Passive indoor positioning method and device

Abstract

The invention discloses a passive indoor positioning method, and the method comprises the steps: enabling the collected radio frequency network link RSS values, collected by a target at all coordinates in a non-empty room, as the training samples, and taking the coordinate numbers as the sample labels; carrying out the two-dimensional double-correlation distributed wavelet filtering of the training samples, and determines he filtered training samples; building an Adaboost.M2 integrated learning model based on a Geni decision tree, employing the filtered training samples and the sample labels for training, and determining the trained model; collecting the RSS values of the target when the target freely moves in a monitored region, taking the RSS values as test samples, carrying out the two-dimensional double-correlation distributed wavelet filtering of the test samples, inputting the filtered test samples into the trained model, and determining a positioning result. The method can distinguish and filter the RSS sample noises and random interference, and keeps the normal hop data. Meanwhile, a positioning process is very strong in generalization capability, and can improve the positioning accuracy and stability. The method also provides a corresponding device.

Description

technical field [0001] The invention relates to the technical field of wireless sensor network indoor personnel target positioning, in particular to the technical field of passive passive indoor target positioning based on radio frequency received signal strength in a non-empty and complex indoor environment. Background technique [0002] The indoor positioning method of human targets based on wireless radio frequency sensor network does not require the active cooperation and carrying of electronic tags by the located targets. , resulting in occlusion, that is, a shadow effect, which affects the received signal strength value (RSS) of network sensor nodes, and the location of the target can be judged by collecting and analyzing RSSI samples. Due to the characteristics of radio frequency signals, this method has the advantages of insensitivity to ambient temperature, humidity, light and non-metallic obstacles, convenient networking and low cost, so it has great potential in t...

AI Technical Summary

Problems solved by technology

[0007] The purpose of the present invention is to provide a passive passive indoor positioning method and device to solve the problem of filtering RSS samples in a more complex environment, that is, a non-open environment, such as an ordinary indoor environment, in the existing passive passive indoor positioning technology based on RSS. The effect is not good, at least one of the problems of low positioning accuracy and weak generalization ability of machine learning methods

Images