Robust ultra-wideband range error mitigation with deep learning at the edge

Ultra-wideband (UWB) is the state-of-the-art and most popular technology for wireless localization. Nevertheless, precise ranging localization in non-line-of-sight (NLoS) conditions still an open research topic. Indeed, multipath effects, reflections, refractions, complexity of indoor radio environment can easily introduce a positive bias measurement, resulting highly inaccurate unsatisfactory position estimation. This article proposes efficient representation learning methodology that exploits latest advancement deep graph optimization techniques to achieve effective error mitigation at edge. Channel Impulse Response (CIR) signals are directly exploited extract high semantic features estimate corrections either NLoS or LoS conditions. Extensive experimentation with different settings configurations has proved effectiveness our demonstrated feasibility robust low computational power UWB range mitigation.