Beam Prediction based on Large Language Models

TL;DR

This paper introduces a novel beam prediction method for mmWave communications using large language models, transforming the problem into a text-based forecasting task and leveraging prompt techniques for improved accuracy and robustness.

Contribution

The paper presents a new LLM-based approach for beam prediction in wireless systems, integrating time series forecasting with text representations and prompt techniques.

Findings

Outperforms traditional models in accuracy

Demonstrates robustness in predictions

Highlights potential of LLMs in wireless communication

Abstract

In this letter, we use large language models (LLMs) to develop a high-performing and robust beam prediction method. We formulate the millimeter wave (mmWave) beam prediction problem as a time series forecasting task, where the historical observations are aggregated through cross-variable attention and then transformed into text-based representations using a trainable tokenizer. By leveraging the prompt-as-prefix (PaP) technique for contextual enrichment, our method harnesses the power of LLMs to predict future optimal beams. Simulation results demonstrate that our LLM-based approach outperforms traditional learning-based models in prediction accuracy as well as robustness, highlighting the significant potential of LLMs in enhancing wireless communication systems.