Outdoor fingerprint positioning method using CSI multipath and machine learning

Abstract

The invention discloses an outdoor fingerprint positioning method using CSI multipath and machine learning. The method comprises the steps: a receiving end acquires a plurality of distinguishable multipath signals in a cell; the multipath signals are subjected to data collection and pre-processed to obtain offline multipath CSI data and the offline multipath CSI data are grouped and numbered; thegrouped multipath CSI data are trained hierarchically in the offline phase, so that the mean square error between the training tag and the output of the network is minimized; a softmax regression classifier is used for regressing and classifying the data after training and a fingerprint library is built to complete the offline phase training; and after CSI information is received from an unknown location user, a CSI signal passes through a neural network forward propagation and regression classifier, the output of the classifier is classified by a KNN algorithm, and the K positions with the highest probability are selected to perform the weighted average calculation to obtain the position of the user. The invention effectively improves the accuracy of outdoor positioning, saves time and labor, has high efficiency and wide application range.

Description

technical field [0001] The invention belongs to the technical field of communication, and in particular relates to an outdoor fingerprint positioning method using CSI multipath and machine learning. Background technique [0002] With the sharp increase in demand for location-based service applications, precise location technology has attracted widespread attention. In the outdoor environment, GPS still occupies a dominant position and can obtain better positioning accuracy in most application scenarios. However, it is well known that GPS cannot be applied to indoor positioning, and there is an urban canyon effect in urban areas with dense buildings and buildings, which limits the application range of GPS to a certain extent. In addition, there are GSM cellular network positioning and wireless positioning technologies. The former has a wider range of applications and can be used in both indoor and outdoor environments, but the positioning accuracy is limited by the radius of...

AI Technical Summary

Problems solved by technology

In outdoor fingerprint positioning, RSS and its combination form are the most commonly used fingerprint features because of their simplicity and ease of measurement. However, because this is a rough description of the signal, the information is not fully utilized, and it is easily affected by shadow fading. , sometimes the positioning accuracy is low

In addition, different feature quantities are required in different scenarios. The selection of feature quantities determines the positioning performance, and manual design and selection of feature quantities is time-consuming, laborious and inefficient.

Images