Generating Light-based Fingerprints for Indoor Localization

2025-08-04

AI Summary

Radio-frequency indoor positioning systems (e.g., Wi-Fi, UWB) suffer from multipath fading, interference, and coverage instability. To address these limitations, this paper proposes a low-cost, high-robustness visible light communication (VLC)-based localization method leveraging spectral fingerprints. We employ the AS7341 spectral sensor to capture ambient spectral features and design a two-stage localization framework: first, TabGAN synthesizes high-fidelity spectral data to mitigate scarcity of real-world measurements; second, a multilayer perceptron (MLP) learns the spectral-to-position mapping. This data augmentation strategy increases data acquisition cost by only 5%, yet substantially expands the effective training set. Experimental evaluation in a U-shaped laboratory with 42 reference points demonstrates that the mean localization error decreases significantly—from 62.9 cm to 49.3 cm (a 20% reduction)—validating both the stability of spectral fingerprints and the generalizability of the proposed method.

Technology Category

Intelligent Robots: Localization, Mapping, and NavigationPlanning, Routing, and Scheduling: Optimization of Spatio-temporal SystemsMachine Learning: Multi-instance/Multi-view Learning

Application Category

Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applicationsSecurity and Privacy: Large-scale security measurementsSearch and Retrieval-Augmented AI: Web evaluation methodologies and metrics

Abstract

Accurate indoor localization underpins applications ranging from wayfinding and emergency response to asset tracking and smart-building services. Radio-frequency solutions (e.g. Wi-Fi, RFID, UWB) are widely adopted but remain vulnerable to multipath fading, interference, and uncontrollable coverage variation. We explore an orthogonal modality -- visible light communication (VLC) -- and demonstrate that the spectral signatures captured by a low-cost AS7341 sensor can serve as robust location fingerprints. We introduce a two-stage framework that (i) trains a multi-layer perceptron (MLP) on real spectral measurements and (ii) enlarges the training corpus with synthetic samples produced by TabGAN. The augmented dataset reduces the mean localization error from 62.9cm to 49.3cm -- a 20% improvement -- while requiring only 5% additional data-collection effort. Experimental results obtained on 42 reference points in a U-shaped laboratory confirm that GAN-based augmentation mitigates data-scarcity issues and enhances generalization.

Problem

• Develops light-based fingerprints for indoor localization accuracy
• Addresses limitations of radio-frequency solutions like multipath fading
• Uses GAN-augmented data to reduce localization error significantly

Innovation

• Uses visible light communication for localization
• Employs MLP and TabGAN for data augmentation
• Reduces error with synthetic spectral samples