VoLTE业务IP承载网潮汐流量预测模型建模与分析  

Modeling and Analysis of Tidal Traffic Prediction for IP Network of VoLTE Services

作　　者：于增源 YU Zengyuan(Heilongjiang Branch of China Mobile Communication Group Terminal Co. ,Ltd. ,Harbin 150028,China)

机构地区：[1]中国移动通信集团终端有限公司黑龙江分公司,黑龙江哈尔滨150028

出　　处：《无线电工程》2020年第12期1073-1081,共9页Radio Engineering

摘　　要：随着运营商VoLTE网络用户数的不断提升及业务应用场景的不断丰富,业务流量也呈高速增长趋势。IP承载网作为VoLTE业务主要承载网络,面对业务的高速发展,网络拥塞及服务质量较低问题逐渐凸显,潮汐网络现象日益严重。针对VoLTE用户服务质量要求和合理的调配资源和负载需求,提出基于注意力机制的BP-ANN神经网络模型,并考虑对流量数据的时间上的相关性,实现对IP承载网流量的预测。仿真结果表明,流量预测准确率基本达到85%以上。With the continuous increase in the number of VoLTE network users and the continuous enrichment of service application scenarios,the service traffic is also showing a rapid growth trend.IP network is the main carrier network of VoLTE services.With the rapid development of services,the problems of network congestion and low quality of service are becoming more and more serious.To meet the users’requirement for quality of service and the requirement for reasonable allocation of resources and loads,a BP-ANN neural network model based on attention mechanism is proposed.The traffic prediction of IP network is realized by considering the temporal correlation of traffic data.Moreover,the simulation results show that the accuracy of the proposed traffic prediction is above 85%.

关 键 词：VoLTE IP承载网 神经网络 潮汐现象 

分 类 号：TN927.2[电子电信—通信与信息系统]