Authors: Nizar Masmoudi, Wael Jaafar

Published on: March 06, 2024

Impact Score: 7.8

Arxiv code: Arxiv:2403.04037

Summary

• What is new: Introduces OCD-FL, a novel scheme for efficient communication and energy consumption in decentralized federated learning.
• Why this is important: Decentralized FL presents challenges like single points of failure, network bottlenecks, high communication costs, and data heterogeneity.
• What the research proposes: OCD-FL optimizes peer selection for FL collaboration to enhance knowledge gain and reduce energy consumption.
• Results: OCD-FL achieves comparable or superior performance to traditional FL while reducing energy consumption by 30% to 80%.

Technical Details

Technological frameworks used: nan

Models used: Federated Learning, Decentralized FL, OCD-FL

Data used: IoT network data

Potential Impact

Companies in IoT, smart devices, and edge computing could greatly benefit; traditional centralized data processing entities might face disruption.

Want to implement this idea in a business?

We have generated a startup concept here: EffiLearn.