Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Deep-learning-based Wi-Fi indoor positioning method

An indoor positioning and deep learning technology, applied in neural learning methods, positioning, instruments, etc., can solve the problems of inaccurate positioning, noisy original RSS data, affecting the accuracy and reliability of subsequent positions, and achieve the effect of improving accuracy

Inactive Publication Date: 2017-08-11
CHONGQING UNIV
View PDF2 Cites 14 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Insufficiency: the original RSS data has a lot of noise and the dimension is high, and the direct application has a great impact on the accuracy of subsequent matching and positioning, and the calculation complexity is very high. In the process of feature selection of the original RSS data using traditional PCA, Although it has the effect of reducing noise and dimensionality on the data
However, due to human subjective factors, during the feature selection process, some deep-level data features between the original data may be filtered out or lost, thereby affecting the accuracy and reliability of subsequent position matching, resulting in inaccurate positioning, and PCA is only good for linear data. For such a complex indoor environment, RSS data is nonlinear due to the influence of occluders or walls. Traditional PCA is not suitable.

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Deep-learning-based Wi-Fi indoor positioning method
  • Deep-learning-based Wi-Fi indoor positioning method
  • Deep-learning-based Wi-Fi indoor positioning method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0046] A Wi-Fi indoor positioning method based on deep learning:

[0047] Build the fingerprint library:

[0048] 1) Sample RSS data collection.

[0049] 1-1) Deploy n Wi-Fi wireless access points (APs) in the indoor area where positioning is required.

[0050] 1-2) Divide the indoor area into m domain blocks, where m is a natural number.

[0051] Take the center of each domain block as the sampling point P k , measure the received signal strength RSS of each APs deployed in step 1-1), each sampling point P kThey all correspond to a position coordinate, and the n pieces of RSS data corresponding to a position coordinate form a row of the multidimensional data sample set R after sorting. k=1, 2...m.

[0052] 2) Data preprocessing, completion and normalization.

[0053] 2-1) Perform missing zero padding on the multidimensional data sample set R obtained in step 1).

[0054] 2-1) Normalization processing to obtain the training set H.

[0055] 3) Deep feature learning.

...

Embodiment 2

[0077] In this embodiment, the method described in Embodiment 1 is applied to the positioning of parking spaces in a large complex indoor parking garage.

[0078] It is worth noting that, for large indoor parking garages, finding vehicles has always been a difficult problem to solve. In the prior art (as attached Image 6 As shown), a camera and a license plate recognition system are installed in each parking space to record the parking position of the vehicle. Although this method solves the above-mentioned problems, the cost of installing the camera is very high, and the later maintenance cost and reliability of the camera are not ideal.

[0079] This embodiment discloses a method for finding a vehicle parking position:

[0080] Build the fingerprint library:

[0081] 1) Sample RSS data collection.

[0082] 1-1) If figure 1 As shown, n Wi-Fi wireless access points (APs) are deployed in the indoor garage.

[0083] 1-2) Divide the indoor area into m domain blocks, where...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a deep-learning-based Wi-Fi indoor positioning method, so that a problem of positioning of people in indoor environments like an indoor garage of a large commercial complex can be solved. With an automatic coding algorithm, depth characteristics are extracted in a non-supervision manner form RSS big data; on the basis of the depth characteristics, a fingerprint base is established, and the fingerprint base is divided into a plurality of areas according to an original data position; and matching positioning is carried out on a to-be-localized position by using a K-nearest neighbors (KNN) method being common among matching algorithms.

Description

technical field [0001] The invention relates to the technical field of indoor positioning and can be applied to indoor positioning of large commercial complexes. Background technique [0002] In the existing Wi-Fi indoor positioning methods, most of them directly measure the received signal strength (RSS) of different wireless access points (APs) at the reference point to establish a fingerprint database or through traditional principal component analysis ( After PCA) processing, the fingerprint library is established, and then the fingerprint library is divided according to a certain method to reduce the complexity of subsequent matching, and finally used for matching and positioning of the points to be located. Insufficiency: the original RSS data has a lot of noise and the dimension is high, and the direct application has a great impact on the accuracy of subsequent matching and positioning, and the calculation complexity is very high. In the process of feature selection ...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
IPC IPC(8): G01S5/02G06N3/08
CPCG01S5/0278G06N3/08
Inventor 王楷熊庆宇孙国坦马龙昆余星姚政赵友金
Owner CHONGQING UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com