Unlock instant, AI-driven research and patent intelligence for your innovation.

Complex environment indoor fingerprint positioning method based on integrated federated learning

A complex environment and fingerprint positioning technology, applied in integrated learning, services based on specific environments, services based on location information, etc., can solve problems such as inability to achieve positioning effects, and achieve the effect of improving iterative convergence speed and accuracy

Active Publication Date: 2022-03-18
JIANGSU INSTITUTE OF EDUCATION
View PDF7 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

In the face of complex indoor environments, the result of forced integration of pure federated learning into a model obviously cannot achieve a more accurate positioning effect

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
  • Complex environment indoor fingerprint positioning method based on integrated federated learning
  • Complex environment indoor fingerprint positioning method based on integrated federated learning
  • Complex environment indoor fingerprint positioning method based on integrated federated learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0027] According to the following figure 1 The flow chart shown further elaborates the technical solution of the complex environment indoor fingerprint positioning method based on integrated federated learning in the present invention.

[0028] The indoor fingerprint positioning method in a complex environment, first of all, arranges the transmitter indoors, collects a large number of RSS data information of fingerprint points according to the grid distribution, divides simple physical partitions, and fits the data in each partition to obtain a regional characteristic. The preliminary model is uploaded to the centralized server. The server classifies the models according to the similarity of each area and divides them into several different groups. Each group performs federated learning internally and generates a positioning model that meets the environment through training. Finally, a multi-model positioning criterion is generated.

[0029] The fingerprint library is establi...

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 complex environment indoor fingerprint positioning method based on integrated federated learning, and the method comprises the steps: carrying out the rough region division according to the difference of the environment, carrying out the fitting of the collected data in each region, uploading the weight vector of a model after fitting, and carrying out the positioning of the indoor fingerprint through employing an integrated federated learning method. According to the similarity of the models, the similar models are divided into one group according to an adopted grouping method to form groups for various environments, federal aggregation is adopted for the models in the groups, new users and changing users are re-grouped in time, low data isomerism in the groups is ensured, the models are optimized by using a limit gradient lifting algorithm, and the user experience is improved. And meanwhile, the influence of model optimization direction offset caused by model transmission speed is eliminated, and finally, the multi-model fingerprint positioning system suitable for a complex and changeable environment is obtained.

Description

technical field [0001] The invention relates to an indoor fingerprint positioning method in a complex environment based on integrated federated learning, and belongs to the technical field of indoor positioning. Background technique [0002] With the development of science and technology, location-based services play a pivotal role in modern society. Whether it is rescue services, medical monitoring, commercial promotion or daily life, location-based services have a wide range of applications. For example, in the recent control of virus spread, positioning technology has played a huge role. Due to the current mature satellite positioning technology, the location information can be obtained through the mobile phone in basically any environment. However, accurate positioning is based on the premise of outdoor. Due to the problem of signal propagation, when the satellite signal reaches the ground, the signal is weak and the penetration is poor, so the indoor environment may n...

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
Patent Type & Authority Applications(China)
IPC IPC(8): H04W64/00H04W4/021H04W4/33G06N20/20G06K9/62
CPCH04W64/006H04W4/021H04W4/33G06N20/20G06F18/22G06F18/23213
Inventor 赵洁倪艺洋孔志鹏朱辉生王玉玺
Owner JIANGSU INSTITUTE OF EDUCATION