Research on model fusion algorithm based on CSI fingerprint localization

Xin Wang; Guifen Chen; Guangjiao Chen

doi:10.1117/12.2685752

15 August 2023 Research on model fusion algorithm based on CSI fingerprint localization

Xin Wang, Guifen Chen, Guangjiao Chen

Proceedings Volume 12719, Second International Conference on Electronic Information Technology (EIT 2023); 127194O (2023) https://doi.org/10.1117/12.2685752
Event: Second International Conference on Electronic Information Technology (EIT 2023), 2023, Wuhan, China

Abstract

High precision indoor positioning technology is the key research content in the field of communication and control. In order to reduce the influence of environmental factors on indoor positioning accuracy, a fingerprint location model fusion algorithm based on channel state information (CSI) was proposed. In the off-line stage, the pauta criterion is used to remove abnormal data, and then the CSI amplitude is filtered by Kalman to make the CSI amplitude have certain periodic characteristics. Then the phase of CSI is corrected by linear transformation, and the processed amplitude and phase are taken as the joint CSI fingerprint. Finally, CatBoost-KNN model fusion localization algorithm was used to train fingerprint data. In the online stage, the coordinates of the pre-processed points to be measured are predicted by the trained CatBoost- KNN model. Simulation results show that the average positioning error of this method is 0.96m. Compared with some existing positioning methods, the positioning accuracy of the improved method is improved, and the algorithm has good practicability in large-scale indoor positioning applications.

Citation Download Citation

Xin Wang, Guifen Chen, and Guangjiao Chen "Research on model fusion algorithm based on CSI fingerprint localization", Proc. SPIE 12719, Second International Conference on Electronic Information Technology (EIT 2023), 127194O (15 August 2023); https://doi.org/10.1117/12.2685752

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available