Radiance-Field Reinforced Pretraining: Scaling Localization Models with Unlabeled Wireless Signals

TL;DR

This paper introduces RFRP, a self-supervised pretraining framework that leverages unlabeled RF data and a neural radiance field to significantly improve indoor localization accuracy across diverse scenes.

Contribution

The paper presents a novel self-supervised pretraining method coupling a localization model with RF-NeRF for better cross-scene generalization using large-scale unlabeled RF data.

Findings

RFRP reduces localization error by over 40% compared to non-pretrained models.

Pretraining with RFRP outperforms supervised pretraining by 21%.

The approach effectively learns representations from over 7 million RF samples.

Abstract

Radio frequency (RF)-based indoor localization offers significant promise for applications such as indoor navigation, augmented reality, and pervasive computing. While deep learning has greatly enhanced localization accuracy and robustness, existing localization models still face major challenges in cross-scene generalization due to their reliance on scene-specific labeled data. To address this, we introduce Radiance-Field Reinforced Pretraining (RFRP). This novel self-supervised pretraining framework couples a large localization model (LM) with a neural radio-frequency radiance field (RF-NeRF) in an asymmetrical autoencoder architecture. In this design, the LM encodes received RF spectra into latent, position-relevant representations, while the RF-NeRF decodes them to reconstruct the original spectra. This alignment between input and output enables effective representation learning using large-scale, unlabeled RF data, which can be collected continuously with minimal effort. To this end, we collected RF samples at 7,327,321 positions across 100 diverse scenes using four common wireless technologies--RFID, BLE, WiFi, and IIoT. Data from 75 scenes were used for training, and the remaining 25 for evaluation. Experimental results show that the RFRP-pretrained LM reduces localization error by over 40% compared to non-pretrained models and by 21% compared to those pretrained using supervised learning.