Robust Networked Federated Learning for Localization

TL;DR

This paper introduces a robust federated localization method using an $L_1$-norm formulation within a distributed sub-gradient framework, effectively handling outliers and non-convexity to improve accuracy and convergence.

Contribution

It presents a novel $L_1$-norm robust approach for federated localization that directly addresses non-convexity and outliers without approximations, enhancing efficiency and accuracy.

Findings

Outperforms existing localization methods in outlier-rich environments.

Converges to a stationary point demonstrating reliability.

Shows superior accuracy and computational efficiency in simulations.

Abstract

This paper addresses the problem of localization, which is inherently non-convex and non-smooth in a federated setting where the data is distributed across a multitude of devices. Due to the decentralized nature of federated environments, distributed learning becomes essential for scalability and adaptability. Moreover, these environments are often plagued by outlier data, which presents substantial challenges to conventional methods, particularly in maintaining estimation accuracy and ensuring algorithm convergence. To mitigate these challenges, we propose a method that adopts an $L_1$-norm robust formulation within a distributed sub-gradient framework, explicitly designed to handle these obstacles. Our approach addresses the problem in its original form, without resorting to iterative simplifications or approximations, resulting in enhanced computational efficiency and improved estimation accuracy. We demonstrate that our method converges to a stationary point, highlighting its effectiveness and reliability. Through numerical simulations, we confirm the superior performance of our approach, notably in outlier-rich environments, which surpasses existing state-of-the-art localization methods.