FedSparQ: Adaptive Sparse Quantization with Error Feedback for Robust & Efficient Federated Learning

Federated learning (FL) enables decentralized model training on private client data but is frequently bottlenecked by the cost of communicating high-dimensional updates over constrained networks. In this paper, we present FedSparQ, a lightweight compression framework that dynamically sparsifies each client's gradient via an adaptive threshold, applies halfprecision quantization to the retained entries, and integrates error-feedback residuals to prevent information loss. FedSparQ requires no manual tuning of sparsity rates or quantization schedules, adapts seamlessly to both homogeneous and heterogeneous data distributions, and is agnostic to model architecture. Through extensive empirical evaluation on vision benchmarks under independent and identically distributed (IID) and non-IID data, we show that FedSparQ substantially reduces communication overhead (reducing by 90 % of bytes sent comparing to FedAvg) while preserving or improving model accuracy (improving by 6 % compared to FedAvg non-compressed solution or to state-of-the-art compression models), and enhancing convergence robustness (by 50 %, compared to the other baselines). Our approach provides a practical, easy-to-deploy solution for bandwidth-constrained federated deployments and lays the groundwork for future extensions in adaptive precision and privacy-preserving protocols.