Communication-Efficient Federated Learning with Dual-Sided Sparse Aggregation for Edge Sensing Systems
He Zhao, Jingwei Li · 2026
Read Communication-Efficient Federated Learning with Dual-Sided Sparse Aggregation for Edge Sensing Systems by He Zhao, Jingwei Li (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-13165942.txt.
Description
Sensors at the edge of a network, from wearables to cameras, gather data too private and too bulky to ship to a centre. He Zhao and Jingwei Li train shared models on such data in 2026 without moving the raw records. Their method thins what is sent in both directions on each round.
The server trims the shared model and sends only the kept parts. Each device then trains the trimmed parts and sends back sparse updates. That fixed sparse plan cuts the traffic and the local compute load at once. A second rule, which keeps local models from drifting too far from the shared one, guards the result when each device's data look unlike the rest. Edge fleets gain a way to learn as one while each keeps its data home.
Catalog record
| CREATOR | He Zhao, Jingwei Li |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 43 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |