Radio-Localization in Underground Narrow-Vein Mines Using Neural Networks with In-built Tracking and Time Diversity

In the mining industry, knowing the position of miners and/or equipments is an important safety measure that reduces risks and improves the security of that facility. Being an indoor environment, wireless transmitted signals in underground narrow-vein mines suffer multiple kinds of distortions due to extreme multipath and non-line of sight (NLOS) conditions. One of the proposed solutions to accurate localization in such challenging environments is based on extracting the channel impulse response (CIR) of the received signal and using the fingerprinting technique combined with cooperative artificial neural networks (ANNs). Such localization systems use the spatial domain where the reference localizing units are implemented at different positions away from the transmitter. In this article, we introduce a localization technique that uses fingerprints successively recorded in time with in-built tracking as an alternative method to localize. Unlike the spatial-domain technique where cooperative localizing units collect memoryless fingerprints from different locations, this technique uses one localizing unit and is capable of estimating the position of a transmitter precisely using its current and previous registered fingerprints in time. Localization using time-domain fingerprinting (i.e., tracking) and ANNs is introduced as a new method that exploits time diversity and improves the accuracy, precision and scalability of the positioning system.