摘要
内生智能的通信网络被认为是第六代移动通信网络发展的关键技术之一.在深入分析开发内生智能网络数据分析功能网元所面临的数据采集、隐私保护、模型管理以及灵活可扩展等挑战的基础上,提出一种具备并行化数据采集与处理能力、高效化模型训练与管理机制以及强容错性和可扩展性的内生智能网络数据分析功能网元开发架构.该架构旨在实现数据采集、数据分析、数据存储、模型决策一体化的目标,从而能有效应对第六代移动通信网络环境中的复杂需求.结合Kubernetes、流式化处理、微服务化等前沿技术,开发了实验室环境中的验证系统平台,进而验证了所提出架构的有效性并分析了系统性能.
Abstract
6G native intelligence is regarded as a key technology in the development of sixth-generation mobile communication networks.This study identifies critical challenges in developing network data analytics function network elements with 6G native intelligence,including data collection,privacy protection,model management,and scalability.Based on these insights,we propose a development architecture that features parallelized data collection and processing,efficient model training and management mechanisms,as well as robust fault tolerance and scalability.The architecture aims to integrate data collection,analysis,storage,and model-driven decision-making to address the complex demands of sixth-generation mobile communication networks effectively.A validation platform was developed in a laboratory environment using cutting-edge technologies such as Kubernetes,stream processing,and microservices.Experimental results confirm the effectiveness of the proposed architecture and provide a detailed performance analysis.
维普
万方数据