Reinforcement Learning for Efficient and Tuning-Free Link Adaptation

TL;DR

This paper introduces a latent Thompson sampling algorithm for wireless link adaptation that automatically learns optimal transmission parameters, significantly improving throughput without manual tuning in dynamic fading channels.

Contribution

The paper proposes a novel latent learning model and a tuning-free LTS algorithm for reinforcement learning-based link adaptation, enhancing performance in fading wireless channels.

Findings

LTS improves link throughput by up to 100% over existing algorithms.

The method automatically tracks channel dynamics without manual parameter tuning.

Latent learning exploits parameter correlations for better adaptation.

Abstract

Wireless links adapt the data transmission parameters to the dynamic channel state -- this is called link adaptation. Classical link adaptation relies on tuning parameters that are challenging to configure for optimal link performance. Recently, reinforcement learning has been proposed to automate link adaptation, where the transmission parameters are modeled as discrete arms of a multi-armed bandit. In this context, we propose a latent learning model for link adaptation that exploits the correlation between data transmission parameters. Further, motivated by the recent success of Thompson sampling for multi-armed bandit problems, we propose a latent Thompson sampling (LTS) algorithm that quickly learns the optimal parameters for a given channel state. We extend LTS to fading wireless channels through a tuning-free mechanism that automatically tracks the channel dynamics. In numerical evaluations with fading wireless channels, LTS improves the link throughout by up to 100% compared to the state-of-the-art link adaptation algorithms.