An Indoor Localization Method Based on Convolutional Neural Network

Abstract

The invention discloses an indoor positioning method based on a convolutional neural network: Step 1, obtain a phase information matrix; Step 2, calibrate the phase information matrix; Step 3, convert all phase information in the final phase information matrix and each phase The matrix coordinates corresponding to the information are used as the training set, and input the constructed channel detection model to obtain the trained channel detection model; Step 4: Input the phase information of the single channel to be detected into the trained channel detection model; Step 5: According to The channel information CH and the initial position PH obtain complete phase information; Step 6: According to the complete phase information, obtain its corresponding likelihood function estimator, and take the positioning area with the largest likelihood function estimator as the target position. The present invention uses a deep learning method to obtain the position of the channel to be measured, and locates the target according to the virtual high-bandwidth phase information obtained at the position, thereby effectively improving the positioning accuracy.

Description

technical field [0001] The invention belongs to the technical field of positioning, and relates to an indoor positioning method based on a convolutional neural network. Background technique [0002] In recent years, with the popularity of mobile devices such as smartphones, tablets, and laptops, location-based applications based on these devices have become an important part of our daily lives. At present, most mobile devices mainly use GPS or Beidou positioning system to provide positioning services, and these two solutions can only effectively locate targets outdoors, and have low accuracy or even fail to work in complex indoor environments. The main reasons are as follows: (1) Since there are many objects between the indoor access point AP and the mobile client, wireless signals are reflected to multiple paths, which is called multipath phenomenon; (2) ) GPS provides positioning accuracy between a few meters, which is more than enough for streets or city blocks in outdoo...

AI Technical Summary

Problems solved by technology

3) Accuracy: The accuracy of positioning should be accurate in the indoor environment. Since some existing positioning systems have increased hardware deployment and quantity, although the positioning accuracy is improved, the above two conditions are not met

Images