Indoor positioning method based on machine learning

Abstract

The invention discloses an indoor positioning method based on machine learning, which ensures that the system can acquire enough line-of-sight data by setting a large number of base stations. The identification and discarding of non-line-of-sight data can greatly improve the accuracy of positioning algorithms. The invention uses a machine learning algorithm described by non-parameters to learn through empirical data without determining any specific model, and the established model can cope with various scenarios that generate errors. The training samples contain various situations in various locations, so that when the environment changes, these models can still adapt to these changes and play a role.

Description

technical field [0001] The invention relates to the field of indoor wireless positioning, in particular to an indoor positioning method based on machine learning. Background technique [0002] A common indoor positioning method is a positioning algorithm based on distance measurement: the coordinates of the target point are calculated by measuring the distance between the target point and each known base station. The ranging schemes include: based on Time of Arrival (TOA, Time of Arrival), based on Received Signal Strength (RSS, Received Signal Strength), based on Time Difference of Arrival (TDOA, Time Difference of Arrival) and using channel impulse response information (such as average excess delay, maximum excess delay, RMS delay spread, rise time and kurtosis, etc.) to estimate distance. Time of arrival and received signal strength, as the most commonly used ranging methods, have their own defects: ranging based on time of arrival requires clock synchronization between ...

AI Technical Summary

Problems solved by technology

Time of arrival and received signal strength, as the most commonly used ranging methods, have their own defects: ranging based on time of arrival requires clock synchronization between the base station and users, but when the number of users is huge, this synchronization becomes quite Complex; the signal strength obtained by different devices is different, and due to the influence of the environment, the ranging accuracy based on the received signal strength is not high

However, there are problems in obtaining a large amount of channel impulse response information, because of cost constraints, most indoor positioning devices cannot meet this requirement

There are many other methods like this, but the realization of these methods is greatly restricted, so the non-line-of-sight situation is still a serious problem in indoor positioning

Images