Self-Evolving In-Context Learning for Direct Pilot-to-Beamformer Design in MU-MISO Systems

2026-07-12

AI Summary

This work addresses the poor generalization, frequent retraining requirements, and sensitivity to channel mismatch exhibited by pilot-based beamforming in MU-MISO systems under diverse channel models. To overcome these limitations, the authors propose an end-to-end framework that integrates an ICL-Transformer with an encoder-decoder network. Leveraging a curriculum learning strategy, a self-evolving context mechanism, and mismatch-aware modeling, the framework enables a smooth transition from supervised imitation to unsupervised optimization while supporting dynamic context construction. Notably, the method achieves rapid adaptation to both seen and unseen channel models without requiring gradient updates. Extensive evaluations demonstrate its significant superiority over WMMSE and existing Transformer-based approaches across various communication scenarios, effectively mitigating channel mismatch issues.

Abstract

We develop an enhanced in-context learning (ICL) framework to improve the performance of pilot-based beamforming in multi-user multiple-input single-output (MU-MISO) systems. The proposed scheme integrates the ICL-Transformer backbone with the pilot encoder-decoder network (EDN) and the beamformer EDN. A crucial feature of our ICL network is that it can handle multiple channel models without retraining, enabled by the construction of model-specific context datasets. To improve convergence and robustness, we introduce three key innovations: (a) a curriculum learning (CL) strategy that smoothly transitions from supervised LMMSE-labeled imitation to unsupervised sum-rate maximization, (b) a self-evolving mechanism that dynamically expands and refines the context datasets for all channel models during CL-based training, and (c) a mismatch-aware extension that incorporates several mismatches into the general ICL framework and bypasses explicit channel calibrations. Ablation studies validate the effectiveness of the in-context architecture and enhanced training strategies. Simulation results over diverse communication environments show that the proposed scheme is able to rapidly adapt to both seen and unseen channel models without gradient-based parameter updates, and can mitigate the mismatch issues via intelligent context constructions. Furthermore, our scheme consistently outperforms the existing beamforming schemes under pilot-based settings, including the WMMSE benchmark and the recent Transformer-based methods.

Problem

• MU-MISO
• beamforming
• channel mismatch
• pilot-based design
• in-context learning

Innovation

• in-context learning
• self-evolving mechanism
• curriculum learning
• beamformer design
• mismatch-aware