Joint Routing and Model Pruning for Decentralized Federated Learning in Bandwidth-Constrained Multi-Hop Wireless Networks

Xiaoyu He; Weicai Li; Tiejun Lv; Xi Yu

cs.LGMar 16, 2026

Joint Routing and Model Pruning for Decentralized Federated Learning in Bandwidth-Constrained Multi-Hop Wireless Networks

Authors: Xiaoyu He, Weicai Li, Tiejun Lv, Xi Yu

Organizations: School of Information and Communication Engineering, Beijing University of Posts and Telecommunications, Beijing 100876, China · Center for Target Cognition Information Processing Science and Technology, and the Key Laboratory of Modern Measurement and Control Technology, Ministry of Education, both at Beijing Information Science and Technology University, Beijing, China

Abstract

Decentralized federated learning (D-FL) enables privacy-preserving training without a central server, but multi-hop model exchanges and aggregation are often bottlenecked by communication resource constraints. To address this issue, we propose a joint routing-and-pruning framework that optimizes routing paths and pruning rates to maintain communication latency within prescribed limits. We analyze how the sum of model biases across all clients affects the convergence bound of D-FL and formulate an optimization problem that maximizes the model retention rate to minimize these biases under communication constraints. Further analysis reveals that each client's model retention rate is path-dependent, which reduces the original problem to a routing optimization. Leveraging this insight, we develop a routing algorithm that selects latency-efficient transmission paths, allowing more parameters to be delivered within the time budget and thereby improving D-FL convergence. Simulations demonstrate that, compared with unpruned systems, the proposed framework reduces average transmission latency by 27.8% and improves testing accuracy by approximately 12%. Furthermore, relative to standard benchmark routing algorithms, the proposed routing method improves accuracy by roughly 8%.

Figures & tables

Notation	Definition
$m,n,N$	Client indices and total number of clients
$\mathcal{V},\mathcal{E}$	Sets of clients and edges
$\mathcal{N}_{n}$	Neighboring nodes of client $n$
$\alpha$	Index of training rounds
$p_{n}$	Ideal aggregation coefficient of client $n$
$p_{m,n}$	Aggregation coefficient of client $m$ at client $n$

TABLE I: Notation and definitions

Reference	Convergence	Multi-hop	Comm. Constraints	Model Pruning	Joint Pruning & Routing
[ 6 ]	$\checkmark$	$\checkmark$	$\times$	$\checkmark$	$\times$
[ 45 ]	$\checkmark$	$\times$	$\checkmark$	$\checkmark$	$\times$
[ 25 ]	$\times$	$\times$	$\checkmark$	$\checkmark$	$\times$
[ 39 ]	$\checkmark$	$\times$	$\checkmark$	$\times$	$\times$
This paper	$\checkmark$	$\checkmark$	$\checkmark$	$\checkmark$	$\checkmark$

TABLE II : Comparison with Related Works in Decentralized Federated Learning

Fig. 1 : An illustration of the multi-hop route for transmitting the local model from the source client to all target clients in the

\alpha

-th round.

Fig. 2 : Workflow of the D-FL rounds with local model update, pruning, multi-hop routing, and aggregation.

Fig. 3 : Schematic of structured pruning. This figure shows the structured pruning process applied to input and output channels in a hidden layer. Blue indicates retained channels, while gray and yellow represent pruned input and output channels, respectively. Each small square denotes a model parameter.

Fig. 4 : An illustration of the TDMA-based collision-free protocol for model exchange in a decentralized network.

Algorithm	Kruskal (MST)	Bellman-Ford	Flooding (Flood-Fill)	Algorithm 2 (Proposed)
Complexity	$N^{2}\log N$	$N^{3}$	$N^{2}$	$\Psi\times N^{2}\log N$

TABLE III : Comparison of Computational Complexity of Different Algorithms

Panel of Figure 8 — Fig. 5 : Illustration of the proposed routing Algorithm 2 at different optimization stages and the impact of parameters $\theta$ and $\Psi$ on performance ( $N=20$ ).

Fig. 6 : Testing accuracy comparison of multi-hop and P2P aggregation strategies on F-CIFAR10 & ResNet-18 across different numbers of clients.

Panel of Figure 10 — Fig. 7 : (a)-(b) Testing accuracy and loss of different pruning schemes on the F-CIFAR10 & ResNet-18; (c)-(d)Testing accuracy and loss of different pruning schemes on the F-CIFAR100 & ResNet-34; (e) Distribution of per-client model transmission time under different pruning schemes on F-CIFAR10 & ResNet-18.

Panel of Figure 11 — Fig. 8 : (a)–(b) Testing accuracy and loss of the proposed joint pruning and routing method on F-CIFAR10 & ResNet-18 under varying bandwidths; (c)–(d) Testing accuracy and loss on F-CIFAR100 & ResNet-34 under varying bandwidths; (e) Distribution of model retention rates and cumulative maximum link weights per hop across 20 clients under varying bandwidths.

Panel of Figure 12 — Fig. 9 : (a)–(b) Testing accuracy and loss of the proposed joint pruning and routing method on F-CIFAR10 & ResNet-18 under different latency thresholds; (c)–(d) Testing accuracy and loss on F-CIFAR100 & ResNet-34 under different latency thresholds; (e) Distribution of model retention rates and cumulative maximum link weights per hop across 20 clients under different latency thresholds.

Panel of Figure 13 — Fig. 10 : Comparison of the different pruning strategies in terms of test accuracy of ResNet18 & F-CIFAR10 under different client numbers and data distributions: (a) Non-IID with 20 clients; (b) IID with 20 clients; (b) IID with 50 clients

Panel of Figure 14 — Fig. 11 : (a) Comparison of the proposed routing algorithm and traditional algorithms in testing accuracy on F-CIFAR10 & ResNet18; (b)-(c) Comparison with benchmark methods in testing accuracy; (d) Distribution of model retention rates and cumulative maximum link weights per hop for P_CLT and benchmark methods across 20 clients.

Panel of Figure 15 — Fig. 12 : Performance comparison between the proposed enhanced routing method and the baseline under communication bottlenecks: (a) Testing accuracy on F-CIFAR10 & ResNet-18; (b) Comparison of fairness metrics.

Federated Learning

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

See all 943 papers

Joint Optimization for Federated Learning and Transmission over Unreliable Wireless Networks with Heterogeneous Data
Sep 13, 2026Changheng Wang, Xianchao Zhang, Zhiqing Wei +3Federated Learning Federated Learning Aggregation
Key Labora-tory of Universal Wireless Communications, Ministry of Education, School of Information and Communication Engineering, Beijing University of Posts and Telecommunications, Beijing 100876, China · Provincial Key Laboratory of Multimodal Perceiving and Intelligent Systems, Jiaxing University, Jiaxing 314001, China · School of Electronics and Information, Northwestern Polytechnical University, Xi’an 710000, China +1
Tram-FL: Reducing Communication and Computation Costs through Sequential Model Circulation in Decentralized Federated Learning
Oct 6, 2026Kota Maejima, Takayuki Nishio, Asato Yamazaki +1Communication-Efficient Distributed Training Federated Learning
School of Engineering, Institute of Science Tokyo, Tokyo 152-8550, Japan
SubFLOT: Submodel Extraction for Efficient and Personalized Federated Learning via Optimal Transport
Apr 8, 2026Zheng Jiang, Nan He, Yiming Chen +1Federated Learning Heterogeneous Federated Learning
Tsinghua University · Beijing University of Technology · Key Laboratory of Pervasive Computing, Ministry of Education

Sep 13, 2026J/K move · Enter open · S save

Joint Optimization for Federated Learning and Transmission over Unreliable Wireless Networks with Heterogeneous Data

Changheng Wang, Xianchao Zhang, Zhiqing Wei, Lingzhu Zhao+2

Key Labora-tory of Universal Wireless Communications, Ministry of Education, School of Information and Communication Engineering, Beijing University of Posts and Telecommunications, Beijing 100876, China · Provincial Key Laboratory of Multimodal Perceiving and Intelligent Systems, Jiaxing University, Jiaxing 314001, China · School of Electronics and Information, Northwestern Polytechnical University, Xi’an 710000, China+1

Federated Learning Federated Learning Aggregation Non-IID Federated Learning Wireless Resource Allocation

Open paper AskPDF arXiv

In wireless federated learning (FL), data heterogeneity and multiple local updates induce client drift, degrading model convergence. It is further affected by unreliable wireless links, as transmission errors may invalidate model updates. To address these challenges, we propose a federated random walk averaging (FedRW) framework, which is a variant of federated averaging (FedAvg) that mitigates data heterogeneity by updating models along random walk (RW) paths and aggregating them at the server. Model parameters are transmitted in packets with retransmission support to improve training quality by mitigating wireless errors along RW paths. Meanwhile, wireless transmission delays hinder the exploration of FedRW. To this end, we formulate a joint optimization problem that integrates learning, RW path selection, and transmission parameter tuning, aiming to minimize the training loss under delay constraints. By deriving an upper bound on the expected convergence of FedRW over unreliable wireless networks, we reduce the problem to a general form agnostic to task type and model architecture. A distributed solution is then proposed, in which the server or clients optimize packet size and maximum number of retransmissions locally, and efficiently select reliable and expandable next-hop nodes via a resilience-aware beam search with dynamic pruning. Simulation results show that FedRW achieves 2.26%-9% higher accuracy than state-of-the-art baselines under high data heterogeneity. Furthermore, the jointly optimized FedRW yields at least 2.78% higher accuracy and faster convergence compared to baselines.

Abstract

Federated Learning

Explore similar work

Joint Optimization for Federated Learning and Transmission over Unreliable Wireless Networks with Heterogeneous Data

Tram-FL: Reducing Communication and Computation Costs through Sequential Model Circulation in Decentralized Federated Learning

SubFLOT: Submodel Extraction for Efficient and Personalized Federated Learning via Optimal Transport