Indoor fingerprint positioning direction based on deep learning

Abstract

In the modern technological innovation and development, although the global positioning system proposed by researchers can provide positioning information for wireless sensors, the application effect in indoor conditions does not meet the expected requirements. Therefore, someone proposed a Wi-Fi indoor positioning model with deep negative correlation learning as the core. Nowadays, with the comprehensive popularity of Wi-Fi technology, a large number of indoor positioning systems based on Wi-Fi signal strength have become the focus of attention in the market. Therefore, on the basis of understanding the concept of deep learning algorithm and negative correlation learning, this paper mainly studies the Wi-Fi positioning model with deep complex correlation learning as the core, so as to provide an effective basis for indoor fingerprint positioning direction. The final experimental results prove that this model can apply the negative correlation learning method to the regression predictor and denoising autoencoder, so that the deep learning method can adapt to the signals that follow the environment and time changes faster, and improve the effectiveness of the overall indoor positioning.