QoS-based Intelligent multi-connectivity for B5G networks

2025-08-22

AI Summary

To address the challenge of heterogeneous Quality-of-Service (QoS) requirements—namely, throughput, latency, and reliability—across diverse applications in Beyond-5G (B5G) networks, this paper proposes an intelligent multi-connectivity framework based on Deep Neural Networks (DNNs). The method employs real-time DNN-based prediction of QoS metrics across multiple base stations to jointly optimize dynamic service cluster selection and inter-base-station data rate allocation. Crucially, it is the first to deeply embed DNNs within the multi-connectivity decision-making closed loop, enabling fine-grained, QoS-aware resource coordination. Experimental results demonstrate a 98% QoS satisfaction rate, a 30% gain in spectral efficiency over state-of-the-art schemes, and superior robustness under high-mobility and dynamic interference conditions. The framework establishes a scalable, unified infrastructure paradigm for differentiated service provisioning in B5G networks.

Technology Category

• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Search and Optimization: Distributed Search
• Machine Learning: Deep Neural Architectures and Foundation Models

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Economics, Online Markets and Human Computation: Incentives in network design for Web infrastructures and ecosystems

Abstract

The rapid advancement of communication technologies has established cellular networks as the backbone for diverse applications, each with distinct quality of service requirements. Meeting these varying demands within a unified infrastructure presents a critical challenge that can be addressed through advanced techniques such as multi-connectivity. Multiconnectivity enables User equipments to connect to multiple BSs simultaneously, facilitating QoS differentiation and provisioning. This paper proposes a QoS-aware multi-connectivity framework leveraging machine learning to enhance network performance. The approach employs deep neural networks to estimate the achievable QoS metrics of BSs, including data rate, reliability, and latency. These predictions inform the selection of serving clusters and data rate allocation, ensuring that the User Equipment connects to the optimal BSs to meet its QoS needs. Performance evaluations demonstrate that the proposed algorithm significantly enhances Quality of Service (QoS) for applications where traditional and state-of-the-art methods are inadequate. Specifically, the algorithm achieves a QoS success rate of 98%. Furthermore, it improves spectrum efficiency by 30% compared to existing multi-connectivity solutions.

Problem

• Optimizing QoS in B5G networks using multi-connectivity
• Predicting BS performance metrics via deep neural networks
• Selecting optimal serving clusters for UE QoS requirements

Innovation

• Machine learning for QoS-aware multi-connectivity
• Deep neural networks predict BS QoS metrics
• Optimal BS selection based on QoS predictions