# On the Convergence Time of Federated Learning Over Wireless Networks Under Imperfect CSI

arXiv: https://arxiv.org/abs/2104.00331 · HTML: https://mlsanity.com/paper/2104.00331/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- Builds on 12 of its 19 references that are in the map.
- First posted on arXiv 1 Apr 2021.

## Builds on (most cited first, 12 of 12)

- [Communication-Efficient Learning of Deep Networks from Decentralized Data](https://mlsanity.com/paper/1602.05629/) (2016)
- [Convergence Time Optimization for Federated Learning over Wireless Networks](https://mlsanity.com/paper/2001.07845/) (2020)
- [Convergence of Update Aware Device Scheduling for Federated Learning at the Wireless Edge](https://mlsanity.com/paper/2001.10402/) (2020)
- [Federated Learning With Quantized Global Model Updates](https://mlsanity.com/paper/2006.10672/) (2020)
- [Opportunities of Federated Learning in Connected, Cooperative and Automated Industrial Systems](https://mlsanity.com/paper/2101.03367/) (2021)
- [Federated Learning under Channel Uncertainty: Joint Client Scheduling and Resource Allocation](https://mlsanity.com/paper/2002.00802/) (2020)
- [Towards 6G Networks: Use Cases and Technologies](https://mlsanity.com/paper/1903.12216/) (2019)
- [Delay Minimization for Federated Learning Over Wireless Communication Networks](https://mlsanity.com/paper/2007.03462/) (2020)
- [Coverage and Rate Analysis for Millimeter Wave Cellular Networks](https://mlsanity.com/paper/1402.6430/) (2014)
- [Timely Communication in Federated Learning](https://mlsanity.com/paper/2012.15831/) (2020)
- [Simplified Ray Tracing for the Millimeter Wave Channel: A Performance Evaluation](https://mlsanity.com/paper/2002.09179/) (2020)
- [From Federated to Fog Learning: Distributed Machine Learning over Heterogeneous Wireless Networks](https://mlsanity.com/paper/2006.03594/) (2020)

## Built on by (most cited first, 9 of 9)

- [SHED: A Newton-type algorithm for federated learning based on incremental Hessian eigenvector sharing](https://mlsanity.com/paper/2202.05800/) (2022)
- [Federated Learning for 6G: Paradigms, Taxonomy, Recent Advances and Insights](https://mlsanity.com/paper/2312.04688/) (2023)
- [Multiple Access in the Era of Distributed Computing and Edge Intelligence](https://mlsanity.com/paper/2403.07903/) (2024)
- [Distributed Resource Allocation for URLLC in IIoT Scenarios: A Multi-Armed Bandit Approach](https://mlsanity.com/paper/2211.12201/) (2022)
- [Towards Decentralized Predictive Quality of Service in Next-Generation Vehicular Networks](https://mlsanity.com/paper/2302.11268/) (2023)
- [Federated Learning-Distillation Alternation for Resource-Constrained IoT](https://mlsanity.com/paper/2505.20456/) (2025)
- [Q-SHED: Distributed Optimization at the Edge via Hessian Eigenvectors Quantization](https://mlsanity.com/paper/2305.10852/) (2023)
- [VR-VFL: Joint Rate and Client Selection for Vehicular Federated Learning Under Imperfect CSI](https://mlsanity.com/paper/2602.03711/) (2026)
- [Hidden in Rounds: Predicting the Time Cost of 802.11 Contention in Federated Learning](https://mlsanity.com/paper/2609.12903/) (2026)

## Abstract (arXiv)

Federated learning (FL) has recently emerged as an attractive decentralized solution for wireless networks to collaboratively train a shared model while keeping data localized. As a general approach, existing FL methods tend to assume perfect knowledge of the Channel State Information (CSI) during the training phase, which may not be easy to acquire in case of fast fading channels. Moreover, literature analyses either consider a fixed number of clients participating in the training of the federated model, or simply assume that all clients operate at the maximum achievable rate to transmit model data. In this paper, we fill these gaps by proposing a training process that takes channel statistics as a bias to minimize the convergence time under imperfect CSI. Numerical experiments demonstrate that it is possible to reduce the training time by neglecting model updates from clients that cannot sustain a minimum predefined transmission rate. We also examine the trade-off between number of clients involved in the training process and model accuracy as a function of different fading regimes.