Skip to content

Official implementation of "FedBand: Adaptive Federated Learning Under Strict Bandwidth Constraints"

Notifications You must be signed in to change notification settings

t-alanazi/FedBand

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

11 Commits
 
 
 
 
 
 

Repository files navigation

FedBand: Adaptive Federated Learning Under Strict Bandwidth Constraints

Official PyTorch implementation of the paper:

FedBand: Adaptive Federated Learning Under Strict Bandwidth Constraints
Taghreed Alanazi; Abdulrahman Fahim; Muntaka Ibnath; Basak Guler; Amit Roy Chowdhury; Ananthram Swami; Evangelos E. Papalexakis; Srikanth V. Krishnamurthy, ICCCN 2025

https://ieeexplore.ieee.org/document/11133779


Overview

FedBand is a communication-efficient Federated Learning (FL) framework that dynamically allocates per-client compression budgets based on model behavior (validation loss or gradient norm).

The method enforces a global bandwidth constraint each round while maximizing accuracy and fairness.

This repository includes:

  • CIFAR-10 images classification experiments
  • UTMobileNet2021 trafic classification experiments
  • Dynamic bandwidth allocation
  • Top-k / sparse update compression
  • Fairness & per-client metrics

Installation

git clone https://github.com/t-alanazi/FedBand.git
cd FedBand
pip install -r requirements.txt

## Run CIFAR-10
python3 -m fedband.run_cifar10

## Run UTMobileNet2021
python3 -m fedband.run_utmobilenet --base_path /path/to/data