We propose a Multi-Armed Bandit algorithm for mmWave beam steering that approaches the performance of state-of-the-art Bayesian algorithms at a fraction of the complexity and without requiring Channel State Information. The algorithm, called Hierarchical Optimal Sampling of Unimodal Bandits, simultaneously exploits the benefits of hierarchical codebooks and the approximate unimodality of rewards to achieve fast beam steering, in a sense that we precisely define to provide fair comparison with existing algorithms. Extensive simulations over slow fading channels demonstrate the appealing performance versus complexity trade-off struck by the algorithm across a wide range of Signal-to-Noise Ratios.
@inproceedings{Blinn2021,
author = {Blinn, Nathan and Boerger, Jana and Bloch, Matthieu R},
booktitle = {IEEE International Conference on Communications},
title = {mmWave Beam Steering with Hierarchical Optimal Sampling for Unimodal Bandits},
year = {2021},
month = jun,
pages = {1-6},
doi = {10.1109/ICC42927.2021.9500373},
file = {:2021-Blinn-ICC-mmWave_Beam_Steering_with_Hierarchical_Optimal_Sampling_for_Unimodal_Bandits.pdf:PDF},
groups = {mmWave},
howpublished = {accepted to the \emph{IEEE 2021 IEEE International Conference on Communications}}
}