SimDeep: Federated 3D Indoor Localization via Similarity-Aware Aggregation

2025-08-02

AI Summary

To address degraded 3D indoor localization accuracy caused by non-independent and identically distributed (non-IID) data and device heterogeneity in complex indoor environments, this paper proposes a privacy-preserving federated learning-based indoor localization framework. The core methodological contribution is a similarity-aware dynamic aggregation strategy that adaptively selects and weights client model updates based on feature-space similarity, thereby mitigating bias induced by non-IID data. The framework employs a lightweight deep neural network and enables collaborative training across heterogeneous devices without sharing raw data. Experimental evaluation on real-world deployments demonstrates a 3D localization accuracy of 92.89%, significantly outperforming both conventional federated and centralized baselines. The approach achieves high accuracy, strong generalization across diverse devices and environments, and practical deployability—making it suitable for real-world privacy-sensitive indoor positioning systems.

Technology Category

• Machine Learning: Distributed Machine Learning & Federated Learning
• Intelligent Robots: Localization, Mapping, and Navigation
• Search and Optimization: Learning to Search

Application Category

• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Systems and Infrastructure for Web, Mobile and WoT: Federated Web and WoT systems, including distributed, federated and edge-based data processing
• Security and Privacy: Large-scale security measurements

Abstract

Indoor localization plays a pivotal role in supporting a wide array of location-based services, including navigation, security, and context-aware computing within intricate indoor environments. Despite considerable advancements, deploying indoor localization systems in real-world scenarios remains challenging, largely because of non-independent and identically distributed (non-IID) data and device heterogeneity. In response, we propose SimDeep, a novel Federated Learning (FL) framework explicitly crafted to overcome these obstacles and effectively manage device heterogeneity. SimDeep incorporates a Similarity Aggregation Strategy, which aggregates client model updates based on data similarity, significantly alleviating the issues posed by non-IID data. Our experimental evaluations indicate that SimDeep achieves an impressive accuracy of 92.89%, surpassing traditional federated and centralized techniques, thus underscoring its viability for real-world deployment.

Problem

• Addresses non-IID data in federated indoor localization
• Manages device heterogeneity in 3D localization systems
• Improves accuracy via similarity-aware model aggregation

Innovation

• Federated Learning for 3D indoor localization
• Similarity-Aware Aggregation strategy
• Handles non-IID data and device heterogeneity