CN110222776A - A kind of indoor Passive Location based on CSI finger print information - Google Patents
A kind of indoor Passive Location based on CSI finger print information Download PDFInfo
- Publication number
- CN110222776A CN110222776A CN201910498687.9A CN201910498687A CN110222776A CN 110222776 A CN110222776 A CN 110222776A CN 201910498687 A CN201910498687 A CN 201910498687A CN 110222776 A CN110222776 A CN 110222776A
- Authority
- CN
- China
- Prior art keywords
- csi
- classification results
- classifier
- training
- finger print
- 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
- 238000012549 training Methods 0.000 claims description 30
- 238000009826 distribution Methods 0.000 claims description 16
- 230000009466 transformation Effects 0.000 claims description 9
- 238000012360 testing method Methods 0.000 claims description 7
- 238000000034 method Methods 0.000 description 7
- 238000005516 engineering process Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000013507 mapping Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000007792 addition Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 239000012141 concentrate Substances 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/2431—Multiple classes
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
- H04W4/30—Services specially adapted for particular environments, situations or purposes
- H04W4/33—Services specially adapted for particular environments, situations or purposes for indoor environments, e.g. buildings
Abstract
The invention discloses a kind of indoor Passive Locations based on CSI finger print information, make full use of the effective information in WiFi environment, indoor Passive Positioning is carried out using CSI phase information as finger print information, matching accuracy rate is improved by Ensemble Learning Algorithms, confidence recurrence is carried out to matching result again, improves the precision of indoor Passive Positioning.
Description
Technical field
The present invention relates to indoor positioning technologies field, specifically a kind of indoor Passive Positioning side based on CSI finger print information
Method.
Background technique
With the development of society and the progress of science and technology, location based service has become essential in people's life
A part.Based on the indoor locating system of WiFi due to the advantages that its equipment cost is low, easy deployment, become indoor positioning in recent years
The hot spot of research.
Traditional WiFi indoor orientation method generally uses received signal strength (RSS) as parameter, but RSS is as MAC
Layer information, only the simple superposition of each subcarrier reception signal strength, is easy to be done by extraneous factors such as wall, furniture
It disturbs, influences positioning accuracy.As OFDM technology and MIMO technology are in the application of WiFi standard, so that obtaining the channel shape of physical layer
State information (CSI) is possibly realized, and CSI is a kind of parameter for reflecting channel status more fine granularity compared with RSS, is replaced using CSI
Positioning accuracy can be improved as finger print information in RSS.
Channel state information is the information based on physical layer, describes the amplitude and phase two of each subcarrier reception signal
A feature.In the prior art, most of indoor orientation methods based on CSI all utilize the amplitude information of CSI, the reason is that,
Although we can get the phase information of CSI from Intel5300 network interface card, because the reasons such as error, get
Original phase is often to be distributed in a jumble at random, can not be used directly to as finger print information;And amplitude information be easy to because
The obstructions of other facilities and sharp fall, to influence positioning accuracy.
Summary of the invention
In view of the above-mentioned problems, the present invention provides a kind of indoor Passive Location based on CSI finger print information.
A kind of indoor Passive Location based on CSI finger print information, including off-line training step and tuning on-line stage,
Using CSI phase information as finger print information,
The off-line training step at least includes the following steps:
Step 1 acquires CSI data to ready-portioned fingerprint point, to the original phase information of its CSI carry out linear transformation with
Phase offset is eliminated, revised phase information is obtained as sample data, establishes CSI fingerprint base;
Step 2 carries out weight training to sample data using Ensemble Learning Algorithms (i.e. adaboost algorithm), obtains one
Strong classifier;
The tuning on-line stage at least includes the following steps:
It is inclined to eliminate phase to carry out linear transformation to the original phase information of its CSI for step 3, acquisition tested point CSI data
It moves, obtains revised phase information as testing data, and in the strong classifier in input step 2 and obtain classification results;
Step 4 counts the classification results of tested point, chooses r classification results most in all classification results,
Mark position coordinate is Lr(x, y), wherein r=1,2 ..., R, mark corresponding classification results quantity difference n1,n2,…,nR, mark
The sum for remembering r classification results is N=n1+n2+…+nR, then each coordinate LrThe Probability p of (x, y) in this N number of resultr=nr/ N,
It is weighted recurrence respectively according to coordinate (x, y) of the probability distribution to r classification results and obtains final output coordinate
Further, the linear transform in the step 1 isWhereinThe revised phase of i-th of subcarrier is represented,Represent the uncorrected phase of i-th of subcarrier, waveiIt represents
The index value of i-th of subcarrier, A, b respectively represent two parameters of linear transformation,
Further, the step 2 specifically includes the following steps:
S1, CSI fingerprint base is divided into training sample and verifying sample, the weight distribution of training sample is initialized, so that often
One training sample is all endowed identical weight wiThe weight initial distribution of=1/N, training sample meet D1(i)=(w1,
W2 ..., wN)=(1/N, 1/N ..., 1/N);
S2, training sample is inputted in multiple classifiers, carries out classifier training, obtains trained multiple weak typings
Device;
S3, verifying sample is inputted in trained Weak Classifier to the classification results for obtaining each classifier;
S4, the minimum Weak Classifier of a current erroneous rate is chosen as t-th of basic classification device Ht, wherein t=1,
2 ..., T, and calculate Weak Classifier HtIn distribution DtOn error be
S5, Weak Classifier H is calculatedtThe shared weight in final classification device
S6, the weight distribution for updating training datasetWherein ZtIt is normal to normalize
Number,
S7, according to Weak Classifier HtWeight αtEach Weak Classifier is combined, i.e.,
S8, a strong classifier is obtained by sign function sign
The present invention makes full use of the effective information in WiFi environment, carries out using CSI phase information as finger print information indoor
Passive Positioning improves matching accuracy rate by Ensemble Learning Algorithms, then carries out confidence recurrence to matching result, improves indoor quilt
The precision of dynamic positioning.
Detailed description of the invention
Fig. 1 is the scene top view of embodiment 1;
Fig. 2 is invention's principle block diagram.
Specific embodiment
The present invention is described in further detail in the following with reference to the drawings and specific embodiments.The embodiment of the present invention be for
It is provided for the sake of example and description, and is not exhaustively or to limit the invention to disclosed form.Much repair
It is obvious for changing and change for the ordinary skill in the art.Selection and description embodiment are in order to more all right
Bright the principle of the present invention and practical application, and make those skilled in the art it will be appreciated that the present invention is suitable for design
The various embodiments with various modifications of special-purpose.
Embodiment 1
Since in environment indoors, signal is influenced by multipath, using single signal and distance mapping relations into
Row trilateration is difficult to reach high-precision positioning, and Microsoft Research proposes in WiFi indoor locating system RADAR for the first time
The sample data of RSSI is referred to as wireless map or location fingerprint by fingerprint location method.Fingerprint location method is by terminal wait estimate
Meter position and the station acquisition to wireless signal (RSSI or CSI) be associated, compare the signal characteristic fingerprint of wireless signal
Information carries out target position estimation.
Fingerprint location method is divided into off-line training step and tuning on-line stage, by taking CSI signal as an example.Off-line training step is
The CSI for acquiring multiple AP handles CSI information, and main includes the analysis of amplitude and phase, realizes between position and CSI
Mapping, establishes wireless map (MP) and CSI fingerprint base;After the tuning on-line stage is handled the real-time CSI value of user terminal
It is matched with MP, obtains location information, realize positioning.
The invention discloses a kind of indoor Passive Locations based on CSI finger print information, comprising the following steps:
One, use TL-WDR6500 wireless router and Intel5300 network interface card as the present embodiment WiFi signal respectively
Transmitting terminal and receiving end.
Two, 16 points for choosing 4 × 4 distributions within a certain area are shown in that Fig. 1 dot, each point choose 200 data, altogether
200 × 16 data of meter as fingerprint point and mark corresponding label.
Three, CSI data are acquired to ready-portioned fingerprint point, linear transformation is carried out to disappear to the original phase information of its CSI
Except phase offset, revised phase information is obtained as sample data, establishes CSI fingerprint base;Wherein linear transform isWhereinThe revised phase of i-th of subcarrier is represented,Generation
The uncorrected phase of i-th of subcarrier of table, waveiThe index value of i-th of subcarrier is represented, A, b respectively represent linear transformation
Two parameters,
Four, using adaboost algorithm to sample data carry out weight training, obtain a strong classifier, specifically include with
Lower step:
1. CSI fingerprint base is divided into training sample and verifying sample, the weight distribution of training sample is initialized, so that each
A training sample is all endowed identical weight wiThe weight initial distribution of=1/N, training sample meet D1(i)=(w1,
W2 ..., wN)=(1/N, 1/N ..., 1/N);
2. pair training sample inputs in multiple classifiers, classifier training is carried out, trained multiple Weak Classifiers are obtained;
3. verifying sample to be inputted in trained Weak Classifier to the classification results for obtaining each classifier;
4. choosing the minimum Weak Classifier of a current erroneous rate as t-th of basic classification device Ht, wherein t=1,
2 ..., T, and calculate Weak Classifier HtIn distribution DtOn error be
5. calculating Weak Classifier HtThe shared weight in final classification device
6. updating the weight distribution of training datasetWherein ZtIt is normal to normalize
Number,
7. according to Weak Classifier HtWeight αtEach Weak Classifier is combined, i.e.,
8. obtaining a strong classifier by sign function sign
Five, tested point CSI data are acquired, tested point is shown in Fig. 1 triangle, carries out to the original phase information of its CSI linear
Transformation obtains revised phase information as testing data, and in the strong classifier in input step 2 to eliminate phase offset
Obtain classification results;Linear transform is the same as fingerprint point CSI phse conversion.
Six, the classification results of tested point are counted, chooses 4 classification results most in all classification results, mark
Note position coordinates are Lr(x, y), wherein r=1,2,3,4, mark corresponding classification results quantity difference n1,n2,n3,n4, label 4
The sum of a classification results is N=n1+n2+n3+n4, then each coordinate LrThe Probability p of (x, y) in this N number of resultr=nr/ N, root
It is weighted recurrence respectively according to coordinate (x, y) of the probability distribution to 4 classification results and obtains final output coordinate
Based on the present embodiment, when fingerprint point spacing is 1m, 40% test point position error concentrates on 0.3m to 0.5m,
80% test point tolerance is within 0.7m;When fingerprint point spacing is 1.5m, 50% test point position error is concentrated on
0.5m to 0.8m, 90% test point tolerance is within 1.1m.Above-mentioned data, compared to merely using amplitude information as fingerprint
Information, positioning accuracy averagely improve 0.3m.
The present invention uses CSI phase information as finger print information, but is not limited to believe only with CSI phase information as fingerprint
CSI amplitude information and phase information can also be combined the indoor Passive Positioning of realization by breath.Fig. 2 is invention's principle block diagram.
Obviously, described embodiment is only a part of the embodiments of the present invention, instead of all the embodiments, this hair
Bright person of ordinary skill in the field can make various modifications or additions to the described embodiments or use
Similar mode substitutes, and however, it does not deviate from the spirit of the invention or beyond the scope of the appended claims.Base
Embodiment in the present invention, this field and those of ordinary skill in the related art are without creative labor
Every other embodiment obtained, all should belong to the scope of protection of the invention.
Claims (3)
1. a kind of indoor Passive Location based on CSI finger print information, including off-line training step and tuning on-line stage,
It is characterized in that, using CSI phase information as finger print information,
The off-line training step at least includes the following steps:
Step 1 acquires CSI data to ready-portioned fingerprint point, carries out linear transformation to the original phase information of its CSI to eliminate
Phase offset obtains revised phase information as sample data, establishes CSI fingerprint base;
Step 2 carries out weight training to sample data using adaboost algorithm, obtains a strong classifier;
The tuning on-line stage at least includes the following steps:
Step 3, acquisition tested point CSI data carry out linear transformation to the original phase information of its CSI to eliminate phase offset,
It obtains revised phase information and obtains classification results as testing data, and in the strong classifier in input step 2;
Step 4 counts the classification results of tested point, chooses r classification results most in all classification results, label
Position coordinates are Lr(x, y), wherein r=1,2 ..., R, mark corresponding classification results quantity difference n1,n2,…,nR, mark r
The sum of a classification results is N=n1+n2+…+nR, then each coordinate LrThe Probability p of (x, y) in this N number of resultr=nr/ N, root
It is weighted recurrence respectively according to coordinate (x, y) of the probability distribution to r classification results and obtains final output coordinate
2. the indoor Passive Location according to claim 1 based on CSI finger print information, which is characterized in that the step
Linear transform in rapid 1 isWhereinRepresent i-th of subcarrier amendment
Phase afterwards,Represent the uncorrected phase of i-th of subcarrier, waveiThe index value of i-th of subcarrier is represented, A, b divide
Two parameters of linear transformation are not represented,
3. the indoor Passive Location according to claim 1 based on CSI finger print information, which is characterized in that the step
Rapid 2 specifically includes the following steps:
S1, CSI fingerprint base is divided into training sample and verifying sample, initializes the weight distribution of training sample, so that each
Training sample is all endowed identical weight wiThe weight initial distribution of=1/N, training sample meet D1(i)=(w1, w2 ...,
WN)=(1/N, 1/N ..., 1/N);
S2, training sample is inputted in multiple classifiers, carries out classifier training, obtains trained multiple Weak Classifiers;
S3, verifying sample is inputted in trained Weak Classifier to the classification results for obtaining each classifier;
S4, the minimum Weak Classifier of a current erroneous rate is chosen as t-th of basic classification device Ht, wherein t=1,2 ..., T,
And calculate Weak Classifier HtIn distribution DtOn error be
S5, Weak Classifier H is calculatedtThe shared weight in final classification device
S6, the weight distribution for updating training datasetWherein ZtFor normaliztion constant,
S7, according to Weak Classifier HtWeight αtEach Weak Classifier is combined, i.e.,
S8, a strong classifier is obtained by sign function sign
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910498687.9A CN110222776A (en) | 2019-06-10 | 2019-06-10 | A kind of indoor Passive Location based on CSI finger print information |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910498687.9A CN110222776A (en) | 2019-06-10 | 2019-06-10 | A kind of indoor Passive Location based on CSI finger print information |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110222776A true CN110222776A (en) | 2019-09-10 |
Family
ID=67816191
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910498687.9A Pending CN110222776A (en) | 2019-06-10 | 2019-06-10 | A kind of indoor Passive Location based on CSI finger print information |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110222776A (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110736963A (en) * | 2019-10-21 | 2020-01-31 | 普联技术有限公司 | indoor Wi-Fi positioning method, device and storage medium based on CSI |
CN110933633A (en) * | 2019-12-05 | 2020-03-27 | 武汉理工大学 | Onboard environment indoor positioning method based on CSI fingerprint feature migration |
CN111641913A (en) * | 2020-04-14 | 2020-09-08 | 浙江大华技术股份有限公司 | Screen brightness control method and device, computer equipment and readable storage medium |
CN112867021A (en) * | 2021-01-13 | 2021-05-28 | 合肥工业大学 | Improved TrAdaBoost-based indoor positioning method for transfer learning |
CN114531729A (en) * | 2022-04-24 | 2022-05-24 | 南昌大学 | Positioning method, system, storage medium and device based on channel state information |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109151731A (en) * | 2018-06-14 | 2019-01-04 | 西北师范大学 | A kind of indoor orientation method based on CSI |
-
2019
- 2019-06-10 CN CN201910498687.9A patent/CN110222776A/en active Pending
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109151731A (en) * | 2018-06-14 | 2019-01-04 | 西北师范大学 | A kind of indoor orientation method based on CSI |
Non-Patent Citations (1)
Title |
---|
YONG ZHANG 等;: "《An Indoor Passive Positioning Method Using CSI Fingerprint Based on Adaboost》", 《IEEE SENSORS JOURNAL》 * |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110736963A (en) * | 2019-10-21 | 2020-01-31 | 普联技术有限公司 | indoor Wi-Fi positioning method, device and storage medium based on CSI |
CN110736963B (en) * | 2019-10-21 | 2022-03-08 | 普联技术有限公司 | Indoor Wi-Fi positioning method and device based on CSI and storage medium |
CN110933633A (en) * | 2019-12-05 | 2020-03-27 | 武汉理工大学 | Onboard environment indoor positioning method based on CSI fingerprint feature migration |
CN110933633B (en) * | 2019-12-05 | 2020-10-16 | 武汉理工大学 | Onboard environment indoor positioning method based on CSI fingerprint feature migration |
CN111641913A (en) * | 2020-04-14 | 2020-09-08 | 浙江大华技术股份有限公司 | Screen brightness control method and device, computer equipment and readable storage medium |
CN112867021A (en) * | 2021-01-13 | 2021-05-28 | 合肥工业大学 | Improved TrAdaBoost-based indoor positioning method for transfer learning |
CN114531729A (en) * | 2022-04-24 | 2022-05-24 | 南昌大学 | Positioning method, system, storage medium and device based on channel state information |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110222776A (en) | A kind of indoor Passive Location based on CSI finger print information | |
Li et al. | AF-DCGAN: Amplitude feature deep convolutional GAN for fingerprint construction in indoor localization systems | |
Honkavirta et al. | A comparative survey of WLAN location fingerprinting methods | |
Jin et al. | Indoor localization with channel impulse response based fingerprint and nonparametric regression | |
CN106793082B (en) | Mobile equipment positioning method in WLAN/Bluetooth heterogeneous network environment | |
CN105704652B (en) | Fingerprint base acquisition and optimization method in a kind of positioning of WLAN/ bluetooth | |
CN106851573A (en) | Joint weighting k nearest neighbor indoor orientation method based on log path loss model | |
CN103957505B (en) | A kind of action trail detection and analysis and service provider system and method based on AP | |
CN110381436B (en) | Rapid fingerprint positioning method based on large-scale MIMO single station system | |
CN103596267A (en) | Fingerprint map matching method based on Euclidean distances | |
CN109029429B (en) | WiFi and geomagnetic fingerprint based multi-classifier global dynamic fusion positioning method | |
CN106793086A (en) | A kind of indoor orientation method | |
KR20190053470A (en) | Positioning system based on deep learnin and construction method thereof | |
CN104038901B (en) | Indoor positioning method for reducing fingerprint data acquisition workload | |
CN109640262B (en) | Positioning method, system, equipment and storage medium based on mixed fingerprints | |
CN103901398B (en) | A kind of location fingerprint localization method based on combination collating sort | |
Li et al. | Multipath-assisted indoor localization using a single receiver | |
CN104540221B (en) | WLAN indoor orientation methods based on semi-supervised SDE algorithms | |
CN107592611B (en) | Wireless positioning method and system for broadband large-scale MIMO system | |
CN107087256A (en) | A kind of fingerprint cluster method and device based on WiFi indoor positionings | |
CN110488222B (en) | UWB positioning method combining SVM (support vector machine) and barycentric coordinate under NLOS (non line of sight) condition | |
CN109951798A (en) | Merge the enhancing location fingerprint indoor orientation method of Wi-Fi and bluetooth | |
CN109922427A (en) | Utilize the intelligent radio positioning system and method for large scale array antenna | |
CN107613559A (en) | A kind of DOA fingerprint base localization methods based on 5G signals | |
CN103618997B (en) | Indoor positioning method and device based on signal intensity probability |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190910 |
|
RJ01 | Rejection of invention patent application after publication |