This is a repository for the implementation of the paper "Green, Quantized Federated Learning over Wireless Networks: An Energy-Efficient Design".
"NBI_method.zip" contains code for the introduced NBI method to find the Pareto boundary
Another python files (such as Train.py) are Pytorch implementation for the proof of concept of Quantized FL. You can set arbitrary precision levels (n and m), the number of local epochs, schedulingset size, or neural networks to run the presented Quantized FL algorithm.