Radio SLAM for 6G Systems at THz Frequencies: Design and Experimental Validation

Abstract

Next-generation wireless networks will see the convergence of communication and sensing, also exploiting the availability of large bandwidths in the THz spectrum and electrically large antenna arrays on handheld devices. In particular, it is envisaged that user devices will be able to automatically scan their surroundings by steering a very narrow antenna beam and collecting echoes reflected by objects and walls. These data will be utilized to derive a map of the surrounding indoor environment and infer users' trajectories using simultaneous localization and mapping (SLAM) techniques. In this article, we address this scenario by proposing original radio-SLAM (R-SLAM) algorithms, derived from image processing techniques, to map the environment and pinpoint the device position in the map starting from measurements sensed by a mobile THz radar. Initially, to fully understand the THz backscattering phenomenon, we provide an experimental characterization of the THz backscattering channel in indoor environments. Then, the performance of the proposed algorithms is assessed using real-world THz radar measurements and is compared with state-of-the-art SLAM techniques, demonstrating the superiority of the proposed approaches.