The focus of this paper is on the challenge of localizing data in a federated setting where the data is spread across multiple devices. This problem is complex due to the decentralized nature of federated environments and the presence of outlier data, which hinders conventional methods in maintaining accuracy and convergence. To address these challenges, we propose a method that utilizes an $L_1$-norm robust formulation within a distributed sub-gradient framework. This approach is specifically designed to handle outlier data without relying on simplifications or approximations, leading to improved computational efficiency and estimation accuracy. We demonstrate through numerical simulations that our method converges to a stationary point, showcasing its effectiveness and reliability. Our approach also outperforms existing localization methods, particularly in outlier-rich environments.
Enhancing Localization with Resilient Networked Federated Learning
