A Full Compression Pipeline for Green Federated Learning in Communication-Constrained Environments

Elouan Colybes; Shirin Salehi; Anke Schmeink

cs.LGApr 13, 2026

A Full Compression Pipeline for Green Federated Learning in Communication-Constrained Environments

Authors: Elouan Colybes, Shirin Salehi, Anke Schmeink

Organizations: Chair of Information Theory and Data Analytics (INDA), RWTH Aachen University, Aachen, Germany

Abstract

Federated Learning (FL) enables collaborative model training across distributed clients without sharing raw data, thereby preserving privacy. However, FL often suffers from significant communication and computational overhead, limiting its scalability and sustainability. In this work, we introduce a Full Compression Pipeline (FCP) for FL in communication-constrained environments. FCP integrates three complementary deep compression techniques (pruning, quantization, and Huffman encoding) into a unified end-to-end framework. By compressing local models and communication payloads, FCP substantially reduces transmission costs and resource consumption while maintaining competitive accuracy. To quantify its impact, we develop an evaluation framework that captures both communication and computation overheads as a unified model cost, allowing a holistic assessment of efficiency trade-offs. The pipeline is evaluated in an independent and identically distributed (IID) and non-IID data setting. In one representative scenario, training a ResNet-12 model on the CIFAR-10 dataset with ten clients and a 2 Mbps bandwidth, the FCP achieves more than 11 $\times$ reduction in model size, with only a 2% drop in accuracy compared to the uncompressed baseline. This results in an FL training that is more than 60% faster.

Figures & tables

Fig. 1 : Our full compression pipeline to reduce communication overhead in federated learning

w^{(i)}_{\text{global}}\;=\;\sum_{s\in S}\frac{n_{s}}{n}\,w_{s}^{(i)}

Algorithm 1 FL with Full Compression Pipeline (FCP)

	Pruning Rate $\gamma$
Number of	0.1	0.2	0.3	0.4	0.5	0.6	0.7	0.8	0.9	0.95
centroids $k$	IID / non-IID CIFAR-10, baseline = 80.7% / 75.6%
256	80.6 / 73.6	80.3 / 75.0	79.8 / 74.5	80.0 / 75.0	80.1 / 72.0	79.9 / 70.6	78.6 / 67.5	77.9 / 65.5	73.9 / 59.3	70.3 / 54.3
128	81.1 / 73.2	80.3 / 73.4	80.3 / 72.3	80.0 / 70.6	79.0 / 69.0	79.9 / 66.6	79.8 / 63.7	77.9 / 59.1	73.9 / 55.7	70.9 / 50.2
64	80.6 / 59.4	79.8 / 59.8	80.3 / 57.0	79.9 / 53.0	80.2 / 57.2	79.5 / 53.4	79.1 / 51.4	77.2 / 50.5	72.8 / 44.1	69.5 / 43.5
32	79.2 / 45.4	79.2 / 45.8	79.1 / 44.9	78.9 / 43.6	78.5 / 42.5	75.5 / 43.8	75.7 / 42.1	74.8 / 40.3	69.6 / 36.2	66.9 / 35.5

TABLE I : FCP performance accuracy values for different pruning rates and quantization levels in multiple scenarios.

	Pruning Rate $\gamma$	0.1	0.2	0.3	0.4	0.5	0.6	0.7	0.8	0.9	0.95
	256	0.227	0.205	0.183	0.160	0.140	0.117	0.090	0.065	0.037	0.022
	128	0.193	0.179	0.158	0.139	0.122	0.101	0.079	0.056	0.032	0.019
Number of	64	0.165	0.151	0.134	0.117	0.104	0.085	0.068	0.048	0.027	0.016
centroids $k$	32	0.133	0.122	0.111	0.097	0.086	0.073	0.057	0.043	0.023	0.014
	16	0.103	0.097	0.085	0.079	0.073	0.063	0.049	0.035	0.021	0.013
	8	0.092	0.089	0.080	0.074	0.069	0.056	0.046	0.034	0.020	0.012

TABLE II : Compression ratio for client communication overhead for different pruning rates and quantization levels.

Federated Learning

Federated learning trains shared models across data holders without requiring them to pool their raw data.

See all 921 papers

Adaptive Heterogeneous Compression for Resource-Efficient Federated Knowledge Distillation
Aug 16, 2026Chenwang Liu, Yijun Liu, Chang Liu +2Federated Knowledge Distillation Federated Learning
School of Information Engineering, Guangdong University of Technology, Guangzhou 510006, China · School of Communication and Information Engineering, Chongqing University of Posts and Telecommunications, Chongqing 400065, China
Enhanced Privacy and Communication Efficiency in Non-IID Federated Learning with Adaptive Quantization and Differential Privacy
Apr 25, 2026Emre Ardıç, Yakup GençNon-IID Federated Learning Differentially Private Federated Learning
Department of Computer Engineering, Gebze Technical University, Gebze, Kocaeli 41400, Türkiye
ERIS: Enhancing the Privacy-Utility-Scalability Trade-Off in Federated Learning via Federated Shard Aggregation
Feb 9, 2026Dario Fenoglio, Pasquale Polverino, Akash Dhasade +3Federated Learning Federated Learning Aggregation
Universit`a della Svizzera italiana Lugano, Switzerland · Carnegie Mellon University Pittsburgh, PA, United States

Aug 16, 2026J/K move · Enter open · S save

Adaptive Heterogeneous Compression for Resource-Efficient Federated Knowledge Distillation

Chenwang Liu, Yijun Liu, Chang Liu, Xu Zhang+1

School of Information Engineering, Guangdong University of Technology, Guangzhou 510006, China · School of Communication and Information Engineering, Chongqing University of Posts and Telecommunications, Chongqing 400065, China

Federated Knowledge Distillation Federated Learning Communication-Efficient Distributed Training Heterogeneous Federated Learning

Open paper AskPDF arXiv

Federated learning (FL) enables privacy-preserving distributed model training but faces challenges from heterogeneous model architectures and limited communication resources at the network edge. Federated knowledge distillation (FedKD) alleviates model heterogeneity by combining prototype-wise parameter aggregation and knowledge transfer across heterogeneous models. However, transmitting gradients still introduces considerable communication overhead, while existing compression approaches typically apply a uniform strategy across clients and ignore their diverse model characteristics and resource capacities. To address this issue, we propose a heterogeneous compression framework for FedKD that enables each client to select a compression strategy from a candidate strategy set. We formulate the compression strategy selection problem as a non-stationary stochastic multi-armed bandit (MAB), where each arm corresponds to a compression strategy. An efficiency-aware reward is designed by jointly considering local optimization improvement, global knowledge alignment, and execution time. Based on this formulation, we develop an Adaptive heterogeneouS Compression algorithm for fEderated kNowledge Distillation (ASCEND), which employs an exponential moving average (EMA)-enhanced

ε

-greedy policy to balance exploration and exploitation. Experimental results on multiple datasets demonstrate that ASCEND effectively adapts to heterogeneous model and resource settings, reducing communication overhead and training time while maintaining competitive model accuracy.

Abstract

Federated Learning

Explore similar work

Adaptive Heterogeneous Compression for Resource-Efficient Federated Knowledge Distillation

Enhanced Privacy and Communication Efficiency in Non-IID Federated Learning with Adaptive Quantization and Differential Privacy

ERIS: Enhancing the Privacy-Utility-Scalability Trade-Off in Federated Learning via Federated Shard Aggregation