Adaptive 3D-RoPE: Physics-Aligned Rotary Positional Encoding for Wireless Foundation Models

TL;DR

This paper introduces Adaptive 3D-RoPE, a physics-aligned rotary positional encoding for wireless foundation models, improving generalization and extrapolation in channel state information tasks by incorporating explicit 3D physics.

Contribution

It proposes a novel adaptive, physics-aligned 3D rotary positional encoding that dynamically modulates positional priors based on wireless channel physics, enhancing model generalization.

Findings

Achieves up to 10.7 dB NMSE reduction in antenna scale extrapolation.

Improves zero-shot NMSE by 1.07 dB across unseen mobility scenarios.

Demonstrates superior performance across 100 datasets in scale extrapolation and zero-shot tasks.

Abstract

Positional encoding plays a pivotal role in determin?ing the extrapolation and generalization performance of wireless foundation models for channel state information (CSI) modeling, latent characterization, and task-specific prediction. However, existing CSI models inherit static or one-dimensional positional priors from natural language and vision architectures, which fundamentally misalign with the intrinsic physics of wireless channels by lacking explicit relative decay, collapsing the 3D spatio-temporal-frequency structure, and remaining scenario?rigid. This paper proposes Adaptive 3D-RoPE, a physics-aligned rotary positional encoding that establishes the structural corner?stone for wireless foundation models. The framework integrates a learnable, axis-decoupled 3D frequency bank to explicitly disentangle multi-dimensional phase dependencies, coupled with a lightweight channel-conditioned controller that dynamically modulates the prior via compact global CSI descriptors. This sample-adaptive mechanism transforms positional encoding from a static transformer component into a dynamic, coherence-aware inductive bias to resolve heterogeneous channel physics. Extensive experiments across 100 datasets demonstrate the superiority of the proposed scheme in both scale extrapolation and zero-shot generalization. Compared to the state-of-the-art, our method achieves up to a 10.7 dB reduction in normalized mean square error (NMSE) under 8 times antenna scale extrapolation. Given the same CSI input scales, our method can also improve zero-shot NMSE by 1.07 dB across unseen mobility scenarios and 0.90 dB in low-frequency-to-millimeter-wave tasks.