Intelligent wireless positioning system and method using large-scale array antenna

Abstract

The invention provides an intelligent wireless positioning system and method using a large-scale array antenna. And the base station side extracts the angle time delay domain channel intensity matrixfrom the uplink channel state information as position fingerprint information. In an offline stage, a base station side utilizes position fingerprint information to train a cascade convolutional neural network, and the trained network structure, weight value and activation function related to neurons of each convolutional neural network and the position fingerprint information of feature points ina cell coverage range are jointly stored in a database; in the online stage, the base station side inputs the position fingerprint information corresponding to the user side mobile terminal device into the cascade convolutional neural network, calculates the position coordinates of the user side and feeds back the position coordinates to the corresponding user side mobile terminal. The method candepend on wireless communication equipment, the positioning precision of the mobile terminal in complex scatterer environments such as urban buildings and indoor environments is remarkably improved,and the time cost and the storage cost of a positioning system are reduced.

Description

technical field [0001] The present invention relates to an intelligent wireless positioning system and method using large-scale array antennas, and in particular to a wireless positioning system and method utilizing convolutional neural network-based channel characteristics in the angular delay domain of a wideband massive MIMO system. Background technique [0002] With the rapid development of smart mobile terminals and applications and services based on geographic location information, people's requirements for the accuracy of wireless positioning of mobile terminals are increasing. Most positioning requirements are generated in complex scatterer environments such as urban buildings and indoors. Wireless signals usually propagate through multiple non-line-of-sight paths, which greatly affects the accuracy of wireless positioning. The wireless positioning method based on location fingerprint information can fully exploit the influence of the scatterer environment on the cha...

AI Technical Summary

Problems solved by technology

[0004] When extracting angular delay domain channel features as location fingerprint information in wideband massive MIMO systems, most of the existing wireless positioning methods based on location fingerprint information will generate large time and storage overhead.

Images