摘要
通过收集某小区7个月内小时粒度的业务数据,引入LSTM算法和Seq2Seq算法,预测未来72 h、168 h及336 h的下行PRB平均利用率及空口总业务量,并对预测结果分析,提出目前所存在的问题及未来规划,指出在未来数据充足的情况下可利用更多数据特征进行大区域多尺度多粒度的业务数据预测并指导网络资源调度及规划设计等。
By collecting the hourly granularity service data of a cell within 7 months,the LSTM algorithm and seq2seq algorithm are introduced to predict the average utilization rate of downlink PRB and the total traffic volume of air port in the future 72 h,168 h and 336 h.It analyzes the results and puts forward the existing problems and future planning,and points out that more data can be used in the future when the data is sufficient feature to predict large area multi-scale and multi-granularity business data and guide network resource scheduling and planning.
作者
刘旭峰
钟志刚
贾元启
王宁
史文祥
马家豪
Liu Xufeng;Zhong Zhigang;Jia Yuanqi;Wang Ning;Shi Wenxiang;Ma Jiahao(School of Information Engineering,Zhengzhou University,Zhengzhou 450001,China;China Information Technology Designing&Consulting Institute Co.,Ltd.,Zhengzhou Branch,Zhengzhou 450007,China;Joint International Laboratory of Intelligent Network and Data Analysis in Henan Province,Zhengzhou 450001,China;Harbin Institute of Technology,Harbin 264209,China)
出处
《邮电设计技术》
2021年第12期58-63,共6页
Designing Techniques of Posts and Telecommunications
基金
国家自然科学基金(61771431)
河南省重点研发与推广专项(202102210328)。
关键词
业务预测算法
网优
扩容
人工智能
Service prediction algorithm
Network optimization
Capacity expansion
Artificial intelligence
