@INPROCEEDINGS{Paol2410:Efficient,
AUTHOR="Emilio Paolini and Andrea Pinto and Luca Valcarenghi and Nicola Andriolli
and Luca Maggiani and Flavio Esposito",
TITLE="Efficient Distributed Learning Over Lossy Wireless Networks",
BOOKTITLE="2024 20th International Conference on Network and Service Management (CNSM)
(CNSM 2024)",
ADDRESS="Prague, Czech Republic",
PAGES=7,
DAYS=27,
MONTH=oct,
YEAR=2024,
ABSTRACT="In the context of NextG Wireless Networks, addressing the challenges of
reliability of wireless communication links is paramount to ensure
efficient Distributed Learning systems. However, many recent solutions have
overlooked key challenges, such as packet-level losses and the impact of
TCP retransmission, that are crucial for the robustness of these systems.
In this paper, we propose the integration of fountain codes into the
distributed learning process to offer a robust mechanism to counteract
packet loss. Specifically, we propose a cumulative strategy logic based on
fountain codes specifically tailored for packet exchanges in Distributed
Learning applications. Our evaluation shows that fountain codes
significantly enhance the efficiency and reliability of distributed
learning model updates under conditions of high packet loss, e.g., a packet
reduction of ≈ 84\% (≈ 60\% at the UE (gNB) side compared to
traditional TCP methods when packet loss probability reaches 0.9 in
Federated Learning scenarios. However, under low packet loss scenarios,
fountain codes computational overhead becomes non-negligible.These results
highlight the potential of fountain codes to serve as a robust alternative
to conventional communication protocols in distributed learning systems,
particularly in environments characterized by unstable network conditions."
}