ASFL: An Adaptive Model Splitting and Resource Allocation Framework for Split Federated Learning

Chuiyang Meng 1 , Graduate Student Member, IEEE , Ming Tang 2 , Member, IEEE , and Vincent W.S. Wong 1 , Fellow, IEEE · 2026

Read ASFL: An Adaptive Model Splitting and Resource Allocation Framework for Split Federated Learning by Chuiyang Meng 1 , Graduate Student Member, IEEE , Ming Tang 2 , Member, IEEE , and Vincent W.S. Wong 1 , Fellow, IEEE (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-04437.txt.

Description

Meng and colleagues split federated learning adaptively with ASFL on wireless networks. Weak clients face high delay and energy costs when training alone. ASFL offloads part of the model to the server and adapts the split and resources during training. Convergence analysis supports a joint performance and allocation problem solved by online block descent. Against five baselines, ASFL converges faster, cutting delay by up to 75 percent and energy by up to 80 percent.

Catalog record

CREATORChuiyang Meng 1 , Graduate Student Member, IEEE , Ming Tang 2 , Member, IEEE , and Vincent W.S. Wong 1 , Fellow, IEEE
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE91740
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2603.04437

Browse the full library · About the Tower of Babel