CN106488559A - A kind of outdoor positioning method based on visibility and server - Google Patents
A kind of outdoor positioning method based on visibility and server Download PDFInfo
- Publication number
- CN106488559A CN106488559A CN201611046249.1A CN201611046249A CN106488559A CN 106488559 A CN106488559 A CN 106488559A CN 201611046249 A CN201611046249 A CN 201611046249A CN 106488559 A CN106488559 A CN 106488559A
- Authority
- CN
- China
- Prior art keywords
- data
- training
- visibility
- signal strength
- neural network
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W64/00—Locating users or terminals or network equipment for network management purposes, e.g. mobility management
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W84/00—Network topologies
- H04W84/02—Hierarchically pre-organised networks, e.g. paging networks, cellular networks, WLAN [Wireless Local Area Network] or WLL [Wireless Local Loop]
- H04W84/10—Small scale networks; Flat hierarchical networks
- H04W84/12—WLAN [Wireless Local Area Networks]
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Position Fixing By Use Of Radio Waves (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
The invention discloses a kind of outdoor positioning method based on visibility, methods described includes step:In S100, acquisition testing region, each WAP receives signal strength data and the visibility data of the signal sent out by client to be detected;S200, respectively by the signal strength data and the visibility data input training after location model input data layer;S300, the Internet based on the location model after training calculate the signal strength data and the visibility data, and determine the position of client to be detected according to the output result of output layer.Location model in the present invention is trained to deep neural network by a large amount of training sample data using the deep neural network after training, lifts Position location accuracy and precision.
Description
Technical field
The present invention relates to wireless local area network technology field, more particularly to a kind of outdoor positioning method based on visibility and clothes
Business device.
Background technology
Location technology worldwide mainly has GPS location, Wi-Fi positioning, bluetooth positioning etc., GPS location at present
Outdoor is mainly used in, Wi-Fi, bluetooth positioning can be not only used for interior, it can also be used to outdoor.As Wi-Fi positions relative maturity,
The particular content of the present invention is introduced with Wi-Fi location technology as background below.With the popularization of wireless router, current big portion
Point public domain all has been carried out more than ten or even tens WiFi signal are covered, and these routers are being propagated to surrounding
While WiFi signal, its physical address and the information such as signal strength signal intensity is also ceaselessly sent, as long as in its signal cover,
Even if not knowing the password of Wi-Fi, these information can be similarly obtained.
General WiFi indoor positioning technologies are the WLAN (WLAN) based on IEEE802.11b/g agreement mostly
Signal strength signal intensity location technology.It is that letter is calculated according to the intensity of the signal for receiving based on the location technology general principle of signal strength signal intensity
The distance between number receiver and signal source, are largely divided into two classes:Triangle intensity algorithm and location fingerprint recognizer.Its
Intermediate cam shape intensity arithmetic accuracy is low, it is difficult to meet indoor positioning requirement;And there is receiving device in general fingerprint recognizer
Different and cause to receive the defect that signal has error.
Content of the invention
For solving above-mentioned technical problem, the present invention provides a kind of outdoor positioning method based on visibility and server, leads to
Cross and the corresponding signal strength data of each WAP and visibility data are gathered, realize based on deep neural network
WiFi outdoor positioning.
The technical scheme that the present invention is provided is as follows:
The invention discloses a kind of outdoor positioning method based on visibility, methods described includes step:S100, collection inspection
Survey signal strength data and visibility data that each WAP in region receives the signal sent out by client to be detected;
S200, respectively by the signal strength data and the visibility data input training after location model input data
Layer;S300, the Internet based on the location model after training calculate the signal strength data and the visibility data, and
Output result according to output layer determines the position of client to be detected.
It is further preferred that the visibility data are gathered in different presetting under visibility state, each is described default
Visibility state corresponds to the visibility data.
It is further preferred that also including step before step S100:S000, training in advance deep neural network, will
Deep neural network after training is used as the location model.
It is further preferred that step S000 further includes step:S001, pre-set training location tags;
Under the different default visibility states of S002, respectively collection, each WAP receives training terminal and marks in each training position
Sign the signal strength data of signal sent out on correspondence position in detection zone and corresponding with current preset visibility state
The visibility data;According to each WAP receive training terminal each described training location tags corresponding to
Position on the signal strength data that signaled and the corresponding visibility data genaration training sample data, by all institutes
State training sample data and training dataset is generated, and send in deep neural network;S003, by the input number of deep neural network
Two channel data layers are defined as according to layer, the node of the two channel datas layer is corresponding with each WAP, according to two-way
The node of track data layer mode corresponding with WAP respectively by each described training sample data each with wirelessly connect
The corresponding signal strength data of access point combines two passages of corresponding visibility data input corresponding node, through the depth
The neutral net output training result corresponding with training location tags described in the training sample data;S004, successively general
The training location tags that the training result of output is corresponding are compared, and deep neural network are entered according to comparative result
Row training, using the deep neural network after training as the location model.
It is further preferred that also including step between step S002 and step S003:S025, respectively to all
Signal strength data and visibility data in the training sample data is normalized;Step S100 and institute
Stating between step S200 also includes step:S150, each WAP to collection receive the letter sent out by client to be detected
Number signal strength data and visibility data be normalized.
The invention also discloses a kind of outdoor positioning server based on visibility, including:Data acquisition module, for adopting
In collection detection zone, each WAP receives signal strength data and the visibility number of the signal sent out by client to be detected
According to;Locating module, for the positioning after training the signal strength data for collecting and the visibility data input
The input data layer of model, the Internet based on location model calculate the signal strength data and the visibility data,
And the position of client to be detected is determined according to the output result of output layer.
It is further preferred that the visibility data are gathered in different presetting under visibility state, each is described default
Visibility state corresponds to the visibility data.
It is further preferred that also including:Training module, for training in advance deep neural network, by the depth after training
Neutral net is used as the location model.
It is further preferred that the training module is further included:Submodule preset by label, for pre-setting for instructing
Experienced training location tags;Training dataset generates submodule, for gathering each nothing under different default visibility states respectively
Line access point receives the letter of the signal sent out on correspondence position in detection zone by training terminal in each training location tags
Number intensity data and the visibility data corresponding with current preset visibility state;Received according to each WAP
To training terminal on the position corresponding to each described training location tags the signal strength data that signaled and corresponding
The visibility data genaration training sample data, all training sample data are generated training dataset, and send into depth
In degree neutral net;Input data layer defines submodule, for the input data layer of deep neural network is defined as two passages
Data Layer, the node of the two channel datas layer are corresponding with each WAP;Training prediction submodule, for according to two
The node of channel data layer mode corresponding with WAP respectively by each described training sample data each with wireless
The corresponding signal strength data of access point combines two passages of corresponding visibility data input corresponding node, through the depth
The degree neutral net output training result corresponding with training location tags described in the training sample data, successively will output
The corresponding training location tags of training result be compared, deep neural network is instructed according to comparative result
Practice, using the deep neural network after training as the location model.
It is further preferred that also including:Data processing module, for respectively to signal in all training sample data
Intensity data and visibility data are normalized, and to be detected for receiving to each WAP of collection
The signal strength data of the signal sent out by client and visibility data are normalized.
Compared with prior art, the present invention is provided a kind of outdoor positioning method based on visibility and server, pass through
Collect the signal strength data of client position to be measured and comprising visibility data, the location model for training be input into,
To be measured client position can determine that, using visibility data binding signal intensity data as the defeated of deep neural network
Enter, improve the precision of outdoor positioning under different weather state.
Description of the drawings
Below by the way of clearly understandable, preferred embodiment be described with reference to the drawings, the present invention is given furtherly
Bright.
Fig. 1 is a kind of key step schematic diagram of the outdoor positioning method based on visibility of the present invention;
The step of Fig. 2 is a kind of training deep neural network based on the outdoor positioning method of visibility of the present invention is illustrated
Figure;
Fig. 3 is a kind of main composition schematic diagram of the outdoor positioning server based on visibility of the present invention;
Fig. 4 is fully composed schematic diagram for a kind of outdoor positioning server based on visibility of the present invention.
Reference:
100th, data acquisition module, 200, locating module, 300, training module, 311, label preset submodule, 312, instruction
Practice data set generation submodule, 313, input data layer define submodule, 314, training prediction submodule, 400, data processing mould
Block.
Specific embodiment
In order to be illustrated more clearly that the embodiment of the present invention or technical scheme of the prior art, below by control description of the drawings
The specific embodiment of the present invention.It should be evident that drawings in the following description are only some embodiments of the present invention, for
For those of ordinary skill in the art, on the premise of not paying creative work, other can also be obtained according to these accompanying drawings
Accompanying drawing, and obtain other embodiments.
For making simplified form, each in figure only schematically show part related to the present invention, and they do not represent
Which is used as the practical structures of product.In addition, so that simplified form is readily appreciated, there is identical structure or function in some in figures
Part, only symbolically depicts one of those, or has only marked one of those.Herein, " one " not only represents
" only this ", it is also possible to represent the situation of " more than one ".
Fig. 1 is a kind of key step schematic diagram of the outdoor positioning method based on visibility of the present invention, as shown in figure 1, one
The outdoor positioning method based on visibility is planted, methods described includes step:Each WAP in S100, acquisition testing region
Receive signal strength data and the visibility data of the signal sent out by client to be detected;S200, respectively by the signal strength signal intensity
The input data layer of the location model after data and visibility data input training;S300, based on the positioning after training
The Internet of model calculates the signal strength data and the visibility data, and is determined according to the output result of output layer
The position of client to be detected.
Specifically, above-mentioned client to be detected (hereinafter referred to as STA) be with smart mobile phone, notebook computer or personal flat board
The intelligent terminals such as computer are carrier.
Wherein, each WAP receives the signal strength data of the signal sent out by client to be detected by with lower section
Formula is obtained:STA sends detection frame within a detection region in real time, and WAP obtains detection frame signal after receiving is strong
Degree, each WAP report signal strength signal intensity to home server or Cloud Server, and server is according to each WAP
The RSSI signal strength signal intensity for reporting generates signal strength data.For example, the form of signal strength data is<RSSI1, RSSI2,
RSSI3, RSSI4, RSSI5>, the RSSI of the STA that RSSI, the RSSI2 of the STA that wherein RSSI1 is received for AP1 are received for AP2, with
This analogizes.
Visibility data described in the present embodiment is gathered in different presetting under visibility states, and each is described default to see
Degree state corresponds to the visibility data.Wherein, it is visibility to be carried out drawing according to the situation of raining to preset visibility state
Point, concrete divided rank can sets itself, such as include visibility 0,1000 meters of visibility, 2000 meters of visibility etc., herein just
Concrete numerical value is adopted in explanation, but the present invention is not construed as limiting to presetting visibility state, can be set according to actually used situation.
Different default visibility states corresponds to different visibility data, and wherein visibility data can be seen according to default
Degree state is indicated, and specific method for expressing the present embodiment takes advantage of 100 as parameter using the inverse of visibility, such as, presets
Visibility state is 100 meters of visibility, then visibility data represent 1/100*100=1.0, and it is 10 public affairs to preset visibility state
In, then visibility data are expressed as 1/10000*100=0.01.
Signal strength data and the visibility number of the client position to be detected gathered in the present invention is described below
According to.Assume now with outdoor AP1, AP2, AP3.Three signal strength datas corresponding with AP, are represented with RSSI, are further added by one
Input:Visibility data input.Wherein visibility data are to carry out calculated parameter according to default visibility state.
Assume there is same client to be detected gathered data, STA1 table under the different default visibility states of same position
The collection of non-rainy day is shown as, STA2 is expressed as the rainy day, 1000 meters of visibility is gathered, is then exemplified below:
Table one
As shown in Table 1, gathered data under same position different weather state, it is clear that signal strength data has differences.
The form of present invention data of collection when outdoor positioning is carried out is shown in above-mentioned table one.
The existing outdoor positioning based on WiFi, the general signal strength data for obtaining STA by AP are used as positioning input
Data, but generally have a factor not to be taken into account:The visibility of air and rainy situation.Therefore, enter in outdoor environment
During row positioning, the signal strength data that STA is received under different default visibility states differs greatly, and the present invention is by visibility number
Enter according to the consideration, binding signal intensity data as input location model initial data, so as to increase outside WIFI room further
The positioning precision of positioning, improves locating accuracy.
It should be noted that when non-rainy, visibility is mainly what air pollution caused, and pollution itself is to radio
Transmission have no meaning, so in the non-rainy day, no matter how visibility is all set to 0 visibility parameter.In the rainy day, visibility is
Related with the size of rainfall, we take advantage of 100 as parameter using the inverse of visibility.
Preferably, also include step before step S100:S000, training in advance deep neural network, after training
Deep neural network as the location model.
Specifically, the output result difference for being obtained according to positioning in the present invention, can include two kinds of specific implementations, side
Formula one is used as location model by deep neural network and exports belonging to client position to be detected that certain pre-sets
The probable value of classification, mode two are to be used as location model by deep neural network directly to export client position to be detected
Preset position coordinates.The present invention is not construed as limiting to concrete training method.
The present invention is using the method for the global parameter training for having supervision:The known signal strength data for receiving and visibility
The physical location of data, causes the output of the Internet of deep neural network and real result by constantly adjustment network parameter
Identical.
The step of Fig. 2 is a kind of training deep neural network based on the outdoor positioning method of visibility of the present invention is illustrated
Figure.Preferably, as shown in Fig. 2 step S000 further includes step:S001, pre-set training location tags;
Under the different default visibility states of S002, respectively collection, each WAP receives training terminal and marks in each training position
Sign the signal strength data of signal sent out on correspondence position in detection zone and corresponding with current preset visibility state
The visibility data;According to each WAP receive training terminal each described training location tags corresponding to
Position on the signal strength data that signaled and the corresponding visibility data genaration training sample data, by all institutes
State training sample data and training dataset is generated, and send in deep neural network;S003, by the input number of deep neural network
Two channel data layers are defined as according to layer, the node of the two channel datas layer is corresponding with each WAP, according to two-way
The node of track data layer mode corresponding with WAP respectively by each described training sample data each with wirelessly connect
The corresponding signal strength data of access point combines two passages of corresponding visibility data input corresponding node, through the depth
The neutral net output training result corresponding with training location tags described in the training sample data;S004, successively general
The training location tags that the training result of output is corresponding are compared, and deep neural network are entered according to comparative result
Row training, using the deep neural network after training as the location model.
Specifically, because signal strength data and visibility data are multidimensional data in the present embodiment, therefore definition is used
The input data layer for making the deep neural network of location model is Three-channel data layer.
Exported belonging to client position to be detected as location model in mode one by deep neural network below
As a example by the probable value of certain classification for pre-setting, the process that the present invention is trained specifically is introduced to deep neural network.
1st, in training, gather each WAP first and training terminal is received in default training location tags in inspection
Survey signal strength data and the visibility data of the signal that is sent out on correspondence position in region.Training position is preset in the present embodiment
Label is put for self-defining, can specifically pass through stress and strain model detection zone, the grid that detection zone is divided into predetermined number divides
Class, by each corresponding default training location tags of grid classification distribution, also can be by setting up flat square seat in detection zone
Mark system, arranges corresponding position coordinates respectively in a coordinate system for default training location tags.Entered in mode one in the present embodiment
Row citing, explains the network number of each default training location tags correspondence position of collection in the present invention below with real data
According to.
(1) all original number that a default training location tags position is collected wherein are assumed in the non-rainy day
According to as follows:
<(- 32,0), (- 52,0), (- 60,0), 34>
Represent:
RSSI1=-32dBm, visibility 0 (non-rainy day)
RSSI2=-52dBm, visibility 0 (non-rainy day)
RSSI3=-60dBm, visibility 0 (non-rainy day)
Label=34, represent this default training location tags are designated 34, represent the net for being designated 34 in monitored area
Lattice position.
(2) assume what a default training location tags position wherein was collected in 1000 meters of rainy day visibility
All initial data are as follows:
<(- 20,0.1), (- 30,0.1), (- 40,0.1), 45>
Represent:
RSSI1=-20dBm, visibility 0.1 (rainy day, 1000 meters of visibility)
RSSI2=-30dBm, visibility 0.1 (rainy day, 1000 meters of visibility)
RSSI3=-40dBm, visibility 0.1 (rainy day, 1000 meters of visibility)
Label=45, represent this default training location tags are designated 45, represent the net for being designated 45 in monitored area
Lattice position.
2nd, the different data that presets under visibility state on each default training location tags are gathered successively, forms training sample
Notebook data.
3rd, training sample data training deep neural network is passed sequentially through.
As the initial data for gathering is 2-D data, therefore, deep neural network of the definition as location model is defeated
It is also two channel data layers to enter data Layer, and the node of two channel data layers is corresponding with each WAP.According to two passages
The node of data Layer mode corresponding with WAP is respectively by each and wireless access in each described training sample data
The corresponding signal strength data of point combines two passages of corresponding visibility data input corresponding node, default with said one
The initial data of collection is input into two as a example by all initial data that the non-rainy day is collected by training location tags position
Channel data layer, the data of each passage input are as shown in Table 2:
Table two
Passage 1 | Passage 2 |
-32 | 0 |
-52 | 0 |
-60 | 0 |
As shown in Table 2, in table two, passage 1 represents the signal strength data corresponding with each WAP of collection, single
Position is DB, and passage 2 represents presets corresponding visibility data under visibility state.Successively by each default training location tags pair
The training sample data that answers send into three passages of the Three-channel data layer of deep neural network, finally output training result and instruction
Practice the error of location tags, the Loss i.e. error for whole network being caused finally by the parameter in percentage regulation neutral net is most
Little.
It should be noted that do not indicate design parameter in entire depth neutral net, because these parameters and specific
The number of space and AP is relevant, not in the range of this patent.
Preferably, also include step between step S002 and step S003:S025, respectively to all instructions
Practice the signal strength data in sample data and visibility data are normalized;Step S100 and the step
Also include step between S200:S150, each WAP to collection receive the letter of the signal sent out by client to be detected
Number intensity data and visibility data are normalized.
As the unit of the initial data for gathering is differed, AP corresponds to RSSI, and visibility is parameter, so collection number
According to when deep neural network is input into, normalizing, (concrete method for normalizing is not especially limited, can be using institute in prior art
There is suitable normalization processing method).As deep neural network is in training, while having taken into account signal strength data and can see
Degrees of data, so the deep neural network for training can take into account the impact of visibility and weather when location prediction is carried out, makes
Must train more accurately, so that positioning is more accurate.
Fig. 3 is a kind of main composition schematic diagram of the outdoor positioning server based on visibility of the present invention, as shown in figure 3,
A kind of outdoor positioning server based on visibility, including:Data acquisition module 100, for each nothing in acquisition testing region
Line access point receives the signal strength data of the signal sent out by client to be detected and visibility data;Locating module 200, is used for
By the input data layer of the location model after the signal strength data for collecting and visibility data input training,
Internet based on location model calculates the signal strength data and the visibility data, and the output according to output layer
As a result determine the position of client to be detected.
Specifically, above-mentioned client to be detected (hereinafter referred to as STA) be with smart mobile phone, notebook computer or personal flat board
The intelligent terminals such as computer are carrier.
Wherein, each WAP receives the signal strength data of the signal sent out by client to be detected by with lower section
Formula is obtained:STA sends detection frame within a detection region in real time, and WAP obtains detection frame signal after receiving is strong
Degree, each WAP report signal strength signal intensity to home server or Cloud Server, and server is according to each WAP
The RSSI for reporting generates signal strength data.For example, the form of signal strength data is<RSSI1, RSSI2, RSSI3, RSSI4,
RSSI5>, the RSSI of the STA that RSSI, the RSSI2 of the STA that wherein RSSI1 is received for AP1 are received for AP2, by that analogy.
Visibility data described in the present embodiment is gathered in different presetting under visibility states, and each is described default to see
Degree state corresponds to the visibility data.Wherein, it is visibility to be carried out drawing according to the situation of raining to preset visibility state
Point, concrete divided rank can sets itself, such as include visibility 0,1000 meters of visibility, 2000 meters of visibility etc., herein just
Concrete numerical value is adopted in explanation, but the present invention is not construed as limiting to presetting visibility state, can be set according to actually used situation.
The existing outdoor positioning based on WiFi, the general signal strength data for obtaining STA by AP are used as positioning input
Data, but generally have a factor not to be taken into account:The visibility of air and rainy situation.Therefore, enter in outdoor environment
During row positioning, the signal strength data that STA is received under different default visibility states differs greatly, and the present invention is by visibility number
Enter according to the consideration, binding signal intensity data as input location model initial data, so as to increase outside WIFI room further
The positioning precision of positioning, improves locating accuracy.
Fig. 4 is fully composed schematic diagram for a kind of outdoor positioning server based on visibility of the present invention.As shown in figure 4,
Preferably, as shown in figure 4, also including:Training module 300, for training in advance deep neural network, by the depth god after training
Through network as the location model.
Specifically, the output result difference for being obtained according to positioning in the present invention, can include two kinds of specific implementations, side
Formula one is used as location model by deep neural network and exports belonging to client position to be detected that certain pre-sets
The probable value of classification, mode two are to be used as location model by deep neural network directly to export client position to be detected
Preset position coordinates.The present invention is not construed as limiting to concrete training method.
The present invention is using the method for the global parameter training for having supervision:Known signal corresponding with each WAP is strong
The physical location of degrees of data and visibility data belongs to certain default training location tags, is caused by constantly adjustment network parameter
The output of the Internet of deep neural network is identical with real result.
Preferably, the training module is further included:Submodule 311 preset by label, for pre-setting for training
Training location tags;Training dataset generates submodule 312, for gathering each nothing under different default visibility states respectively
Line access point receives the letter of the signal sent out on correspondence position in detection zone by training terminal in each training location tags
Number intensity data and the visibility data corresponding with current preset visibility state;Received according to each WAP
To training terminal on the position corresponding to each described training location tags the signal strength data that signaled and corresponding
The visibility data genaration training sample data, all training sample data are generated training dataset, and send into depth
In degree neutral net;Input data layer defines submodule 313, for the input data layer of deep neural network is defined as two-way
Track data layer, the node of the two channel datas layer are corresponding with each WAP;Training prediction submodule 314, is used for
According to the node of two channel data layers mode corresponding with WAP respectively by each described training sample data each
Signal strength data corresponding with WAP combines two passages of corresponding visibility data input corresponding node, passes through
The deep neural network output training result corresponding with training location tags described in the training sample data, successively
The training location tags corresponding for the training result of output are compared, according to comparative result to deep neural network
It is trained, using the deep neural network after training as the location model.
It should be noted that the training process for above-mentioned training module 300 refers to the inventive method part for training
The explanation of deep neural network, is no longer repeated herein.The contents such as information exchange, implementation procedure in book server between each module
Same design is based on said method embodiment, particular content can be found in the narration in the inventive method embodiment, herein no longer
Repeat.
Preferably, also include:Data processing module 400, for strong to signal in all training sample data respectively
Degrees of data and visibility data are normalized, and for receiving visitor to be detected to each WAP of collection
The signal strength data of the signal sent out by family end and visibility data are normalized.
As the unit of the initial data for gathering is differed, AP corresponds to RSSI, and visibility is parameter, so gathered data
When deep neural network is input into, should normalize that (concrete method for normalizing is not especially limited, and can own using in prior art
Suitable normalization processing method).As deep neural network is in training, while having taken into account signal strength data and visibility
Data, so the deep neural network for training can take into account the impact of visibility and weather when location prediction is carried out so that
Training is more accurate, so that positioning is more accurate.
The contents such as information exchange, implementation procedure in book server between each module and said method embodiment are based on same
Design, particular content can be found in the narration in the inventive method embodiment, and here is omitted.
It should be noted that above-described embodiment all can independent assortment as needed.The above is only the preferred of the present invention
Embodiment, it is noted that for those skilled in the art, in the premise without departing from the principle of the invention
Under, some improvements and modifications can also be made, these improvements and modifications also should be regarded as protection scope of the present invention.
Claims (10)
1. a kind of outdoor positioning method based on visibility, it is characterised in that methods described includes step:
In S100, acquisition testing region, each WAP receives the signal strength data of the signal sent out by client to be detected
With visibility data;
S200, respectively by the signal strength data and the visibility data input training after location model input number
According to layer;
S300, the Internet based on the location model after training calculate the signal strength data and the visibility data,
And the position of client to be detected is determined according to the output result of output layer.
2. the outdoor positioning method based on visibility as claimed in claim 1, it is characterised in that the visibility data be
Different presetting under visibility state gathers, and each described default visibility state corresponds to the visibility data.
3. the outdoor positioning method based on visibility as claimed in claim 2, it is characterised in that before step S100 also
Including step:S000, training in advance deep neural network, using the deep neural network after training as the location model.
4. the outdoor positioning method based on visibility as claimed in claim 3, it is characterised in that step S000 is further
Including step:
S001, pre-set training location tags;
Under the different default visibility states of S002, respectively collection, each WAP receives training terminal in each training position
Put the signal sent out on correspondence position in detection zone by label signal strength data and with current preset visibility state
The corresponding visibility data;Training terminal is received in each training location tags institute according to each WAP
The signal strength data signaled on corresponding position and the corresponding visibility data genaration training sample data, by institute
There is the number of training according to generation training dataset, and send in deep neural network;
S003, the input data layer of deep neural network is defined as two channel data layers, the node of the two channel datas layer
Corresponding with each WAP, according to the node of two channel data layers mode corresponding with WAP respectively by each
In the training sample data, each signal strength data corresponding with WAP combines corresponding visibility data input
Two passages of corresponding node, through deep neural network output and the mark of training position described in the training sample data
Sign corresponding training result;
S004, successively the training location tags corresponding for the training result of output are compared, according to comparative result
Deep neural network is trained, using the deep neural network after training as the location model.
5. the outdoor positioning method based on visibility as claimed in claim 4, it is characterised in that:
Also include step between step S002 and step S003:
S025, respectively place is normalized to the signal strength data in all training sample data and visibility data
Reason;
Also include step between step S100 and step S200:
S150, each WAP to collection receive signal strength data and the energy of the signal sent out by client to be detected
See that degrees of data is normalized.
6. a kind of outdoor positioning server based on visibility, it is characterised in that include:
Data acquisition module, receives, for each WAP in acquisition testing region, the signal sent out by client to be detected
Signal strength data and visibility data;
Locating module, for the positioning after training the signal strength data for collecting and the visibility data input
The input data layer of model, the Internet based on location model calculate the signal strength data and the visibility data,
And the position of client to be detected is determined according to the output result of output layer.
7. the outdoor positioning server based on visibility as claimed in claim 6, it is characterised in that the visibility data are
Gather in different presetting under visibility state, each described default visibility state corresponds to the visibility data.
8. the outdoor positioning server based on visibility as claimed in claim 7, it is characterised in that also include:
Training module, for training in advance deep neural network, using the deep neural network after training as the location model.
9. the outdoor positioning server based on visibility as claimed in claim 8, it is characterised in that the training module enters
Step includes:
Submodule preset by label, for pre-setting the training location tags for training;
Training dataset generates submodule, receives for gathering each WAP under different default visibility states respectively
The signal strength data of signal that training terminal is sent out on correspondence position in detection zone in each training location tags and
The visibility data corresponding with current preset visibility state;Training terminal is received every according to each WAP
The signal strength data signaled on position corresponding to the individual training location tags and the corresponding visibility data
Training sample data are generated, all training sample data is generated training dataset, and is sent in deep neural network;
Input data layer defines submodule, for the input data layer of deep neural network is defined as two channel data layers, institute
The node for stating two channel data layers is corresponding with each WAP;
Training prediction submodule, for node according to two channel data layers mode corresponding with WAP respectively by each
In the training sample data, each signal strength data corresponding with WAP combines corresponding visibility data input
Two passages of corresponding node, through deep neural network output and the mark of training position described in the training sample data
Corresponding training result is signed, successively the training location tags corresponding for the training result of output is compared, root
Deep neural network is trained according to comparative result, using the deep neural network after training as the location model.
10. the outdoor positioning server based on visibility as claimed in claim 9, it is characterised in that also include:
Data processing module, for respectively to the signal strength data in all training sample data and visibility data
It is normalized, and for receiving the signal of the signal sent out by client to be detected to each WAP of collection
Intensity data and visibility data are normalized.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611046249.1A CN106488559A (en) | 2016-11-22 | 2016-11-22 | A kind of outdoor positioning method based on visibility and server |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611046249.1A CN106488559A (en) | 2016-11-22 | 2016-11-22 | A kind of outdoor positioning method based on visibility and server |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106488559A true CN106488559A (en) | 2017-03-08 |
Family
ID=58275525
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201611046249.1A Pending CN106488559A (en) | 2016-11-22 | 2016-11-22 | A kind of outdoor positioning method based on visibility and server |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106488559A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107064913A (en) * | 2017-03-10 | 2017-08-18 | 上海斐讯数据通信技术有限公司 | A kind of wireless location method and system based on deep learning |
CN107886049A (en) * | 2017-10-16 | 2018-04-06 | 江苏省气象服务中心 | A kind of visibility identification method for early warning based on camera probe |
CN112533137A (en) * | 2020-11-26 | 2021-03-19 | 北京爱笔科技有限公司 | Device positioning method and device, electronic device and computer storage medium |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103945533A (en) * | 2014-05-15 | 2014-07-23 | 济南嘉科电子技术有限公司 | Big data based wireless real-time position positioning method |
CN104244113A (en) * | 2014-10-08 | 2014-12-24 | 中国科学院自动化研究所 | Method for generating video abstract on basis of deep learning technology |
-
2016
- 2016-11-22 CN CN201611046249.1A patent/CN106488559A/en active Pending
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103945533A (en) * | 2014-05-15 | 2014-07-23 | 济南嘉科电子技术有限公司 | Big data based wireless real-time position positioning method |
CN104244113A (en) * | 2014-10-08 | 2014-12-24 | 中国科学院自动化研究所 | Method for generating video abstract on basis of deep learning technology |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107064913A (en) * | 2017-03-10 | 2017-08-18 | 上海斐讯数据通信技术有限公司 | A kind of wireless location method and system based on deep learning |
CN107886049A (en) * | 2017-10-16 | 2018-04-06 | 江苏省气象服务中心 | A kind of visibility identification method for early warning based on camera probe |
CN112533137A (en) * | 2020-11-26 | 2021-03-19 | 北京爱笔科技有限公司 | Device positioning method and device, electronic device and computer storage medium |
CN112533137B (en) * | 2020-11-26 | 2023-10-17 | 北京爱笔科技有限公司 | Positioning method and device of equipment, electronic equipment and computer storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106507476A (en) | A kind of WiFi localization methods and server and location model construction method | |
CN106535134A (en) | Multi-room locating method based on wifi and server | |
CN106535326A (en) | WiFi locating method based on depth neural network and server | |
US10334456B2 (en) | Optimizing cellular networks using deep learning | |
CN110418354A (en) | It is a kind of that propagation model wireless network planning method is exempted from based on machine learning | |
CN106792507A (en) | A kind of WiFi localization methods and server based on network data | |
US11277755B2 (en) | Method and apparatus for analyzing communication environment based on property information of an object | |
CN106792769A (en) | A kind of WiFi localization methods and server and location model method for building up | |
CN106792553A (en) | A kind of many floor location methods and server based on wifi | |
CN106793067A (en) | A kind of many floor indoor orientation methods and server based on joint network | |
CN106131958A (en) | A kind of based on channel condition information with the indoor Passive Location of support vector machine | |
CN106792506A (en) | A kind of WiFi localization methods and server | |
US20180139623A1 (en) | Method and apparatus for analyzing communication environment based on property information of an object | |
CN106793070A (en) | A kind of WiFi localization methods and server based on reinforcement deep neural network | |
CN106488559A (en) | A kind of outdoor positioning method based on visibility and server | |
CN108271165A (en) | A kind of Forecasting Methodology and system of base station network signal covering state | |
CN103974369A (en) | Channel selection method and system based on WLAN wireless channel grading mechanism | |
CN106982436A (en) | The detection method of adjustment and device of a kind of indoor and outdoor network coverage | |
CN112243249A (en) | LTE new access anchor point cell parameter configuration method and device under 5G NSA networking | |
CN103458434B (en) | Method and device for determining antenna feeder parameters | |
CN109874104A (en) | User location localization method, device, equipment and medium | |
CN104039008B (en) | A kind of hybrid locating method | |
CN107736071A (en) | It is determined that the apparatus and method of the wave beam for guiding the signal that be sent to user equipment | |
CN107817466A (en) | Based on the indoor orientation method for stacking limited Boltzmann machine and random forests algorithm | |
CN107451605A (en) | A kind of simple target recognition methods based on channel condition information and SVMs |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170308 |