摘要
为实现业务需求和网络能力的实时精准匹配,需要在网络端引入智能采集处理功能对交互数据进行分析处理,实现对6G网络状态的实时感知和估计预测。设计一种基于机器学习的网络状态感知分析方法,通过隐树模型的多层节点感知能力,能够根据通信网络历史数据,从数据层面对当前的通信网络状态进行量化感知,并对网络后续状态进行初步预测,可以为业务传输层的拥塞控制和移动网络层资源调度提供指导。
In order to achieve real-time and accurate matching between business requirements and network capabilities,it is necessary to introduce intelligent collection and processing functions on the network side to analyze and process interactive data,which can realize real-time perception,estimation and prediction of 6G network state.A network state-aware analysis method based on machine learning is designed.Specifically,through the multi-layer node-aware capability of the latent tree model,based on the history data of the communication network,the current communication network state is quantitatively perceived from the data level,and the subsequent network state is preliminarily predicted.The proposed method can provide guidance for congestion control at the business transport layer and resource scheduling at the mobile network layer.
作者
任育峰
REN Yufeng(The 20th Reseh Institute of CETC,Xi'an 710000,China)
出处
《移动通信》
2022年第9期30-34,共5页
Mobile Communications
关键词
网络状态感知
贝叶斯网络
数据分析
network state perception
Bayesian network
data analysis
