ICWLM: A Multi-Task Wireless Large Model via In-Context Learning

TL;DR

ICWLM introduces a wireless-native foundation model trained from scratch to jointly address multi-task physical-layer problems (precoding and channel prediction) using in-context learning. By avoiding cross-domain adaptation to pre-trained LLMs and employing a causal transformer with RoPE and SwiGLU, ICWLM demonstrates competitive performance to task-specific methods while generalizing to unseen configurations with few demonstrations. The approach integrates data formulation, model architecture, and a multi-task training schedule to enable rapid adaptation in dynamic wireless environments, reducing deployment complexity. Practical impact lies in unified intelligent resource management for future wireless networks, with potential extensions to additional tasks and hardware embodiment.

Abstract

The rapid evolution of wireless communication technologies, particularly massive multiple-input multiple-output (mMIMO) and millimeter-wave (mmWave), introduces significant network complexity and computational demands. Significant research efforts have been made to improve physical layer performance by resorting to deep learning (DL) methods, which, however, are usually task-specific and struggle with data scarcity and generalization. To address these challenges, we propose a novel In-Context Wireless Large Model (ICWLM), a wireless-native foundation model designed for simultaneous multi-task learning at the physical layer. Unlike conventional methods that adapt wireless data to pre-trained large language models (LLMs), ICWLM is trained directly on large-scale, mixed wireless datasets from scratch. It jointly solves multiple classical physical layer problems, including multi-user precoding (sum-rate maximization and max-min SINR) and channel prediction. A key innovation of ICWLM is its utilization of in-context learning (ICL), enabling the model to adapt to varying system configurations and channel conditions with minimal demonstration pairs, eliminating the need for extensive retraining. Extensive simulation results demonstrate that ICWLM achieves competitive performance compared to task-specific methods while exhibiting remarkable generalization capabilities to unseen system configurations. This work offers a promising paradigm for developing unified and adaptive AI models for future wireless networks, potentially reducing deployment complexity and enhancing intelligent resource management.

Figures (14)

• Figure 1: Comparison of different wireless large models.
• Figure 2: Illustration of In-Context Learning (ICL) mechanism.
• Figure 3: Schematic diagram of multi-user downlink transmission system.
• Figure 4: UPA antenna configuration in 3D-Cartesian coordinate system.
• Figure 5: The proposed ICL framework for multi-task wireless large model.