Exportar Publicação

Nwokoro, I. S., Osaghae, E. O., Sambo, M. Q., Eze, I. F., Akinfenwa, T. O., Oshodin, O. G...Zacciah, K. A.-O. (2025). Federated learning with memory-based cognitive engine to optimize multi-service 5G QoS in a privacy-preserving framework. Journal of Networking and Communication Systems. 8 (4)

I. S. Nwokoro et al.,  "Federated learning with memory-based cognitive engine to optimize multi-service 5G QoS in a privacy-preserving framework", in Journal of Networking and Communication Systems, vol. 8, no. 4, 2025

@article{nwokoro2025_1777296557584,
	author = "Nwokoro, I. S. and Osaghae, E. O. and Sambo, M. Q. and Eze, I. F. and Akinfenwa, T. O. and Oshodin, O. G and Nwatu, A. N. and Zacciah, K. A.-O.",
	title = "Federated learning with memory-based cognitive engine to optimize multi-service 5G QoS in a privacy-preserving framework",
	journal = "Journal of Networking and Communication Systems",
	year = "2025",
	volume = "8",
	number = "4",
	url = "https://publisher.resbee.org/jnacs/archive/v8i4/a3.html"
}

TY  - JOUR
TI  - Federated learning with memory-based cognitive engine to optimize multi-service 5G QoS in a privacy-preserving framework
T2  - Journal of Networking and Communication Systems
VL  - 8
IS  - 4
AU  - Nwokoro, I. S.
AU  - Osaghae, E. O.
AU  - Sambo, M. Q.
AU  - Eze, I. F.
AU  - Akinfenwa, T. O.
AU  - Oshodin, O. G
AU  - Nwatu, A. N.
AU  - Zacciah, K. A.-O.
PY  - 2025
SN  - 2582-3817
UR  - https://publisher.resbee.org/jnacs/archive/v8i4/a3.html
AB  - Fifth-Generation New Radio (5G-NR) networks support heterogeneous services with highly diverse Quality of Service (QoS) requirements, including enhanced Mobile Broadband (eMBB), massive Machine-Type Communications (mMTC), and ultra-reliable low-latency communications (URLLC). These service categories are designed to address distinct application domains such as immersive multimedia streaming, large-scale Internet of Things (IoT) deployments, and mission- critical industrial automation. Ensuring fair, efficient, and trustworthy resource allocation across these services under dynamic traffic conditions remains a critical challenge, particularly when conventional centralized QoS management approaches face scalability, privacy, and adaptability limitations. This paper proposes a decentralized, trust-aware QoS management framework that integrates Federated Learning (FL) with a memory-based Cognitive Smart Engine (CSE) for adaptive multi-service resource allocation in 5G networks. FL enables collaborative QoS prediction across distributed network nodes without exposing raw user data, thereby preserving privacy and enhancing network trust. The CSE leverages historical QoS knowledge and reinforcement learning to dynamically optimize resource allocation while improving Subscriber Comfort Experience (SCE) and Service Level Agreement (SLA) compliance. The proposed framework is evaluated using a simulation environment combining MATLAB/Simulink and Python-based deep learning tools under variable traffic loads for eMBB, mMTC, and URLLC services. Simulation results demonstrate that the FL + CSE framework reduces average latency by up to 57%, improves packet delivery success by 7–10%, and increases SLA compliance by approximately 9% compared to centralized QoS management. These findings highlight the effectiveness of decentralized intelligence for scalable, privacy- preserving, and adaptive QoS management in 5G and beyond-5G networks.
ER  -