Millimeter wave (mmWave) communications are an essential component of 5G-and-beyond ultra-dense Gbit/s wireless networks, but also pose significant challenges related to the communication environment.
Especially beam-training and tracking, device association, and fast handovers for highly directional mmWave links may potentially incur a high overhead. At the same time, such mechanisms would benefit greatly from accurate knowledge about the environment and device locations that can be provided through simultaneous localization and mapping (SLAM) algorithms.
In this paper, we tackle the above issues by proposing CLAM, a distributed mmWave SLAM algorithm that works with no initial information about the network deployment or the environment, and achieves low computational complexity thanks to a fundamental reformulation of the angle-differences-of-arrival mmWave anchor location estimation problem.
All information required by CLAM is collected by a mmWave device thanks to beam training and tracking mechanisms inherent to mmWave networks, at no additional overhead. Our results show that CLAM achieves submeter accuracy in the great majority of cases. These results are validated via an extensive experimental measurement campaign carried out with 60-GHz mmWave hardware.
About Joan Palacios Beltrán
Joan Palacios finished his Bachelor degree of mathematics in 2015 from Univesrsitat de València in Spain. He joined IMDEA on September 2015 as a PhD student. He does not like long bios.
This event will be conducted in English