CN109151724A - Sighting distance based on channel impulse response Energy distribution/obstructed path recognition methods - Google Patents
Sighting distance based on channel impulse response Energy distribution/obstructed path recognition methods Download PDFInfo
- Publication number
- CN109151724A CN109151724A CN201810057363.7A CN201810057363A CN109151724A CN 109151724 A CN109151724 A CN 109151724A CN 201810057363 A CN201810057363 A CN 201810057363A CN 109151724 A CN109151724 A CN 109151724A
- Authority
- CN
- China
- Prior art keywords
- energy
- sighting distance
- path
- csi
- classifier
- 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.)
- Granted
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
- H04W4/02—Services making use of location information
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L25/00—Baseband systems
- H04L25/02—Details ; arrangements for supplying electrical power along data transmission lines
- H04L25/0202—Channel estimation
- H04L25/0212—Channel estimation of impulse response
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W64/00—Locating users or terminals or network equipment for network management purposes, e.g. mobility management
- H04W64/006—Locating users or terminals or network equipment for network management purposes, e.g. mobility management with additional information processing, e.g. for direction or speed determination
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Power Engineering (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
The invention discloses a kind of sighting distance based on channel impulse response Energy distribution/obstructed path recognition methods, without offline acquisition data, and manpower cost caused by data acquisition can be re-started to avoid the variation because of environment.In addition, this method applies also in dynamic environment, when signal transmission path is changed between sighting distance/non line of sight, the variation of transmission path can be also identified in real time.
Description
Technical field
The present invention relates to location-based service application field, especially a kind of sighting distance based on channel impulse response Energy distribution/
Obstructed path recognition methods.
Background technique
In recent years, the application based on location-based service (Location based Services, LBS) receives researcher's
Extensive concern.Location-based service can be applied to many fields, such as carry out Item Management to intelligent repository, carry out in megastore
Intelligent navigation in case of emergency provides the location information etc. of trapped person.With the rapid development of current mobile computing, people
Requirement to location-based service precision is also higher and higher.
In the application of location-based service, wireless indoor location technology becomes research as one of key technology therein
The focus of personnel.Currently, wireless indoor location technology is broadly divided into two classes: indoor positioning technologies based on signal strength and being based on
The indoor positioning technologies of channel state information.And channel state information becomes main study subject with the advantage of its stability.Shadow
The principal element for ringing indoor position accuracy is the influence of indoor complicated multipath, therefore how will from these complicated transmission paths
Sighting distance/obstructed path is separated, and is the effective ways for being currently able to improve indoor position accuracy.Existing sighting distance/non line of sight
The recognition methods in path needs specific equipment or carries out information off-line acquisition or need to be added some artificial deliberately operations,
Higher accuracy of identification can be reached, there is certain limitation.
Compared with the open signal transmission environment in outdoor, indoor environment is due to the separation and obstruction by a large amount of indoor objects
(such as indoor furniture, personnel's interference) etc. influences, so that the transmission environment of wireless signal becomes increasingly complex.The transmission of these signals
Path includes sighting distance and obstructed path, is increased by the delay that non-line-of-sight propagation bring error would generally generate signal, signal
Change of strength retrogression and angle of arrival etc. influences.Therefore, how to identify indoor signal propagation distance be sighting distance or non line of sight,
Influence according to indoor varying environment to signal propagation path judges, to reach the mesh for improving indoor objects positioning accuracy
's.
Indoors in signal communications system, indoor signal can generate obstructed path propagation effect because of indoor object, than
Such as increase of non-line-of-sight propagation distance, the delay of phase change and time of arrival (toa), therefore, receiving end received signal is normal
It is likely to will appear variation more sharply in a relatively short period of time.This influence generated by obstructed path, to a certain degree
On can reduce indoor position accuracy.How by the identification interior path LOS/NLOS, avoids or reduce by the path NLOS bring
Error becomes the research emphasis of present indoor positioning technologies researcher.
Existing related work realizes the identification work of indoor view distance/obstructed path at present.Pi-Chun Chen[1]It mentions
Go out a kind of to identify the path NLoS using the positioning residual error algorithm that is weighted positioning result.Document devises one in [2]
The method that kind identifies the path NLoS using QUADRATIC PROGRAMMING METHOD FOR.Later Li Cong et al.[3]Further in reaching time-difference
On the basis that (Time Difference of Arrival, abbreviation TDOA) is defined, when having studied location of mobile users arrival
Between poor non-line-of-sight propagation identification and correction, propose the method for a base station NLoS Path Recognition.Document [5] is mentioned using UWB
The side counted based on poly-diameter channel characteristic (such as kurtosis, average delay time and square root) to identify the path NLoS is gone out
Method.This method needs specific bandwidth to carry out the identification of realizing route, is not suitable for the narrowband network system of indoor ubiquitous deployment now
System.Wylie et al.[6]It puts measured range information in different times using different base station, believes in conjunction with the measurement of different time
The information such as standard deviation of breath difference and measurement noise devise the distinguished number that can judge to whether there is sighting distance between base station,
This method usually requires the identification that enough reference points participate in path, and the reference point of sighting distance is in the great majority.However in reality
In the environment of border, it is difficult to exist even without the reference point of sighting distance sometimes so that all environment-identifications all have this condition, because
This, this sighting distance recognition methods can not be suitable for most of occasions.
Document [4], which is proposed, carries out LoS path detection using Kalman technology.Document [7] by acquiring in room offline
A large amount of signal characteristics, certain several signal characteristic collection therein are trained using support vector machines to know to path
Not.But these methods are required to acquire a large amount of data in advance, carry out off-line training, and by different under different scenes
Signal characteristic.So each indoor environment changes, and is required to resurvey data, allows for identification workload so very
Greatly, it is not suitable for labile indoor scene.Document [8] [9] algorithm needs receiving end to be constantly in movement during realization
In the state of, the path LOS and NLOS is identified by increasing the randomness of receiving end signal, and is needed to mass data packet
Signal characteristic is counted to reach high-precision discrimination, can generate additional manpower costs and real-time is poor.PhaseU
Algorithm[10]During LOS under realizing dynamic environment is identified, the movement for needing target object to carry certain embedded sensors is set
It is standby, the path LOS and NLOS is distinguished by information such as the direction of acquisition sensor and gravity, there is certain application limitation.
Explanation of nouns used in the present invention is as follows:
Sighting distance: referring between transmitting terminal and receiving end can be mutually in the distance of " seeing ", and signal is from transmitting terminal to reception
A kind of mode at end.
Non line of sight: after the signal of transmitting terminal encounters barrier during straightline propagation to be hindered, the propagation path of signal
It can change, signal can reach receiving end by the circulation ways such as reflecting or reflecting.
A kind of channel state information (Channel State Information, CSI): channel attribute of communication link.It
The individual features of signal on different frequency bands are described in a frequency domain.
Channel impulse response (Channel Impulse Response, CIR): a kind of channel attribute of communication link,
The shock effect of signal generation is described in time domain.
Summary of the invention
The present invention is intended to provide a kind of sighting distance based on physical layer information/obstructed path recognition methods, avoids by offline
Time caused by training and manpower costs solve in the indoor environment of propagation path complexity, differentiate sighting distance/obstructed path
The problem of, to improve the precision of positioning, achieve the purpose that high precision position service quality.
In order to solve the above technical problems, the technical scheme adopted by the invention is that: one kind being based on channel impulse response time domain
The sighting distance of energy distribution information/obstructed path recognition methods, which comprises the following steps:
1) the corresponding CSI data of continuous K data packet are obtained；
2) inversefouriertransform is carried out to each CSI data, obtains CSI channel impulse response, each CSI Channel Impulse is rung
It should be expressed as the vector V=[s1 ..., s30] comprising 30 elements, indicate the energy point in the corresponding time domain of the data packet
Cloth, wherein si indicates the energy amplitude at i-th of moment in CSI channel impulse response；I=1,2 ... ..., 30；
3) following feature is calculated based on obtained K time domain energy distributed intelligence:
A) mean value and variance of ceiling capacity amplitude corresponding sequence；
B) mean value and variance of secondary big energy amplitude corresponding sequence；
C) ceiling capacity amplitude accounts for the mean value of the ratio of all energy；
D) maximum and time big energy amplitude accounts for the mean value of the ratio of all energy；
E) highest 5 time serieses of energy correspond to the ratio mean value that gross energy accounts for all energy；
F) maximum and time big energy amplitude difference；
G) absolute value of the difference of ceiling capacity amplitude corresponding sequence and time big energy corresponding sequence；
4) it is based on the feature of step 3), train the classifier C1 of sighting distance and obstructed path using machine learning method and is divided
Class device C2, step include:
A) the corresponding CSI information of data packet under sighting distance and obstructed path is acquired, calculates in step 3) listed feature simultaneously
It is marked, as training sample；
B) classifier of sighting distance and obstructed path is respectively trained based on training sample, is denoted as classifier C1 and classifier
C2；
5) when actually being judged, the corresponding CSI feature of several data packets of continuous acquisition, and as classifier C1 and
The input of classifier C2, and determine that output result is as follows: if C1 output is positive example, C2 output is counter-example, determines currently to be view
Away from；If C1 output is counter-example and C2 output is positive example, it is determined as being currently non line of sight；Otherwise, compare C1 and C2 output result
Confidence level, using output with a high credibility as final result.
In step 2), the specific implementation process for obtaining CSI channel impulse response includes:
1) each CSI data indicate the amplitude and phase of a subcarrier:
Wherein, H (fk) indicate centre carrier frequency fkCSI information, | | H (fk) | | and
∠H(fk) respectively indicate H (fk) amplitude and phase information；
2) to H (fk) inversefouriertransform is carried out, obtain H (fk) channel impulse response:
Wherein, ai, θiAnd τiAmplitude, phase and the time delay on i-th of path are respectively indicated, N refers to all path numbers, δ
(τ) is delta function.
The specific implementation process of step 4) includes:
A) the corresponding CSI information of data packet under sighting distance and obstructed path is acquired, calculates in step 3) listed feature simultaneously
It is marked, as training sample；
B) classifier of sighting distance and obstructed path is respectively trained based on training sample, is denoted as classifier C1 and classifier
C2。
Compared with prior art, the advantageous effect of present invention is that: the present invention, and can without offline acquisition data
Data are re-started to avoid the variation because of environment acquires generated manpower cost.In addition, this method applies also for moving
In state environment, when signal transmission path is changed between sighting distance/non line of sight, the change of transmission path can be also identified in real time
Change.
Detailed description of the invention
Fig. 1 is the corresponding energy profile of different time sequence in the channel impulse response under los path；
Fig. 2 is the corresponding energy profile of different time sequence in the channel impulse response under obstructed path；
Relationship of the Fig. 3 between data packet number and Path Recognition rate；
Fig. 4 is influence of the time to discrimination.
Specific embodiment
(1) channel state information of Wi-Fi is obtained
Currently, CSI information can be obtained using commercial network interface card Intel 5300.In an experiment, our unified uses are equipped with
The notebook of the commercial network interface card of Intel 5300 acquires CSI information with the wireless router of IEEE 802.11n standard is supported.Pen
Remember equipped with suitable operating system and correlation CSI integration tool on this, after notebook connects above Wi-Fi Hotspot, just in terminal
Ping order is sent to router with every 2 milliseconds of frequency, then saves received integrated data package informatin.
In CSI information, each subcarrier indicates signal status information of the transmission signal from transmitting terminal to receiving end, utilizes Intel 5300
Network interface card can obtain 30 subcarrier informations.What each CSI was indicated is the amplitude and phase of each subcarrier:
Wherein, H (fk) indicate centre carrier frequency fkCSI, | | H (fk) | | and ∠ H (fk) respectively indicate its amplitude
And phase information.
(2) channel impulse response
The CSI information that receiving end obtains belongs to channel frequency domain response (the Channel Frequency of wireless signal
Response, CFR), we are by carrying out anti-Fourier transform (Inverse Fourier to channel state information
Transform, IFT), its available channel impulse response (Channel Impulse Response, CIR):
Wherein, ai, θiAnd τiDistribution indicates amplitude, phase and the time delay on i-th of path, and N refers to all path numbers, δ
(τ) is delta function.
After obtaining the sub-carrier signal feature that each different time reaches according to CIR, pass through the letter reached to different delay
Road information, so that it may different characteristic information of the analysis under the path LOS and NLOS.
(3) channel characteristics select
After we acquire more CSI information, a very stable feature letter is had found when carrying out anti-Fourier analysis to it
Breath, for some data packet: under LOS path, possess can the subcarrier of magnitude can always concentrate and appear in some
Regular time point (such as the energy value in the time series that always the 7th, 8 time point reaches in our experimental situation
It is maximum), and under the path NLOS, can magnitude subcarrier but show very random (maximum energy value possibly is present at the 3rd
A time point of arrival or the 15th, 30 time point).
Due to being influenced under signal propagation path indoors NLOS environment by barrier, signal propagation path can change
Become, so that the transmitting path of each sub-carrier selection and different under LOS environment, therefore arrival time also can be with accordingly
It is inconsistent under LOS environment.According to upper figure under the different paths LOS/NLOS, the signal energy of CIR is distributed difference, we are to most
This feature of the corresponding time series of high-energy value has carried out detailed analysis comparison.
Indoors in the realization of location technology, in order to improve positioning accuracy, the acquisition in certain time is often utilized
A series of signal information is analyzed.Originally, we have carried out comparative analysis to the CIR Energy distribution of a CSI data packet, are
Prove the validity of this feature value, we, which are arranged in every group of experiment, acquires 300 data packets, to 300 data packets
It time series can be counted corresponding to magnitude.It was found that under LOS environment, the highest energy of 300 data packets
The probability for being worth corresponding same time sequence is 80% or more, and in a nlos environment 300 can the magnitude corresponding time
Sequence but shows more dispersed, each packet can time for reaching of magnitude it is also more random.
By being compared in the signal energy distribution map that different time reaches to the data packet under 300 varying environments
It can be found that signal propagation path is more stable under LOS path, can magnitude appear in sequence at the same time
Ratio is up to 80%, and under the path NLOS, can the time series that occurs of magnitude be in stochastic regime, same temporal
Ratio is smaller.Not accidental under LOS/NLOS environment in order to verify the CIR feature, we are different real then in different time
It tests on place and distinct device and is tested respectively.
Pass through the experimental result of different time and place, it was demonstrated that using CIR can magnitude time series as system
Feature is counted to identify that LOS/NLOS is not accidental.Under LOS environment, because its signal propagation path does not have the dry of other barriers
It disturbs, the transmitting path of multiple data is relatively stable；And in a nlos environment, since the barrier between link is to signal of communication
Interference, after causing signal to be propagated by obstructed paths such as reflection, refractions, time of arrival (toa) can change, and corresponding
The distribution of signal energy also can be because the appearance of barrier with the experimental data under above-mentioned multiple varying environment without showing
Using CIR can magnitude time series distribution probability come identify the path LOS/NLOS have universality.
(4) algorithm designs
Based on CSI data collected, following feature is calculated, and based on these features two classifier C1 (sighting distance) of training
With C2 (non line of sight).Specific step is as follows:
1) the corresponding CSI data of continuous K data packet are obtained；
2) inversefouriertransform is carried out to each CSI data, obtains channel impulse response, each CSI channel pulse is corresponding
It is expressed as the vector V=[s1 ..., s30] comprising 30 elements, indicates the energy point in the corresponding time domain of the data packet
Cloth, wherein si indicates the energy amplitude at i-th of moment in response；
3) following feature is calculated based on obtained K time domain energy distributed intelligence
A) mean value and variance of ceiling capacity amplitude corresponding sequence；
B) mean value and variance of secondary big energy amplitude corresponding sequence；
C) ceiling capacity amplitude accounts for the mean value of the ratio of all energy；
D) maximum and time big energy amplitude accounts for the mean value of the ratio of all energy；
E) highest 5 time serieses of energy correspond to the ratio mean value that gross energy accounts for all energy；
F) maximum and time big energy amplitude difference；
G) absolute value of the difference of ceiling capacity amplitude corresponding sequence and time big energy corresponding sequence
4) it is based on preceding feature, using the classifier of machine learning method training sighting distance and obstructed path, step includes:
A) the corresponding CSI information of data packet under sighting distance and obstructed path is acquired, listed feature in 3) is calculated and is carried out
Label, as training sample；
B) classifier of sighting distance and obstructed path is respectively trained based on training sample, is denoted as classifier C1 and C2；
5) when actually being judged, the corresponding CSI feature of several data packets of continuous acquisition, and as classifier C1 and
The input of C2, and determine that output result is as follows:
If a) C1 output is positive example and C2 output is counter-example, determine currently to be sighting distance；
If b) C1 output is counter-example and C2 output is positive example, it is determined as being currently non line of sight；
C) otherwise, compare the confidence level of C1 and C2 output result, as a result with output with a high credibility.
When we acquire data in the actual environment, data transmission frequency is set as 2 milliseconds, packet capture number is set
It is 300.After we take a large amount of CSI data, data are handled using Matlab, and to the view based on physical layer information
Performance Evaluation has been carried out away from/obstructed path recognition methods.
(1) influence of the number of different data packet to discrimination
When the data amount check of every group of experiment acquisition is more, the path status indicated is then more accurate, correspondingly, algorithm pair
The recognition result of los path is also more accurate.Consider that the requirement of los path discrimination the high, answers in practical applications
It is more ideal with effect.Especially in the environmental demand of high request, comparatively ideal Path Recognition rate is needed, therefore, we will be real
The number for testing acquisition data packet continues growing in the case where 300, and whether the Path Recognition accuracy rate for analyzing the algorithm also can
Rise with the increase of data packet number.
It will be apparent that Fig. 3 illustrate different data packet number be on sighting distance discrimination it is influential, data packet number is more,
Its discrimination is also higher.When data packet number is 50, LoS/NLoS discrimination is respectively 90.1% and 88.8%, by data
When packet number increases to 300,94% and 92.7% has also been respectively increased in discrimination, substantially has reached the water of LiFi algorithm
It is flat.In order to which preferably compared with LiFi algorithm, the number of data packet is set as and the data packet of LiFi algorithm by we
When number is all 2000, the accuracy of this algorithm is correct compared to LiFi algorithm 90% and 93.09% up to 97.5 and 94.3%
Rate, this algorithm take advantage on discrimination.
(2) assessment of algorithm real-time energy
And we are in data acquisition, the frequency for obtaining data is 2 milliseconds, so we only need to 50 data
100 milliseconds of time is expended to obtain data, discrimination 90%.In contrast, LiFi, which reaches identical precision, will but expend 4
The data acquisition time of second.From the point of view of real-time performance, our algorithm ratio LiFi algorithm is stronger.This algorithm is to 100 milliseconds
Under acquisition time, 90.1% discrimination can achieve, as shown in figure 4, the accuracy is to a certain extent acceptable
In range, do not have to precision but during special emphasis applies requirement of real-time is higher especially, this algorithm has certain superior
Property.
When system is more demanding to accuracy, it can suitably increase data amount check in data acquisition phase, can achieve
Improve accuracy of identification.It is demonstrated experimentally that reaching 94% discrimination only needs 600 milli of time-consuming when handling 300 data
Second, it is strong compared to LiFi algorithm real-time, it is high-efficient.When the number to data packet increases to 2000, need to consume 4000 at this time
The acquisition time of millisecond, sighting distance discrimination may be up to 97.5%, considerably beyond LiFi Path Recognition algorithm in terms of discrimination.
We compare experimental result and newest LiFi algorithm.LiFi algorithm is proposed to be believed based on channel status
The method of characteristic statistics is ceased to realize indoor view distance Path Recognition.This method main thought is increased using the movement of receiving end
The signal randomness in signal non-line-of-sight propagation path, it according to the skewness and kurtosis of signal as signal characteristic, according to its point
Cloth distinguishes sighting distance/obstructed path.When data volume is less, the degree of bias and peak value performance on different paths are not obvious,
Therefore, LiFi algorithm needs more data volume to prove its different presentation in sighting distance and obstructed path, then in number
The more time will be expended according to acquisition aspect.Also, the realization of the algorithm needs to guarantee receiving end when receiving data always
It is kept in motion, therefore, LiFi algorithm has certain limitation in its application aspect.
The discrimination of 1 different classifications device of table
By Comparison of experiment results, LiFi algorithm is when handling 50 data packets, the accuracy rate point of sighting distance and non line of sight
Not Wei 77.5% and 82.5%, when data amount check increases to 2000, accuracy rate can be promoted to 90% and 93.09%.And I
Algorithm when handling 50 data packets, sighting distance/non line of sight discrimination is up to 94.6% and 93%, to 2000
When data packet is handled, discrimination is up to 97.5% and 94.3%.And it can thus be seen that this algorithm is in terms of real-time
Advantage more than LiFi algorithm.
Bibliography:
[1]Chen P C.A Non-Line-Of-Sight Error Mitigation Algorithm In
Location Estimation[C]//Wireless Communications and NETWORKING Conference,
1999.WCNC.IEEE Xplore,1999:316-320vol.1.
[2]Wang X,Wang Z,O'Dea B.A TOA-based location algorithm reducing the
errors due to non-line-of-sight(NLoS)propagation[J].IEEE Transactions on
Vehicular Technology,2003,52(1):112-116.
[3]Cong L,Zhuang W.Non-line-of-sight error mitigation in TDOA mobile
location[C]//Global Telecommunications Conference,2001.GLOBECOM'01.IEEE.IEEE,
2001,1:680-684.
[4]Yu K,Dutkiewicz E.Improved kalman filtering algorithms for mobile
tracking in NLoS scenarios[C]//Wireless Communications and Networking
Conference(WCNC),2012IEEE.IEEE,2012:2390-2394.
[5]Zhang J,Salmi J,Lohan E S.Analysis of kurtosis-based LoS/NLoS
identification using indoor MIMO channel measurement[J].IEEE transactions on
vehicular technology,2013,62(6):2871-2874.
[6]Wylie M P,Holtzman J.The non-line of sight problem in mobile
location estimation[C]//IEEE International Conference on Universal Personal
Communications.IEEE Xplore,1996:827-831 vol.2.
[7]Xiao Z,Wen H,Markham A,et al.Non-line-of-sight identification and
mitigation using received signal strength[J].IEEE Transactions on Wireless
Communications,2015,14(3):1689-1702.
[8]Wu C,Yang Z,Zhou Z,et al.PhaseU:Real-time LoS identification with
WiFi[C]//2015 IEEE Conference on Computer Communications(INFOCOM).IEEE,2015:
2038-2046.
[9]Zhou Z,Yang Z,Wu C,et al.WiFi-Based Indoor Line-of-Sight
Identification[J].IEEE Transactions on Wireless Communications,2015,14(11):
6125-6136.
[10]Zhou Z,Yang Z,Wu C,et al.Lifi:Line-of-sight identification with
wifi[C]//IEEE INFOCOM 2014-IEEE Conference on Computer Communications.IEEE,
2014:2688-2696.
Claims (3)
1. a kind of sighting distance based on channel impulse response time domain energy distributed intelligence/obstructed path recognition methods, feature exist
In, comprising the following steps:
1) the corresponding CSI data of continuous K data packet are obtained；
2) inversefouriertransform is carried out to each CSI data, obtains CSI channel impulse response, each CSI channel impulse response table
It is shown as the vector V=[s1 ..., s30] comprising 30 elements, indicates the Energy distribution in the corresponding time domain of the data packet,
Wherein si indicates the energy amplitude at i-th of moment in CSI channel impulse response；I=1,2 ... ..., 30；
3) following feature is calculated based on obtained K time domain energy distributed intelligence:
A) mean value and variance of ceiling capacity amplitude corresponding sequence；
B) mean value and variance of secondary big energy amplitude corresponding sequence；
C) ceiling capacity amplitude accounts for the mean value of the ratio of all energy；
D) maximum and time big energy amplitude accounts for the mean value of the ratio of all energy；
E) highest 5 time serieses of energy correspond to the ratio mean value that gross energy accounts for all energy；
F) maximum and time big energy amplitude difference；
G) absolute value of the difference of ceiling capacity amplitude corresponding sequence and time big energy corresponding sequence；
4) it is based on the feature of step 3), utilizes the classifier C1 and classifier of machine learning method training sighting distance and obstructed path
C2, step include:
A) the corresponding CSI information of data packet under sighting distance and obstructed path is acquired, listed feature in step 3) is calculated and is carried out
Label, as training sample；
B) classifier of sighting distance and obstructed path is respectively trained based on training sample, is denoted as classifier C1 and classifier C2；
5) when actually being judged, the corresponding CSI feature of several data packets of continuous acquisition, and as classifier C1 and classification
The input of device C2, and determine that output result is as follows: if C1 output is positive example, C2 output is counter-example, is determined currently for sighting distance；
If C1 output is counter-example and C2 output is positive example, it is determined as being currently non line of sight；Otherwise, compare C1 and C2 output result can
Reliability, using output with a high credibility as final result.
2. the sighting distance based on physical layer information/obstructed path recognition methods according to claim 1, which is characterized in that
In step 2), the specific implementation process for obtaining CSI channel impulse response includes:
1) each CSI data indicate the amplitude and phase of a subcarrier:
Wherein, H (fk) indicate centre carrier frequency fkCSI information, | | H (fk) | | and ∠ H (fk) respectively indicate H (fk)
Amplitude and phase information；
2) to H (fk) inversefouriertransform is carried out, obtain H (fk) channel impulse response:
Wherein, ai, θiAnd τiAmplitude, phase and the time delay on i-th of path are respectively indicated, N refers to all path numbers, and δ (τ) is
Delta function.
3. the sighting distance based on physical layer information/obstructed path recognition methods according to claim 1, which is characterized in that
The specific implementation process of step 4) includes:
A) the corresponding CSI information of data packet under sighting distance and obstructed path is acquired, listed feature in step 3) is calculated and is carried out
Label, as training sample；
B) classifier of sighting distance and obstructed path is respectively trained based on training sample, is denoted as classifier C1 and classifier C2.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810057363.7A CN109151724B (en) | 2018-01-22 | 2018-01-22 | Line-of-sight/non-line-of-sight path identification method based on channel impulse response energy distribution |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810057363.7A CN109151724B (en) | 2018-01-22 | 2018-01-22 | Line-of-sight/non-line-of-sight path identification method based on channel impulse response energy distribution |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109151724A true CN109151724A (en) | 2019-01-04 |
| CN109151724B CN109151724B (en) | 2020-12-22 |
Family
ID=64801616
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810057363.7A Active CN109151724B (en) | 2018-01-22 | 2018-01-22 | Line-of-sight/non-line-of-sight path identification method based on channel impulse response energy distribution |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109151724B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113194427A (en) * | 2021-04-30 | 2021-07-30 | 长安大学 | Identification method, system and device based on soft-decision visible and non-visible channels |
| WO2022170632A1 (en) * | 2021-02-15 | 2022-08-18 | 苏州优它科技有限公司 | Positioning simulation method based on line-of-sight path identification |
| CN116582815A (en) * | 2023-05-22 | 2023-08-11 | 青岛柯锐思德电子科技有限公司 | LOS and NLOS scene judging method based on ranging channel evaluation |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103399297A (en) * | 2013-08-16 | 2013-11-20 | 哈尔滨工业大学深圳研究生院 | Machine learning based ultra-wideband NLOS (non line of sight) identification method |
| CN106792808A (en) * | 2016-12-08 | 2017-05-31 | 南京邮电大学 | Los path recognition methods under a kind of indoor environment based on channel condition information |
| WO2017100706A1 (en) * | 2015-12-09 | 2017-06-15 | Origin Wireless, Inc. | Method, apparatus, and systems for wireless event detection and monitoring |
| CN107037402A (en) * | 2017-05-24 | 2017-08-11 | 南京图贝斯韦智能技术有限公司 | A kind of localization method under indoor nlos environment based on UWB rangings |
-
2018
- 2018-01-22 CN CN201810057363.7A patent/CN109151724B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103399297A (en) * | 2013-08-16 | 2013-11-20 | 哈尔滨工业大学深圳研究生院 | Machine learning based ultra-wideband NLOS (non line of sight) identification method |
| WO2017100706A1 (en) * | 2015-12-09 | 2017-06-15 | Origin Wireless, Inc. | Method, apparatus, and systems for wireless event detection and monitoring |
| CN106792808A (en) * | 2016-12-08 | 2017-05-31 | 南京邮电大学 | Los path recognition methods under a kind of indoor environment based on channel condition information |
| CN107037402A (en) * | 2017-05-24 | 2017-08-11 | 南京图贝斯韦智能技术有限公司 | A kind of localization method under indoor nlos environment based on UWB rangings |
Non-Patent Citations (2)
| Title |
|---|
| JENS SCHROEDER ET. AL: "NLOS detection algorithms for Ultra-Wideband localization", 《2007 4TH WORKSHOP ON POSITIONING, NAVIGATION AND COMMUNICATION》 * |
| 蔡雄: "基于CSI的室内WiFi定位技术研究", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2022170632A1 (en) * | 2021-02-15 | 2022-08-18 | 苏州优它科技有限公司 | Positioning simulation method based on line-of-sight path identification |
| CN113194427A (en) * | 2021-04-30 | 2021-07-30 | 长安大学 | Identification method, system and device based on soft-decision visible and non-visible channels |
| CN116582815A (en) * | 2023-05-22 | 2023-08-11 | 青岛柯锐思德电子科技有限公司 | LOS and NLOS scene judging method based on ranging channel evaluation |
| CN116582815B (en) * | 2023-05-22 | 2023-10-24 | 青岛柯锐思德电子科技有限公司 | LOS and NLOS scene judging method based on ranging channel evaluation |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109151724B (en) | 2020-12-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109151707A (en) | A kind of sighting distance/obstructed path recognition methods under moving condition | |
| Seow et al. | Non-line-of-sight localization in multipath environments | |
| Jiang et al. | Communicating is crowdsourcing: Wi-Fi indoor localization with CSI-based speed estimation | |
| US7574221B2 (en) | Method for estimating jointly time-of-arrival of signals and terminal location | |
| WO2018133264A1 (en) | Indoor automatic human body positioning detection method and system | |
| CN106792808A (en) | Los path recognition methods under a kind of indoor environment based on channel condition information | |
| CN104883732B (en) | A kind of enhanced indoor passive passive human body localization method | |
| Boudhir et al. | New technique of wireless sensor networks localization based on energy consumption | |
| CN109155984A (en) | Determine method, localization method and the relevant device of channel time delay | |
| US9462421B1 (en) | Indoor-outdoor detector for estimating the location of a wireless terminal | |
| CN109151724A (en) | Sighting distance based on channel impulse response Energy distribution/obstructed path recognition methods | |
| CN110475221A (en) | A kind of personnel's action recognition and location estimation method based on channel state information | |
| Sung et al. | Accurate indoor positioning for UWB-based personal devices using deep learning | |
| CN109819394A (en) | Based on the WiFi indoor orientation method mixed with ultrasonic wave and its system | |
| CN106411433A (en) | WLAN-based fine-grained indoor passive intrusion detection method | |
| Chen et al. | Indoor pedestrian tracking with sparse RSS fingerprints | |
| CN108226912B (en) | Sparse network-based non-contact object perception positioning method and system | |
| Mizmizi et al. | Design of RSSI based fingerprinting with reduced quantization measures | |
| CN108882171A (en) | A kind of personnel's trace tracking method based on CSI | |
| Wang et al. | Through the wall detection and localization of autonomous mobile device in indoor scenario | |
| Hashem et al. | Deepnar: Robust time-based sub-meter indoor localization using deep learning | |
| Fan et al. | Single-site indoor fingerprint localization based on MIMO-CSI | |
| CN107911863B (en) | A method of the position malice AP is determined based on simple gesture | |
| CN105531600B (en) | Time analysis in wireless network for user velocity estimation | |
| Ciurana et al. | SofTOA: Software ranging for TOA-based positioning of WLAN terminals |
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 |