Submitted to Journal of Network and Computer Applications
DecoyTrace: Toxic Decoys for Active Defense in Decentralized Federated Learning
DecoyTrace uses cyber deception to defend serverless DFL. A mobile decoy node sends bait models to neighbours, flags poisoned responses, isolates the source and recovers training, evaluated in 60 NEBULA configurations.
Abstract
Decentralized Federated Learning (DFL) eliminates the central aggregation server, reducing the single point of observation that traditional defenses against attacks rely on. As a result, peer-to-peer networks become exposed to malicious updates containing backdoors or semantic poisoning, since such updates can remain close to benign ones in the parameter space while behaving very differently. This may evade defenses based on passive parameter inspection. However, existing deception-based defenses have mainly been designed for centralized FL and do not jointly address local observation, poisoning propagation, source attribution, and containment in strictly serverless DFL. To address these limitations, this paper presents DecoyTrace, a proactive cyber deception-based defense for strictly serverless DFL environments. DecoyTrace deploys a mobile DecoyNode that generates decoy challenges using chaotic maps, disseminates a dual model (clean vs. decoy) based on neighbor trust, and evaluates them using three-state semantic metrics. Upon confirmation, a distributed protocol isolates the source and performs a model reset or recovery to preserve training progress. Evaluated across sixty configurations on the NEBULA platform (five datasets, three topologies, and four attack/defense scenarios), DecoyTrace systematically restores lost utility. The F1-score remains within 0.03 of the baseline on MNIST/FashionMNIST (mitigating drops of up to 0.37), matches or exceeds the baseline on EMNIST and CIFAR-100, and remains between 0.05 and 0.10 below the baseline on CIFAR-10, the most visually complex convolutional scenario evaluated. Furthermore, containment reduces CPU and network usage by up to two-thirds. These results demonstrate the feasibility of unifying deception, identification, and containment in DFL, while also identifying its limitations in complex tasks and multi-attractor threat models.
Authors
Pedro Beltrán López, Enrique Tomás Martínez Beltrán, Pantaleone Nespoli, Manuel Gil Pérez, Alberto Huertas Celdrán
Keywords
- Decentralized Federated Learning
- Cyber Deception
- Active Defense
- Poisoning Attacks
- Backdoor Attacks
- Attack Mitigation
Related publications
Publications connected by topic, method or application.
Computer Networks
RepuNet: A Reputation System for Mitigating Malicious Clients in DFL
Isaac Marroqui Penalva, Enrique Tomás Martínez Beltrán, Manuel Gil Pérez, Alberto Huertas Celdrán
RepuNet studies how to reduce the influence of malicious DFL clients. It combines model and communication signals into a reputation score that adjusts each peer's weight during aggregation, with an evaluation in NEBULA.
Information Fusion
Decentralized Federated Learning with Multimodal Prototypes for Heterogeneous Data
Enrique Tomás Martínez Beltrán, Gérôme Bovet, Gregorio Martínez Pérez, Alberto Huertas Celdrán
Modalis addresses clients with different or missing modalities in decentralized learning. It exchanges compact class prototypes and aligns local representations to support collaboration across heterogeneous data.
XI Jornadas Nacionales de Investigación en Ciberseguridad (JNIC 2026)
MadHoney: Señuelos Tóxicos para la Defensa Activa en el Aprendizaje Federado Descentralizado
Pedro Beltrán López, Enrique Tomás Martínez Beltrán, Pantaleone Nespoli, Manuel Gil Pérez, Gregorio Martínez Pérez, Alberto Huertas Celdrán
Related Research

Jan 2026 / Dec 2026
TITAN: Trustworthy and Intelligent Threat Analysis
A framework that joins MLOps and LLMOps for cyber defense, audits every AI decision in real time and executes only validated responses.

Apr 2023 / Nov 2023
DEFENDIS: Decentralized Federated Learning for IoT Device Identification and Security
DEFENDIS develops a framework for uniquely identifying IoT devices in a distributed manner while solving security threats through decentralized federated learning.

Nov 2025 / Jan 2026
Cybersecurity and Distributed Federated Learning
An international research exchange program investigating threat mitigation models for node-to-node federated learning.