A Wireless Foundation Model for Multi-Task Prediction

TL;DR

This paper introduces a unified wireless foundation model that leverages a causal Transformer backbone and novel training strategies to improve multi-task prediction accuracy and generalization in dynamic mobile networks.

Contribution

The paper presents a novel multi-task wireless prediction model with univariate decomposition, interval encoding, and patch masking, enabling zero-shot learning and better generalization.

Findings

Strong generalization to unseen scenarios

Zero-shot performance on new tasks surpassing baselines

Effective multi-task prediction across diverse intervals

Abstract

With the growing complexity and dynamics of the mobile communication networks, accurately predicting key system parameters, such as channel state information (CSI), user location, and network traffic, has become essential for a wide range of physical (PHY)-layer and medium access control (MAC)-layer tasks. Although traditional deep learning (DL)-based methods have been widely applied to such prediction tasks, they often struggle to generalize across different scenarios and tasks. In response, we propose a unified foundation model for multi-task prediction in wireless networks that supports diverse prediction intervals. The proposed model enforces univariate decomposition to unify heterogeneous tasks, encodes granularity for interval awareness, and uses a causal Transformer backbone for accurate predictions. Additionally, we introduce a patch masking strategy during training to support arbitrary input lengths. After trained on large-scale datasets, the proposed foundation model demonstrates strong generalization to unseen scenarios and achieves zero-shot performance on new tasks that surpass traditional full-shot baselines.