VCFedAdam: A Verifiable and Compressed FedAdam Aggregation Scheme for Federated Learning in Distributed Sensor Networks †

Zhiwei Si, Xiuheng Liao, Ziang Wu, Tianhui Li, Chunhua Su · 2026

Read VCFedAdam: A Verifiable and Compressed FedAdam Aggregation Scheme for Federated Learning in Distributed Sensor Networks † by Zhiwei Si, Xiuheng Liao, Ziang Wu, Tianhui Li, Chunhua Su (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/pmc-13611767.txt.

Description

Zhiwei Si and colleagues design a training scheme for sensor networks that learns together without leaking data or trusting the server, in this computing paper from twenty twenty-six. In federated learning, devices train a shared model locally and send only updates. But updates can leak private facts, and a dishonest server can fake the combined result.

Their scheme, VCFedAdam, compresses each update into a small masked sketch, so little travels and little can leak. A commitment check then lets devices verify the server's combined answer rather than take it on faith. At a quarter of the original size, computing cost fell by about ninety percent and traffic by about three quarters, against standard secure methods. Accuracy on two standard image tests dipped by about one percent. The authors aim the design at the small, battery-bound sensors that need such thrift most.

Catalog record

CREATORZhiwei Si, Xiuheng Liao, Ziang Wu, Tianhui Li, Chunhua Su
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE56 KB
VERSION1.0
LICENSECC BY 4.0
RIGHTSCC BY
SOURCEEurope PMC

Source: https://europepmc.org/articles/PMC13611767

Browse the full library · About the Tower of Babel