796DEST_PATH_IMAGE207
the dimension(s) of (a) is,
Figure 296337DEST_PATH_IMAGE101
representing fusion of the electromagnetic-related data and the instituteAnd obtaining complete electromagnetic data after the geographic relevant data, wherein the electromagnetic map is obtained after the complete electromagnetic data is visualized.
Solving equation (12) specifically includes the following procedure.
Initializing intermediate parameters of the solution process
Figure 296654DEST_PATH_IMAGE103
，
Figure 780856DEST_PATH_IMAGE104
。
The loop is iterated through a plurality of cycles,
Figure 561206DEST_PATH_IMAGE105
denotes the first
Figure 40729DEST_PATH_IMAGE113
The number of sub-iterations is,
Figure 352893DEST_PATH_IMAGE183
representing the total number of iterations, then:
Figure 58811DEST_PATH_IMAGE209
（13）
Figure 973678DEST_PATH_IMAGE186
（14）
Figure 672423DEST_PATH_IMAGE111
（15）
wherein the content of the first and second substances,
Figure 686646DEST_PATH_IMAGE112
is shown as
Figure 473337DEST_PATH_IMAGE106
Intermediate parameters after sub-iteration
Figure 863998DEST_PATH_IMAGE114
The value of (a) is,
Figure 721708DEST_PATH_IMAGE187
are all intermediate parameters, and
Figure 641254DEST_PATH_IMAGE116
，
Figure 321765DEST_PATH_IMAGE117
obtained after the last iteration, and then the final iteration is carried out,
Figure 250538DEST_PATH_IMAGE134
representing pairs of numbers, including
Figure 962754DEST_PATH_IMAGE188
，
Figure 53201DEST_PATH_IMAGE114
Has a dimension of
Figure 486588DEST_PATH_IMAGE189
，
Figure 953472DEST_PATH_IMAGE121
Has a dimension of
Figure 647759DEST_PATH_IMAGE122
And equation (16) holds.
Figure 375019DEST_PATH_IMAGE190
（16）
Wherein, in formula (13)
Figure 358018DEST_PATH_IMAGE124
Operation satisfaction
Figure 628594DEST_PATH_IMAGE191
，
Figure 52753DEST_PATH_IMAGE126
，
Figure 609636DEST_PATH_IMAGE192
，
Figure 811423DEST_PATH_IMAGE128
And is provided with
Figure 88952DEST_PATH_IMAGE129
，
Figure 164355DEST_PATH_IMAGE130
，
Figure 33085DEST_PATH_IMAGE131
，
Figure 66376DEST_PATH_IMAGE193
，
Figure 944333DEST_PATH_IMAGE133
Is shown in
Figure 343084DEST_PATH_IMAGE134
The quadrature operation is performed.
The orthogonal operation may be defined as:
Figure 382716DEST_PATH_IMAGE210
then, there is formula (17):
Figure 824674DEST_PATH_IMAGE211
（17）
wherein the content of the first and second substances,
Figure DEST_PATH_IMAGE212
indicates that the maximum value limiting the array is not more than n 2 And the minimum value is not more than n 1 And satisfy the definitions
Figure 709584DEST_PATH_IMAGE213
。
Therefore, the strong electromagnetic correlation electromagnetic data which are obtained by reconstruction and the strong geographical correlation electromagnetic data are subjected to weight distribution, so that the strong correlation electromagnetic data which are more consistent with the actual situation can be obtained, the precision is improved, and the high-precision electromagnetic map reconstruction is realized.
The invention discloses a system for constructing an electromagnetic map based on weight distribution in a second aspect. The electromagnetic map comprises electromagnetic related data and geographic related data; FIG. 2 is a schematic diagram of a system for building an electromagnetic map based on weight assignment, according to an embodiment of the present invention; as shown in fig. 2, the system 200 includes: a first processing unit 201 configured to: deploying a plurality of sensing nodes in a target area to acquire a sensing matrix, acquiring electromagnetic observation data of the target area through sampling, and constructing the electromagnetic relevant data of the electromagnetic map based on the sensing matrix and the electromagnetic observation data; a second processing unit 202 configured to: acquiring geographic sampling data of the target area through sampling based on the deployed sensing nodes, and predicting geographic related data of non-sampling positions according to geographic correlation among the geographic sampling data so as to obtain the geographic related data of the electromagnetic map; a third processing unit 203 configured to: weights are assigned to the electromagnetic-related data of the electromagnetic map and the geographical-related data of the electromagnetic map, and a complete electromagnetic map of the target area is constructed by fusing the weighted electromagnetic-related data and the geographical-related data.
According to the system of the second aspect of the present invention, the first processing unit 201 is specifically configured to perform the following processes.
Deploying a plurality of sensing nodes in the target area to acquire a sensing matrix, specifically comprising: equally dividing the target area into a plurality of sub-areas and numbering, wherein the number is
Figure 166104DEST_PATH_IMAGE001
The same number of sensing nodes are deployed in each sub-area, and the deployment mode is random deployment.
The electromagnetic correlation data is characterized by equation (1):
Figure 970112DEST_PATH_IMAGE002
（1）
wherein the content of the first and second substances,
Figure 305891DEST_PATH_IMAGE004
representing said electromagnetic correlation data to be constructed,
Figure 260072DEST_PATH_IMAGE214
，
Figure 695733DEST_PATH_IMAGE006
representing the total amount of electromagnetic data within the target region,
Figure 998538DEST_PATH_IMAGE007
representing the electromagnetic observation data as a function of time,
Figure DEST_PATH_IMAGE215
，
Figure 903171DEST_PATH_IMAGE007
in which comprises
Figure 275026DEST_PATH_IMAGE009
The number of the pieces of data is set,
Figure 830772DEST_PATH_IMAGE010
representing an observation matrix.
Based on the electromagnetic correlation data
Figure 648687DEST_PATH_IMAGE011
In a sparse matrix
Figure 430829DEST_PATH_IMAGE012
Projection onto
Figure 992391DEST_PATH_IMAGE013
Will solve for the electromagnetic correlation data
Figure 399714DEST_PATH_IMAGE004
Conversion to solve formula (2):
Figure 122951DEST_PATH_IMAGE216
（2）
wherein the content of the first and second substances,
Figure 923548DEST_PATH_IMAGE015
representing calculation matrices
Figure 351118DEST_PATH_IMAGE016
The norm of the number of the first-order-of-arrival,
Figure 553560DEST_PATH_IMAGE013
representing said electromagnetic related data
Figure 772665DEST_PATH_IMAGE017
In a sparse matrix
Figure 60558DEST_PATH_IMAGE018
The projection of the image onto the optical system,
Figure 167185DEST_PATH_IMAGE019
to represent the said perception matrix or matrices,
Figure 552030DEST_PATH_IMAGE020
，
Figure 210544DEST_PATH_IMAGE018
representing the sparse matrix, characterized by equation (3):
Figure 48050DEST_PATH_IMAGE021
（3）
wherein the content of the first and second substances,
Figure 486598DEST_PATH_IMAGE022
。
solving equation (2) specifically includes, for each sub-region:
initializing each parameter: residual error
Figure 866894DEST_PATH_IMAGE023
Indexing matrix
Figure 696310DEST_PATH_IMAGE024
Number of elements
Figure 427637DEST_PATH_IMAGE025
，
Figure 266280DEST_PATH_IMAGE026
Representing step size, the size of which is the number of sub-regions and the number of iterations
Figure 622787DEST_PATH_IMAGE027
，
Figure 498471DEST_PATH_IMAGE028
Representing the perception matrix
Figure 576148DEST_PATH_IMAGE019
To (1) a
Figure 687324DEST_PATH_IMAGE029
Column, phase index
Figure 776633DEST_PATH_IMAGE030
Said projection
Figure 541327DEST_PATH_IMAGE013
Is estimated value of
Figure 713158DEST_PATH_IMAGE031
。
Calculating out
Figure 628024DEST_PATH_IMAGE032
Before selecting the element values according to the calculated values
Figure 165316DEST_PATH_IMAGE033
A value of an element, and comparing the value of the element
Figure 100911DEST_PATH_IMAGE033
Value of each element in the perception matrix
Figure 762968DEST_PATH_IMAGE034
The set of corresponding column sequence numbers in
Figure 481525DEST_PATH_IMAGE035
。
Calculating out
Figure 404482DEST_PATH_IMAGE036
，
Figure 917503DEST_PATH_IMAGE037
(ii) a Computing a least squares solution
Figure 63926DEST_PATH_IMAGE038
(ii) a According to the least square solution
Figure 320595DEST_PATH_IMAGE039
The absolute value of each element item is selected from the values
Figure 707845DEST_PATH_IMAGE033
Individual element terms and extracting the
Figure 657346DEST_PATH_IMAGE033
Element item in the perception matrix
Figure 418629DEST_PATH_IMAGE034
Composition of corresponding column in (1)
Figure 617004DEST_PATH_IMAGE040
Formation of corresponding column numbers
Figure 514553DEST_PATH_IMAGE041
。
A new residual value is calculated and,
Figure 775901DEST_PATH_IMAGE042
，
Figure 24480DEST_PATH_IMAGE043
(ii) a Judgment of
Figure 763897DEST_PATH_IMAGE044
Whether the result is true or not; if so, then
Figure 67619DEST_PATH_IMAGE045
Update, update
Figure 968710DEST_PATH_IMAGE046
Will be
Figure 642268DEST_PATH_IMAGE046
In the corresponding position of
Figure 388638DEST_PATH_IMAGE047
Is set to
Figure 664374DEST_PATH_IMAGE048
(ii) a If not, further judgment is made
Figure 329842DEST_PATH_IMAGE049
Whether the result is true; if so, then
Figure 693958DEST_PATH_IMAGE050
，
Figure 306336DEST_PATH_IMAGE051
，
Figure 702158DEST_PATH_IMAGE052
And recalculating the new residual valuesr tc (ii) a If not, then
Figure 413893DEST_PATH_IMAGE045
，
Figure 124360DEST_PATH_IMAGE053
，
Figure 274850DEST_PATH_IMAGE052
And recalculating the new residual valuesr tc 。
Thereby solving the electromagnetic correlation data in the sub-area
Figure 525178DEST_PATH_IMAGE054
Fusing electromagnetic correlation data of the respective sub-regions
Figure DEST_PATH_IMAGE217
，
Figure 407815DEST_PATH_IMAGE056
The electromagnetic related data of each sub-area after fusion is passed through a median filter to obtain the electromagnetic related data in the electromagnetic map in the target area
Figure 605578DEST_PATH_IMAGE057
Wherein, in the step (A),
Figure 887655DEST_PATH_IMAGE058
representing a median filtering process.
According to the system of the second aspect of the present invention, the second processing unit 202 is specifically configured to: the geographical sampling data of any sampling point position is expressed by the reference signal receiving power of the sampling point position:
Figure 992489DEST_PATH_IMAGE059
（4）
wherein the content of the first and second substances,
Figure 108344DEST_PATH_IMAGE060
is representative of the reference signal received power,
Figure 934349DEST_PATH_IMAGE061
representing geography of any of the sample pointsThe position of the mobile phone is determined,
Figure 551275DEST_PATH_IMAGE062
a function representing a prediction model of the model,
Figure 513546DEST_PATH_IMAGE063
，
Figure 62951DEST_PATH_IMAGE064
a parameter representing the prediction model is determined,
Figure 376252DEST_PATH_IMAGE065
the number of functions of the prediction model and the number of parameters of the prediction model are both N,
Figure 131DEST_PATH_IMAGE066
representing a random process.
Then the equation (5) holds:
Figure 410384DEST_PATH_IMAGE067
（5）
wherein, the first and the second end of the pipe are connected with each other,
Figure 727096DEST_PATH_IMAGE068
representing the random process
Figure 728025DEST_PATH_IMAGE066
In the expectation that the position of the target is not changed,
Figure 155596DEST_PATH_IMAGE069
indicating the geographic location of the two sample points,
Figure 420355DEST_PATH_IMAGE070
a correlation model is represented that is representative of,
Figure 907968DEST_PATH_IMAGE071
a parameter representative of the correlation model is determined,
Figure 399123DEST_PATH_IMAGE072
representing the random process
Figure 630385DEST_PATH_IMAGE066
The variance of (c).
Setting the sample point of the geographic sampling data as
Figure 749650DEST_PATH_IMAGE073
The corresponding sample value is
Figure 139656DEST_PATH_IMAGE074
To give formula (6):
Figure 977162DEST_PATH_IMAGE075
（6）
wherein, the first and the second end of the pipe are connected with each other,
Figure 746535DEST_PATH_IMAGE076
to represent
Figure 985886DEST_PATH_IMAGE064
The predicted value of (a) is determined,
Figure 425089DEST_PATH_IMAGE077
to represent
Figure 15470DEST_PATH_IMAGE078
The predicted value of (a) is determined,
Figure 854113DEST_PATH_IMAGE079
to comprise
Figure 702900DEST_PATH_IMAGE080
In which
Figure 578583DEST_PATH_IMAGE218
The matrix of the type is such that,
Figure 125102DEST_PATH_IMAGE081
，
Figure 767436DEST_PATH_IMAGE082
to represent
Figure 122325DEST_PATH_IMAGE083
A correlation matrix of type sample points, the constituent elements of which are
Figure 559123DEST_PATH_IMAGE084
。
Defining a correlation function:
Figure 855587DEST_PATH_IMAGE085
（7）
wherein, the first and the second end of the pipe are connected with each other,
Figure 583503DEST_PATH_IMAGE086
is shown and
Figure 917532DEST_PATH_IMAGE061
the geographical location of a different one of the other sampling points,
Figure 400597DEST_PATH_IMAGE087
is the dimension of the sample point; and has the following formula (8):
Figure 718446DEST_PATH_IMAGE088
（8）
wherein the content of the first and second substances,
Figure 575019DEST_PATH_IMAGE089
representation solving
Figure 435659DEST_PATH_IMAGE082
The determinant of (1) adopts a spherical model, and comprises the following components:
Figure DEST_PATH_IMAGE219
（9）
wherein the content of the first and second substances,
Figure 292888DEST_PATH_IMAGE092
then, there is formula (10):
Figure 566874DEST_PATH_IMAGE093
（10）
wherein the content of the first and second substances,
Figure 820614DEST_PATH_IMAGE094
to represent
Figure 270181DEST_PATH_IMAGE095
The predicted value of (2).
Substituting the non-sampled position into formula (10) to obtain the geographic related data of the non-sampled position, and finally obtaining the geographic related data of the electromagnetic map
Figure 954103DEST_PATH_IMAGE096
。
According to the system of the second aspect of the present invention, the third processing unit 203 is specifically configured to:
the weight distribution result is:
Figure 325172DEST_PATH_IMAGE220
（11）
constructing the fused complete electromagnetic map is characterized by equation (12):
Figure 385532DEST_PATH_IMAGE098
（12）
wherein the content of the first and second substances,min J( )it is shown that the minimum value is found,
Figure 548660DEST_PATH_IMAGE099
a 2-norm representing the computation matrix,
Figure 275921DEST_PATH_IMAGE100
presentation pair
Figure 258920DEST_PATH_IMAGE101
The constraint of the full variable is carried out,
Figure 591812DEST_PATH_IMAGE102
a parameter of the constraint is represented by,
Figure 343868DEST_PATH_IMAGE206
is composed of
Figure 369593DEST_PATH_IMAGE101
And
Figure 839888DEST_PATH_IMAGE207
the dimension(s) of (a) is,
Figure 976471DEST_PATH_IMAGE101
and representing complete electromagnetic data obtained by fusing the electromagnetic related data and the geographic related data, wherein the electromagnetic map is obtained by visualizing the complete electromagnetic data.
Solving equation (12) specifically includes the following procedure.
Initializing intermediate parameters of the solution process
Figure 721049DEST_PATH_IMAGE103
，
Figure 917675DEST_PATH_IMAGE104
。
The loop is iterated through the loop,
Figure 140846DEST_PATH_IMAGE105
denotes the first
Figure 487645DEST_PATH_IMAGE106
The number of iterations is then repeated,
Figure 479872DEST_PATH_IMAGE107
representing the total number of iterations, then:
Figure DEST_PATH_IMAGE221
（13）
Figure 454256DEST_PATH_IMAGE110
（14）
Figure 305669DEST_PATH_IMAGE111
（15）
wherein the content of the first and second substances,
Figure 49634DEST_PATH_IMAGE112
is shown as
Figure 37313DEST_PATH_IMAGE113
Intermediate parameters after sub-iteration
Figure 575741DEST_PATH_IMAGE114
The value of (a) is,
Figure 911520DEST_PATH_IMAGE115
are all intermediate parameters, and
Figure 600122DEST_PATH_IMAGE116
，
Figure 566941DEST_PATH_IMAGE117
obtained after the last iteration of the process,
Figure 682795DEST_PATH_IMAGE118
represents a pair of numbers comprising
Figure 529308DEST_PATH_IMAGE119
，
Figure 880655DEST_PATH_IMAGE114
Has the dimension of
Figure 436401DEST_PATH_IMAGE120
，
Figure 723157DEST_PATH_IMAGE121
Has the dimension of
Figure 770879DEST_PATH_IMAGE122
。
Equation (16) holds:
Figure 925916DEST_PATH_IMAGE123
（16）
wherein, in formula (13)
Figure 739764DEST_PATH_IMAGE124
Operation satisfaction
Figure 790897DEST_PATH_IMAGE125
，
Figure 325914DEST_PATH_IMAGE126
，
Figure 19064DEST_PATH_IMAGE127
，
Figure 424769DEST_PATH_IMAGE128
And is provided with
Figure 909452DEST_PATH_IMAGE129
，
Figure 259662DEST_PATH_IMAGE130
，
Figure 959765DEST_PATH_IMAGE131
，
Figure 344610DEST_PATH_IMAGE132
，
Figure 3124DEST_PATH_IMAGE133
Is shown in
Figure 981576DEST_PATH_IMAGE134
The quadrature operation is performed.
A third aspect of the invention discloses an electronic device. The electronic device comprises a memory storing a computer program and a processor implementing the steps of a method for constructing an electromagnetic map based on weight assignment according to any one of the first aspect of the present disclosure when the computer program is executed by the processor.
Fig. 3 is a block diagram of an electronic device according to an embodiment of the present invention, and as shown in fig. 3, the electronic device includes a processor, a memory, a communication interface, a display screen, and an input device, which are connected by a system bus. Wherein the processor of the electronic device is configured to provide computing and control capabilities. The memory of the electronic equipment comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of an operating system and computer programs in the non-volatile storage medium. The communication interface of the electronic device is used for carrying out wired or wireless communication with an external terminal, and the wireless communication can be realized through WIFI, an operator network, near Field Communication (NFC) or other technologies. The display screen of the electronic equipment can be a liquid crystal display screen or an electronic ink display screen, and the input device of the electronic equipment can be a touch layer covered on the display screen, a key, a track ball or a touch pad arranged on the shell of the electronic equipment, an external keyboard, a touch pad or a mouse and the like.
It will be understood by those skilled in the art that the structure shown in fig. 3 is only a partial block diagram related to the technical solution of the present disclosure, and does not constitute a limitation to the electronic device to which the solution of the present invention is applied, and a specific electronic device may include more or less components than those shown in the figure, or combine some components, or have different arrangements of components.
A fourth aspect of the invention discloses a computer-readable storage medium. The computer-readable storage medium has stored thereon a computer program which, when executed by a processor, implements the steps in a method of constructing an electromagnetic map based on weight assignment according to any one of the first aspect of the present disclosure.
In conclusion, the technical scheme provided by the invention can effectively overcome the defect that the accuracy of the kriging interpolation method in the electromagnetic map construction is greatly influenced by abnormal values; the inverse distance weighting interpolation method has low accuracy and serious bulls eye phenomenon; the result change of the nearest neighbor method is discontinuous; the problems that parameter setting is not easy by the Sheberd interpolation method and a local polynomial method is sensitive to the neighborhood distance are solved. Meanwhile, sufficient prior information is not needed, and compared with a method based on a propagation model, the method has the advantages that the application range is wider; the obtained electromagnetic data distribution is closer to real electromagnetic data, and the method has the characteristics of high precision, good robustness and strong consistency.
Specific examples
The method comprises the steps of extracting a 4000-meter area in a Brussels map, arranging three base stations (hereinafter referred to as radiation sources) in the area, thereby constructing an LTE mobile communication network, taking the radio coverage condition of the network as experimental data, and adopting RSRP as a parameter for measuring the strength of electromagnetic signals. The maximum transmitting power is 43dBm, the frequency bandwidth is set to 2110FDD-10MHz, the radius of a cell is 350 meters, and the used propagation model is an Okumura-Hata model. The electromagnetic resolution of the area is 20 m, so 40000 grids can be obtained, namely the total amount of electromagnetic data is 40000, the area is divided into four sub-areas, 200 low-cost distributed sensing nodes are randomly scattered in the four sub-areas respectively, and 800 nodes are scattered in the area in total, namely the percentage of sampling points (sensing nodes) is 2%.
In order to verify the effect of the method provided by the invention, an inverse distance weighted interpolation method (IDW), a nearest neighbor method (NN), a kriging interpolation method (KGA), an improved schilder interpolation method (MSM) and a local polynomial method (LP) are applied to the reconstruction of the electromagnetic map and compared with the method provided by the invention. In an electromagnetic map constructed by an inverse distance weighting interpolation method, the distribution of the isomagnetic lines is very disordered, the abnormal values are many, the bullseye phenomenon is serious, and the map quality is very low; in the electromagnetic map constructed by the nearest neighbor method, the isomagnetic lines are disordered and serrated, the numerical change is discontinuous, and the map quality is low; the electromagnetic map constructed by the kriging interpolation method has poor construction effect in the area where the radiation source is located, the form of the isomagnetic line is disordered, more abnormal values exist, the receiving power distribution condition of the area is difficult to clearly display, and the map quality is low; the electromagnetic map constructed by the improved Sheberd interpolation method has clear isometric lines and better form, but has a bullseye phenomenon and holes, so that the map has better quality; the medium magnetic wire in the electromagnetic map constructed by the local polynomial method has good shape and no bulls eye phenomenon, but abnormal values exist in a piece, the electromagnetic distribution condition of the area where the radiation source is located cannot be accurately displayed, and the quality of the map is very low. The electromagnetic map constructed by the invention has good shape of the medium magnetic wire, particularly the area where the radiation source is positioned, can better describe the change of the receiving power of the area, has less abnormal values, does not have the bulls eye phenomenon, and has good quality of the constructed map.
Besides the quality of the map construction, the accuracy of the obtained result is also the key to measure the effect of the used method. Table 1 shows the root mean square error and the coefficient of determination R of the six methods when the sampling point ratio is 2% 2 。
Table 1: data accuracy measurement
Figure 16528DEST_PATH_IMAGE222
As can be seen from table 1, the present invention has the smallest root mean square error value and the largest determining coefficient, so that the average deviation degree of the result obtained by the method and the actual electromagnetic data is the smallest, the data distribution of the obtained result is closest to the distribution of the actual data, and the accuracy is the highest among the six methods. Theoretically, the more sensing nodes used for sampling, the higher the accuracy of the map construction. However, if the number of available nodes is less than the predetermined number for uncontrollable reasons, the method must ensure that the map is constructed with high precision and with a small enough variation compared to the predetermined case, i.e. with good robustness.
In addition, when the sampling point proportion is reduced, the total increase amplitude of the root mean square error of the method is slightly larger than that of a kriging interpolation method and an inverse distance weighting interpolation method, but is smaller than that of other methods; however, when the number of sampling nodes is increased, the error fluctuation phenomenon exists in the kriging interpolation method, and the root mean square error value of the inverse distance weighting interpolation method is far larger than that of the present invention, so that the present invention has good robustness and stable effect.
The invention provides a scheme for constructing an electromagnetic map based on weight distribution by fully considering the weakness of the existing electromagnetic map construction method and through the ideas of relevance distribution and precision re-improvement. According to actual requirements, a target area is divided in a rasterization mode, the electromagnetic signal resolution of the area is determined, namely only the electromagnetic signal strength (electromagnetic data) at specific interval positions is researched, the electromagnetic data at the center of each grid is taken as a representative of all the electromagnetic data in the grid, meanwhile, the geographic position of the center is taken as the geographic position of corresponding data, so that the total amount of the electromagnetic data is determined, the data volume is controlled in the processing range of a data processing center, the data acquisition is ensured to be more regular, and a certain number of sensing nodes are randomly deployed to perform data sampling on the electromagnetic signals in the target area; and reconstructing a group of strong geographically related electromagnetic data and a group of strong electromagnetically related electromagnetic data according to the sampled data. The electromagnetic data distribution obtained through the weight distribution is closer to real electromagnetic data, namely the correlation is stronger; and then, the precision of the strong correlation data is improved, so that the overall electromagnetic data of the high-precision target area can be obtained, and finally, a complete electromagnetic map is obtained through drawing an equal-strength line.
In conclusion, the scheme for constructing the electromagnetic map based on the weight distribution has the effect superior to that of the conventional reverse distance weighting interpolation method, the nearest neighbor method, the kriging interpolation method, the improved schilder interpolation method and the local polynomial method, and the required sampling point accounts for 2 percent at least. The method can be widely applied to the field of wireless communication networks, such as the deployment and optimization of 4G/5G/6G networks, the fields of frequency spectrum resource use evaluation, electromagnetic frequency spectrum management and control, electromagnetic situation perception, battlefield electromagnetic situation control and the like, and the technology has important theory and application value.
It should be noted that the technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, however, as long as there is no contradiction between the combinations of the technical features, the scope of the present description should be considered. The above-mentioned embodiments only express several embodiments of the present invention, and the description thereof is specific and detailed, but not to be understood as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the inventive concept, which falls within the scope of the present invention. Therefore, the protection scope of the present patent shall be subject to the appended claims.

Claims (10)

1. A method for constructing an electromagnetic map based on weight distribution is characterized in that the electromagnetic map comprises electromagnetic related data and geographic related data; the method comprises the following steps:
s1, deploying a plurality of sensing nodes in a target area to obtain a sensing matrix, acquiring electromagnetic observation data of the target area through sampling, and constructing the electromagnetic related data of the electromagnetic map based on the sensing matrix and the electromagnetic observation data;
s2, acquiring geographic sampling data of the target area through sampling based on the deployed sensing nodes, and predicting geographic related data of an unsampled position according to geographic correlation among the geographic sampling data to obtain the geographic related data of the electromagnetic map;
s3, distributing weights to the electromagnetic relevant data of the electromagnetic map and the geographic relevant data of the electromagnetic map, and constructing the complete electromagnetic map of the target area by fusing the electromagnetic relevant data and the geographic relevant data distributed with the weights.
2. The method for constructing an electromagnetic map based on weight distribution according to claim 1, wherein in step S1:
deploying a plurality of sensing nodes in the target area to acquire a sensing matrix, specifically comprising: the target area and the likeDividing the sub-area into a plurality of sub-areas and numbering the sub-areas
Figure 553555DEST_PATH_IMAGE002
Deploying the same number of sensing nodes in each sub-area in a random deployment mode;
the electromagnetic correlation data is characterized by equation (1):
Figure 608711DEST_PATH_IMAGE004
（1）
wherein, the first and the second end of the pipe are connected with each other,
Figure 326131DEST_PATH_IMAGE005
representing said electromagnetic correlation data to be constructed,
Figure 659024DEST_PATH_IMAGE007
，
Figure 676658DEST_PATH_IMAGE008
representing the total amount of electromagnetic data within the target region,
Figure 436804DEST_PATH_IMAGE009
representing the electromagnetic observation data in a form of a plurality of electromagnetic observations,
Figure 844782DEST_PATH_IMAGE010
，
Figure 981366DEST_PATH_IMAGE009
in which comprises
Figure 584998DEST_PATH_IMAGE011
The number of the pieces of data is set,
Figure 516045DEST_PATH_IMAGE012
representing an observation matrix;
based on the electromagnetic correlation data
Figure 208057DEST_PATH_IMAGE013
In a sparse matrix
Figure 882752DEST_PATH_IMAGE014
Projection onto
Figure 343821DEST_PATH_IMAGE015
Will solve for the electromagnetic correlation data
Figure 711348DEST_PATH_IMAGE016
Conversion to solve formula (2):
Figure 156236DEST_PATH_IMAGE017
（2）
wherein the content of the first and second substances,
Figure 369042DEST_PATH_IMAGE018
representing calculation matrices
Figure 783204DEST_PATH_IMAGE019
The norm of the number of the first-order-of-arrival,
Figure 56054DEST_PATH_IMAGE015
representing said electromagnetic related data
Figure 722658DEST_PATH_IMAGE016
In a sparse matrix
Figure 739156DEST_PATH_IMAGE020
The projection of the image onto the image plane is performed,
Figure 174817DEST_PATH_IMAGE021
-representing said perceptual matrix by means of a perceptual matrix,
Figure 884147DEST_PATH_IMAGE022
，
Figure 975730DEST_PATH_IMAGE020
representing the sparse matrix, characterized by equation (3):
Figure 795919DEST_PATH_IMAGE023
（3）
wherein the content of the first and second substances,
Figure 83156DEST_PATH_IMAGE024
；
solving equation (2) specifically includes, for each sub-region:
initializing each parameter: residual error
Figure 963388DEST_PATH_IMAGE025
Index matrix
Figure 339005DEST_PATH_IMAGE026
Number of elements
Figure 697305DEST_PATH_IMAGE027
，
Figure 904296DEST_PATH_IMAGE028
Representing step size, the size of which is the number of sub-regions and the number of iterations
Figure 955428DEST_PATH_IMAGE029
，
Figure 21605DEST_PATH_IMAGE030
Representing the perception matrix
Figure 183596DEST_PATH_IMAGE031
To (1) a
Figure 179846DEST_PATH_IMAGE032
Column, phase index
Figure 401880DEST_PATH_IMAGE033
Said projection
Figure 548827DEST_PATH_IMAGE015
Is estimated value of
Figure 514509DEST_PATH_IMAGE034
；
Calculating out
Figure 633775DEST_PATH_IMAGE035
Before selecting from the calculated values of each element
Figure 761131DEST_PATH_IMAGE036
A value of an element, and comparing the value of the element
Figure 598637DEST_PATH_IMAGE036
Value of each element in the perception matrix
Figure 305693DEST_PATH_IMAGE031
The set of corresponding column sequence numbers in
Figure 542115DEST_PATH_IMAGE037
；
Computing
Figure 637110DEST_PATH_IMAGE038
，
Figure 961912DEST_PATH_IMAGE039
；
Computing a least squares solution
Figure 472659DEST_PATH_IMAGE040
；
According to the least square solution
Figure 300938DEST_PATH_IMAGE041
The absolute value of each element item is selected from the values
Figure 35675DEST_PATH_IMAGE042
Individual element terms and extracting the
Figure 847774DEST_PATH_IMAGE042
Element item in the perception matrix
Figure 424861DEST_PATH_IMAGE031
Of (1) corresponding column composition
Figure 576488DEST_PATH_IMAGE043
Corresponding column number constitution
Figure 747706DEST_PATH_IMAGE044
；
A new residual value is calculated and,
Figure 781521DEST_PATH_IMAGE045
，
Figure 634071DEST_PATH_IMAGE046
；
judgment of
Figure 436942DEST_PATH_IMAGE047
Whether the result is true;
if so, then
Figure 513482DEST_PATH_IMAGE048
Update, update
Figure 297243DEST_PATH_IMAGE049
Will be
Figure 484642DEST_PATH_IMAGE049
In the corresponding position of
Figure 407598DEST_PATH_IMAGE050
Is set to
Figure 858302DEST_PATH_IMAGE051
；
If not, further judgment is made
Figure 866710DEST_PATH_IMAGE052
Whether the result is true; if so, then
Figure 857799DEST_PATH_IMAGE053
，
Figure 900842DEST_PATH_IMAGE054
，
Figure 319185DEST_PATH_IMAGE055
And recalculating the new residual valuesr tc (ii) a If not, then
Figure 749641DEST_PATH_IMAGE056
，
Figure 544422DEST_PATH_IMAGE057
，
Figure 176392DEST_PATH_IMAGE058
And recalculating the new residual valuesr tc ；
Thereby solving the electromagnetic correlation data in the sub-area
Figure 765636DEST_PATH_IMAGE059
Fusing electromagnetic correlation data of the respective sub-regions
Figure 14215DEST_PATH_IMAGE060
，
Figure 347107DEST_PATH_IMAGE001
The electromagnetic related data of each sub-area after fusion is passed through a median filter to obtain the electromagnetic related data in the electromagnetic map in the target area
Figure 36846DEST_PATH_IMAGE061
Wherein, in the process,
Figure 325220DEST_PATH_IMAGE062
the median filtering process is indicated.
3. The method for constructing the electromagnetic map based on the weight distribution as claimed in claim 2, wherein in the step S2:
the geographical sampling data of any sampling point position is represented by the reference signal received power of the sampling point position:
Figure 998778DEST_PATH_IMAGE063
（4）
wherein, the first and the second end of the pipe are connected with each other,
Figure 135361DEST_PATH_IMAGE064
is representative of the reference signal received power,
Figure 679606DEST_PATH_IMAGE065
represents the geographical location of any of the sample points,
Figure 876232DEST_PATH_IMAGE066
a function representing a prediction model of the target,
Figure 568245DEST_PATH_IMAGE067
，
Figure 242940DEST_PATH_IMAGE068
a parameter representing the prediction model is determined,
Figure 724516DEST_PATH_IMAGE069
the number of functions of the prediction model and the number of parameters of the prediction model are both N,
Figure 92043DEST_PATH_IMAGE070
representing a random process;
then the equation (5) holds:
Figure 802510DEST_PATH_IMAGE071
（5）
wherein the content of the first and second substances,
Figure 280896DEST_PATH_IMAGE072
representing the random process
Figure 330892DEST_PATH_IMAGE070
In the expectation that the position of the target is not changed,
Figure 807004DEST_PATH_IMAGE073
indicating the geographical location of the two sample points,
Figure 473608DEST_PATH_IMAGE074
a correlation model is represented that is representative of,
Figure 752755DEST_PATH_IMAGE075
a parameter representative of the correlation model is determined,
Figure 188416DEST_PATH_IMAGE076
representing the random process
Figure 835429DEST_PATH_IMAGE070
The variance of (a);
let the sample point of the geographic sampling data be
Figure 989330DEST_PATH_IMAGE077
The corresponding sample value is
Figure 809518DEST_PATH_IMAGE078
To give formula (6):
Figure 99685DEST_PATH_IMAGE079
（6）
wherein the content of the first and second substances,
Figure 245496DEST_PATH_IMAGE080
to represent
Figure 621114DEST_PATH_IMAGE068
The predicted value of (a) is determined,
Figure 976484DEST_PATH_IMAGE081
represent
Figure 589999DEST_PATH_IMAGE082
The predicted value of (a) is determined,
Figure 844394DEST_PATH_IMAGE083
is composed of
Figure 176149DEST_PATH_IMAGE084
The matrix of the composition is composed of a plurality of matrixes,
Figure 72561DEST_PATH_IMAGE085
，
Figure 803232DEST_PATH_IMAGE086
to represent
Figure 228529DEST_PATH_IMAGE087
A correlation matrix of type sample points, the constituent elements of which are
Figure 47580DEST_PATH_IMAGE088
；
Defining a correlation function:
Figure 216524DEST_PATH_IMAGE089
（7）
wherein the content of the first and second substances,
Figure 335790DEST_PATH_IMAGE090
is shown and
Figure 994305DEST_PATH_IMAGE065
the geographical location of a different one of the other sampling points,
Figure 32143DEST_PATH_IMAGE091
is the dimension of the sample point; and has the following formula (8):
Figure 801516DEST_PATH_IMAGE092
（8）
wherein the content of the first and second substances,
Figure 775288DEST_PATH_IMAGE093
representation solution
Figure 542387DEST_PATH_IMAGE094
The determinant (b) using the spherical model includes:
Figure 867189DEST_PATH_IMAGE096
（9）
wherein the content of the first and second substances,
Figure 174674DEST_PATH_IMAGE097
then, there is formula (10):
Figure 471794DEST_PATH_IMAGE098
（10）
wherein the content of the first and second substances,
Figure 672444DEST_PATH_IMAGE099
to represent
Figure 484542DEST_PATH_IMAGE100
The predicted value of (2);
substituting the non-sampled position into formula (10) to obtain the geographic related data of the non-sampled position, and finally obtaining the geographic related data of the electromagnetic map
Figure 64559DEST_PATH_IMAGE101
。
4. The method for constructing an electromagnetic map based on weight distribution as claimed in claim 3, wherein in said step S3:
the weight distribution result is:
Figure 481765DEST_PATH_IMAGE102
（11）
constructing the fused complete electromagnetic map is characterized by equation (12):
Figure 121825DEST_PATH_IMAGE103
（12）
wherein, the first and the second end of the pipe are connected with each other,min J( )it is shown that the minimum value is found,
Figure 421219DEST_PATH_IMAGE104
a 2-norm representing the computation matrix,
Figure 5260DEST_PATH_IMAGE105
presentation pair
Figure 339289DEST_PATH_IMAGE106
The constraint of the whole variable is carried out,
Figure 619092DEST_PATH_IMAGE107
a parameter of the constraint is represented by,
Figure 343465DEST_PATH_IMAGE106
representing complete electromagnetic data obtained by fusing the electromagnetic related data and the geographic related data, wherein the electromagnetic map is obtained by visualizing the complete electromagnetic data;
solving equation (12) specifically includes:
initializing intermediate parameters of the solution process
Figure 530864DEST_PATH_IMAGE108
，
Figure 657083DEST_PATH_IMAGE109
；
The loop is iterated through the loop,
Figure 839278DEST_PATH_IMAGE110
denotes the first
Figure 113265DEST_PATH_IMAGE111
The number of iterations is then repeated,
Figure 776459DEST_PATH_IMAGE112
representing the total number of iterations, then:
Figure 491605DEST_PATH_IMAGE114
（13）
Figure 133718DEST_PATH_IMAGE115
（14）
Figure 567104DEST_PATH_IMAGE117
（15）
wherein the content of the first and second substances,
Figure 627464DEST_PATH_IMAGE118
denotes the first
Figure 790592DEST_PATH_IMAGE111
Intermediate parameters after sub-iteration
Figure 51941DEST_PATH_IMAGE119
The value of (a) is set to (b),
Figure 831678DEST_PATH_IMAGE120
are all intermediate parameters, and
Figure 367832DEST_PATH_IMAGE121
，
Figure 385467DEST_PATH_IMAGE122
obtained after the last iteration, and then the final iteration is carried out,
Figure 673841DEST_PATH_IMAGE123
representing pairs of numbers, including
Figure 81820DEST_PATH_IMAGE124
，
Figure 421666DEST_PATH_IMAGE119
Has the dimension of
Figure 293807DEST_PATH_IMAGE125
，
Figure 693695DEST_PATH_IMAGE126
Has the dimension of
Figure 385708DEST_PATH_IMAGE127
And equation (16) holds:
Figure 260735DEST_PATH_IMAGE128
（16）
wherein, in formula (13)
Figure 128328DEST_PATH_IMAGE129
Operation satisfies
Figure 167960DEST_PATH_IMAGE130
，
Figure 19372DEST_PATH_IMAGE131
，
Figure 166932DEST_PATH_IMAGE132
，
Figure 154611DEST_PATH_IMAGE133
And is provided with
Figure 896302DEST_PATH_IMAGE134
，
Figure 766169DEST_PATH_IMAGE135
，
Figure 248578DEST_PATH_IMAGE136
，
Figure 621922DEST_PATH_IMAGE137
，
Figure 3356DEST_PATH_IMAGE138
Is shown in
Figure 422836DEST_PATH_IMAGE123
The quadrature operation is performed.
5. A system for constructing an electromagnetic map based on weight assignment, wherein the electromagnetic map comprises electromagnetic-related data and geographic-related data; the system comprises:
a first processing unit configured to: deploying a plurality of sensing nodes in a target area to acquire a sensing matrix, acquiring electromagnetic observation data of the target area through sampling, and constructing the electromagnetic relevant data of the electromagnetic map based on the sensing matrix and the electromagnetic observation data;
a second processing unit configured to: acquiring geographic sampling data of the target area through sampling based on the deployed sensing nodes, and predicting geographic related data of non-sampling positions according to geographic correlation among the geographic sampling data so as to obtain the geographic related data of the electromagnetic map;
a third processing unit configured to: and distributing weights to the electromagnetic relevant data of the electromagnetic map and the geographic relevant data of the electromagnetic map, and constructing a complete electromagnetic map of the target area by fusing the electromagnetic relevant data and the geographic relevant data distributed with the weights.
6. The system for constructing an electromagnetic map based on weight assignment as claimed in claim 5, wherein said first processing unit is specifically configured to:
deploying a plurality of sensing nodes in the target area to acquire a sensing matrix, specifically comprising: equally dividing the target area into a plurality of sub-areas and numbering, wherein the number is
Figure 177778DEST_PATH_IMAGE139
Deploying the same number of sensing nodes in each sub-area in a random deployment mode;
the electromagnetic correlation data is characterized by equation (1):
Figure 140049DEST_PATH_IMAGE140
（1）
wherein the content of the first and second substances,
Figure 161225DEST_PATH_IMAGE005
representing said electromagnetic correlation data to be constructed,
Figure 943368DEST_PATH_IMAGE141
，
Figure 502001DEST_PATH_IMAGE008
representing the total amount of electromagnetic data within the target region,
Figure 849936DEST_PATH_IMAGE009
representing the electromagnetic observation data as a function of time,
Figure 838752DEST_PATH_IMAGE010
，
Figure 639349DEST_PATH_IMAGE009
in which comprises
Figure 939356DEST_PATH_IMAGE011
The number of the pieces of data is set,
Figure 672957DEST_PATH_IMAGE012
representing an observation matrix;
based on the electromagnetic correlation data
Figure 160570DEST_PATH_IMAGE013
In a sparse matrix
Figure 448463DEST_PATH_IMAGE014
Projection onto
Figure 148566DEST_PATH_IMAGE015
Will solve for the electromagnetic correlation data
Figure 757181DEST_PATH_IMAGE016
Conversion to solve formula (2):
Figure 353378DEST_PATH_IMAGE017
（2）
wherein the content of the first and second substances,
Figure 456463DEST_PATH_IMAGE142
representing a computational matrix
Figure 632361DEST_PATH_IMAGE019
The norm of the number of the first-order-of-arrival,
Figure 871712DEST_PATH_IMAGE015
representing said electromagnetic related data
Figure 839144DEST_PATH_IMAGE016
In a sparse matrix
Figure 836050DEST_PATH_IMAGE020
The projection of the image onto the image plane is performed,
Figure 81218DEST_PATH_IMAGE021
-representing said perceptual matrix by means of a perceptual matrix,
Figure 388790DEST_PATH_IMAGE022
，
Figure 733315DEST_PATH_IMAGE020
representing the sparse matrix, characterized by equation (3):
Figure 14255DEST_PATH_IMAGE143
（3）
wherein the content of the first and second substances,
Figure 794604DEST_PATH_IMAGE024
；
solving equation (2) specifically includes, for each sub-region:
initializing each parameter: residual error
Figure 946231DEST_PATH_IMAGE025
Indexing matrix
Figure 320712DEST_PATH_IMAGE026
Number of elements
Figure 292210DEST_PATH_IMAGE027
，
Figure 675918DEST_PATH_IMAGE028
Representing step size, the size of which is the number of the sub-regions and the iteration number
Figure 882384DEST_PATH_IMAGE029
，
Figure 427766DEST_PATH_IMAGE030
Representing the perception matrix
Figure 683298DEST_PATH_IMAGE031
To (1) a
Figure 401855DEST_PATH_IMAGE032
Column, phase index
Figure 528074DEST_PATH_IMAGE144
Said projection
Figure 444690DEST_PATH_IMAGE015
Is estimated value of
Figure 656359DEST_PATH_IMAGE034
；
Computing
Figure 850712DEST_PATH_IMAGE035
Before selecting from the calculated values of each element
Figure 424912DEST_PATH_IMAGE036
A value of an element, and comparing the value
Figure 46518DEST_PATH_IMAGE036
Value of each element in the perception matrix
Figure 765991DEST_PATH_IMAGE031
The set of corresponding column sequence numbers in
Figure 701717DEST_PATH_IMAGE037
；
Computing
Figure 271370DEST_PATH_IMAGE038
，
Figure 329456DEST_PATH_IMAGE039
；
Computing a least squares solution
Figure 312455DEST_PATH_IMAGE040
；
According to the least square solution
Figure 111260DEST_PATH_IMAGE145
The absolute value of each element item is selected from
Figure 535419DEST_PATH_IMAGE036
The element items are extracted
Figure 233248DEST_PATH_IMAGE036
Element item in the perceptionMatrix array
Figure 703543DEST_PATH_IMAGE031
Of (1) corresponding column composition
Figure 43389DEST_PATH_IMAGE043
Corresponding column number constitution
Figure 584704DEST_PATH_IMAGE146
；
A new residual value is calculated and,
Figure 453434DEST_PATH_IMAGE045
，
Figure 411026DEST_PATH_IMAGE147
；
judgment of
Figure 23404DEST_PATH_IMAGE148
Whether the result is true or not;
if so, then
Figure 419226DEST_PATH_IMAGE048
Update
Figure 458857DEST_PATH_IMAGE149
Will be
Figure 903745DEST_PATH_IMAGE149
In the corresponding position of
Figure 523076DEST_PATH_IMAGE050
Is set to
Figure 776334DEST_PATH_IMAGE051
；
If not, further judgment is made
Figure 780675DEST_PATH_IMAGE052
Whether the result is true or not; if so, then
Figure 650542DEST_PATH_IMAGE053
，
Figure 198198DEST_PATH_IMAGE150
，
Figure 571541DEST_PATH_IMAGE151
And recalculating the new residual valuesr tc (ii) a If not, then
Figure 749713DEST_PATH_IMAGE152
，
Figure 369525DEST_PATH_IMAGE057
，
Figure 127397DEST_PATH_IMAGE055
And recalculating the new residual valuesr tc ；
Thereby solving the electromagnetic relevant data in the subarea
Figure 886406DEST_PATH_IMAGE153
Fusing electromagnetic correlation data of the respective sub-regions
Figure 907582DEST_PATH_IMAGE154
，
Figure 17621DEST_PATH_IMAGE001
The electromagnetic relevant data of each sub-area after fusion is passed through a median filter to obtain the electromagnetic relevant data in the electromagnetic map in the target area
Figure 310674DEST_PATH_IMAGE061
Wherein, in the step (A),
Figure 924190DEST_PATH_IMAGE062
the median filtering process is indicated.
7. The system for constructing an electromagnetic map based on weight assignment as claimed in claim 6, wherein said second processing unit is specifically configured to:
the geographical sampling data of any sampling point position is represented by the reference signal received power of the sampling point position:
Figure 444164DEST_PATH_IMAGE063
（4）
wherein the content of the first and second substances,
Figure 775919DEST_PATH_IMAGE064
is representative of the received power of the reference signal,
Figure 406752DEST_PATH_IMAGE065
represents the geographical location of any of the sample points,
Figure 934160DEST_PATH_IMAGE066
a function representing a prediction model of the model,
Figure 562719DEST_PATH_IMAGE067
，
Figure 647350DEST_PATH_IMAGE068
a parameter representing the prediction model is determined,
Figure 81873DEST_PATH_IMAGE069
the number of functions of the prediction model and the number of parameters of the prediction model are both N,
Figure 669980DEST_PATH_IMAGE070
representing a random process;
then the equation (5) holds:
Figure 755527DEST_PATH_IMAGE155
（5）
wherein the content of the first and second substances,
Figure 593033DEST_PATH_IMAGE072
representing the random process
Figure 565669DEST_PATH_IMAGE070
In the expectation of the above-mentioned method,
Figure 8282DEST_PATH_IMAGE156
indicating the geographical location of the two sample points,
Figure 40960DEST_PATH_IMAGE157
a correlation model is represented that is representative of,
Figure 365763DEST_PATH_IMAGE075
a parameter representing the correlation model is determined,
Figure 342421DEST_PATH_IMAGE158
representing the random process
Figure 905121DEST_PATH_IMAGE070
The variance of (a);
setting the sample point of the geographic sampling data as
Figure 905438DEST_PATH_IMAGE159
The corresponding sample value is
Figure 655219DEST_PATH_IMAGE078
To yield formula (6):
Figure 235236DEST_PATH_IMAGE079
（6）
wherein the content of the first and second substances,
Figure 649512DEST_PATH_IMAGE080
to represent
Figure 758414DEST_PATH_IMAGE068
The predicted value of (a) is determined,
Figure 261071DEST_PATH_IMAGE081
to represent
Figure 379199DEST_PATH_IMAGE076
The predicted value of (a) is obtained,
Figure 447649DEST_PATH_IMAGE083
is composed of
Figure 993031DEST_PATH_IMAGE084
The matrix of the composition is composed of a plurality of matrixes,
Figure 980054DEST_PATH_IMAGE085
，
Figure 636295DEST_PATH_IMAGE086
represent
Figure 293672DEST_PATH_IMAGE087
A correlation matrix of type sample points, the constituent elements of which are
Figure 806693DEST_PATH_IMAGE088
；
Defining a correlation function:
Figure 18363DEST_PATH_IMAGE089
（7）
wherein, the first and the second end of the pipe are connected with each other,
Figure 478294DEST_PATH_IMAGE090
is shown and
Figure 518407DEST_PATH_IMAGE065
the geographical location of a different one of the other sampling points,
Figure 140012DEST_PATH_IMAGE091
is the dimension of the sample point; and has the following formula (8):
Figure 635716DEST_PATH_IMAGE092
（8）
wherein the content of the first and second substances,
Figure 430496DEST_PATH_IMAGE093
representation solution
Figure 62466DEST_PATH_IMAGE160
The determinant of (1) adopts a spherical model, and comprises the following components:
Figure 386131DEST_PATH_IMAGE162
（9）
wherein the content of the first and second substances,
Figure 103551DEST_PATH_IMAGE097
then, there is formula (10):
Figure DEST_PATH_IMAGE163
（10）
wherein the content of the first and second substances,
Figure 43301DEST_PATH_IMAGE164
to represent
Figure DEST_PATH_IMAGE165
The predicted value of (2);
substituting the non-sampled position into formula (10) to obtain the geographic related data of the non-sampled position, and finally obtaining the geographic related data of the electromagnetic map
Figure 936302DEST_PATH_IMAGE101
。
8. The system for building an electromagnetic map based on weight assignment according to claim 7, wherein the third processing unit is specifically configured to:
the weight distribution result is:
Figure 696447DEST_PATH_IMAGE166
（11）
constructing the fused complete electromagnetic map is characterized by equation (12):
Figure 101496DEST_PATH_IMAGE103
（12）
wherein, the first and the second end of the pipe are connected with each other,min J( )it is shown that the minimum value is found,
Figure 503659DEST_PATH_IMAGE104
represents the 2-norm of the computational matrix,
Figure 313483DEST_PATH_IMAGE105
presentation pair
Figure 510109DEST_PATH_IMAGE106
The constraint of the whole variable is carried out,
Figure 405384DEST_PATH_IMAGE107
the parameters of the constraint are represented by a representation,
Figure 345658DEST_PATH_IMAGE106
representing fusion of the electromagnetic-related data andthe complete electromagnetic data is obtained after the geographic relevant data is described, and the electromagnetic map is obtained after the complete electromagnetic data is visualized;
solving equation (12) specifically includes:
initializing intermediate parameters of the solution process
Figure 9989DEST_PATH_IMAGE108
，
Figure 843428DEST_PATH_IMAGE109
；
The loop is iterated through the loop,
Figure 491578DEST_PATH_IMAGE110
denotes the first
Figure 173226DEST_PATH_IMAGE111
The number of sub-iterations is,
Figure 19960DEST_PATH_IMAGE112
representing the total number of iterations, then:
Figure DEST_PATH_IMAGE167
（13）
Figure 699334DEST_PATH_IMAGE168
（14）
Figure 855288DEST_PATH_IMAGE116
（15）
wherein the content of the first and second substances,
Figure 340627DEST_PATH_IMAGE118
is shown as
Figure 979550DEST_PATH_IMAGE111
Sub-iterationIntermediate parameter of the last
Figure 157721DEST_PATH_IMAGE119
The value of (a) is,
Figure 514885DEST_PATH_IMAGE120
are all intermediate parameters, and
Figure 597723DEST_PATH_IMAGE121
，
Figure 356731DEST_PATH_IMAGE122
obtained after the last iteration of the process,
Figure 440225DEST_PATH_IMAGE123
representing pairs of numbers, including
Figure 81422DEST_PATH_IMAGE124
，
Figure 908563DEST_PATH_IMAGE119
Has a dimension of
Figure 787658DEST_PATH_IMAGE125
，
Figure 39123DEST_PATH_IMAGE126
Has the dimension of
Figure 370878DEST_PATH_IMAGE127
And equation (16) holds:
Figure DEST_PATH_IMAGE169
（16）
wherein, in formula (13)
Figure 204973DEST_PATH_IMAGE129
Operation satisfaction
Figure 407416DEST_PATH_IMAGE130
，
Figure 895029DEST_PATH_IMAGE131
，
Figure 445571DEST_PATH_IMAGE132
，
Figure 676832DEST_PATH_IMAGE133
And is provided with
Figure 999361DEST_PATH_IMAGE134
，
Figure 126716DEST_PATH_IMAGE135
，
Figure 229802DEST_PATH_IMAGE136
，
Figure 202437DEST_PATH_IMAGE137
，
Figure 376542DEST_PATH_IMAGE138
Is shown in
Figure 878061DEST_PATH_IMAGE123
The quadrature operation is performed.
9. An electronic device, characterized in that the electronic device comprises a memory and a processor, the memory stores a computer program, and the processor, when executing the computer program, implements the steps in a method for constructing an electromagnetic map based on weight assignment as claimed in any one of claims 1-4.
10. A computer-readable storage medium, having stored thereon a computer program which, when being executed by a processor, carries out the steps of a method of constructing an electromagnetic map based on weight assignment as claimed in any one of claims 1 to 4.
CN202210785057.1A 2022-07-06 2022-07-06 Method and system for constructing electromagnetic map based on weight distribution Active CN114895111B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210785057.1A CN114895111B (en) 2022-07-06 2022-07-06 Method and system for constructing electromagnetic map based on weight distribution

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210785057.1A CN114895111B (en) 2022-07-06 2022-07-06 Method and system for constructing electromagnetic map based on weight distribution

Publications (2)

Publication Number Publication Date
CN114895111A CN114895111A (en) 2022-08-12
CN114895111B true CN114895111B (en) 2022-10-21

Family

ID=82730412

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210785057.1A Active CN114895111B (en) 2022-07-06 2022-07-06 Method and system for constructing electromagnetic map based on weight distribution

Country Status (1)

Country Link
CN (1) CN114895111B (en)

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103389517A (en) * 2013-07-30 2013-11-13 高建东 High-accuracy geomagnetic vector measurement method and device
CN104714257A (en) * 2015-01-29 2015-06-17 哈尔滨工程大学 Local geomagnetic map construction method based on multifractal kriging interpolation of gradual interpolation correction
CN110260855B (en) * 2019-06-24 2021-06-29 北京壹氢科技有限公司 Indoor pedestrian navigation positioning method integrating pedestrian dead reckoning, geomagnetic information and indoor map information
WO2021257725A1 (en) * 2020-06-17 2021-12-23 Astra Navigation, Inc. Correlating overlapping magnetic measurement data from multiple magnetic navigation devices and updating a geomagnetic map with that data
CN112887903B (en) * 2021-01-26 2022-12-09 东南大学 Communication positioning perception integrated method based on soft information fusion
CN113376597A (en) * 2021-06-18 2021-09-10 西安电子科技大学 Complex terrain electromagnetic scattering rapid simulation method based on digital elevation map and GPU
CN113466571B (en) * 2021-09-06 2022-02-22 中国人民解放军国防科技大学 Method and system for constructing electromagnetic map

Also Published As

Publication number Publication date
CN114895111A (en) 2022-08-12

Similar Documents

Publication Publication Date Title
CN108271165B (en) Method and system for predicting coverage state of base station network signal
Han et al. Radio environment map construction by kriging algorithm based on mobile crowd sensing
CN104618045B (en) Collected data-based wireless channel transmission model establishing method and system
CN108271117A (en) A kind of LTE network coverage evaluating method and device
KR102553327B1 (en) Method of image proccessing, Computer readable storage medium of recording the method and image processing apparatus
CN108271171B (en) Building network condition statistical method and device
CN113466571B (en) Method and system for constructing electromagnetic map
CN108966120A (en) A kind of three side localization method of combination and system for dynamic cluster network improvement
Fernandes et al. Comparison of artificial intelligence and semi-empirical methodologies for estimation of coverage in mobile networks
US10021668B2 (en) Wireless characteristic display apparatus, wireless characteristic display method, and computer readable medium
Raimbault et al. Generating urban morphologies at large scales
CN107438264A (en) Estate performance analysis method and device
Stusek et al. LPWAN coverage assessment planning without explicit knowledge of base station locations
CN114895111B (en) Method and system for constructing electromagnetic map based on weight distribution
CN111083724A (en) OCWG-TOPSIS-based user association method in fog computing network
CN105357707B (en) Indoor distributed mobile communication signal coverage prediction method based on Kriging regression algorithm
Lovén et al. Towards EDISON: An edge-native approach to distributed interpolation of environmental data
Zhao et al. GSMAC: GAN-based signal map construction with active crowdsourcing
Xia et al. Real time localization algorithm based on local linear embedding optimization in sensor networks
Maiti et al. Three dimensional measuring points locating algorithm based texture-patched matrix completion for indoor 3D REM design
Zheng et al. Measurement method of distributed nodes in wireless sensor networks based on multiple attributes
Huang et al. TV band radio environment mapping in Beijing
Saikia et al. Wireless sensor node deployment strategy for hilly terrains–a surface approximation based approach
Boubrima et al. Leveraging the potential of wsn for an efficient correction of air pollution fine-grained simulations
CN115022961B (en) Positioning method and equipment

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant