We introduce fast millimeter-wave base station (BS) and its antenna sector selection for user equipment based on its location. Using a conditional random field inference model with specially designed parameters, which are robust to change of environment, InferBeam allows the use of measurement samples on best beam selection at a small number of locations to infer the rest dynamically. Compared to beam-sweeping based approaches in the literature, InferBeam can drastically reduce the setup cost for beam alignment for a new environment, and also the latency in acquiring a new beam under intermittent blockage. We have evaluated InferBeam using a discrete event simulation. Our results indicate that the system can make best beam selection for 98% of locations in test environments comprising smallsized apartment or office spaces, while sampling fewer than 1% of locations. InferBeam is a complete protocol for best beam inference that can be integrated into millimeter-wave standards for accelerating the much-needed fast and economic beam alignment capability.
@article{arxiv.1802.03373,
title = {InferBeam: A Fast Beam Alignment Protocol for Millimeter-wave Networking},
author = {Sai Qian Zhang and H. T. Kung and Youngjune Gwon},
journal= {arXiv preprint arXiv:1802.03373},
year = {2018}
}