网络智能,以“智”赋“动”
|
摘要
通过分析一种较为系统的思维模型,提出了在网络智能中"以智赋动"的观点。通过网元及用户智能识别、网络资源智能调度、网络智能保障、网络业务智能编排4个主要智能能力,分别赋予网络自动化用户通信环境适应能力、自动化网络资源优化能力、自动化故障修复能力、自动化业务适应能力。介绍了当前系统节能、智能无线资源管理(RRM)、边缘智能、智能网规网优、智能运维保障、网络安全6个应用级子方案。最后提出了MAPE-K、智能服务侧挂、模型驱动、三层共享等顶层设计思路,作为后续网络智能化架构设计的参照。
Based on the systematic thinking model,"artificial intelligence endows network automation"is put forward. Through the intelligent identification of network elements and users, intelligent scheduling of network resources, intelligent guarantee of network resources and intelligent arrangement of network business, the network automation users are endowed with communication environment adaptability, automation network resource optimization capability, automatic fault repair capability, and automation business adaptability. Then six sub-schemes of system are proposed, including energy saving,intelligent radio resource management(RRM), edge intelligence, intelligent network regulation network excellence, intelligent operation and maintenance guarantee, and network security. Finally, the MAPE-K, intelligent service side-hanging, model-driven,three-tier sharing and other top-level design ideas are proposed as the reference for the following intelligent network architecture design.
Based on the systematic thinking model,"artificial intelligence endows network automation"is put forward. Through the intelligent identification of network elements and users, intelligent scheduling of network resources, intelligent guarantee of network resources and intelligent arrangement of network business, the network automation users are endowed with communication environment adaptability, automation network resource optimization capability, automatic fault repair capability, and automation business adaptability. Then six sub-schemes of system are proposed, including energy saving,intelligent radio resource management(RRM), edge intelligence, intelligent network regulation network excellence, intelligent operation and maintenance guarantee, and network security. Finally, the MAPE-K, intelligent service side-hanging, model-driven,three-tier sharing and other top-level design ideas are proposed as the reference for the following intelligent network architecture design.
引文
[1]LILLIAN B.AT&T Exploring Artificial Intelligence for UAV Tower Inspections[EB/OL].(2018-1-10)[2019-01-15].https://unmanned-aerial.com/att-exploring-artificialintelligence-uav-tower-inspections
[2]KOLEY B.The Zero Touch Network.CNSM2016[EB/OL].(2018-01-10)[2019-01-15].https://research.google.com/pubs/pub45687.html
[3]Gartner.Gartner Inc.2017 Market Guide for AIOps Platforms[EB/OL].(2018-01-05)[2019-01-15].https://www.moogsoft.com/resources/aiops/guide/gartner-2017-aiops-marketguide/
[4]Cisco.2018 Intent-Based Networking Building the bridge between business and IT[EB/OL].(2018-08-31)[2019-01-15].https://securenetworkers.com/2018/08/31/intentbased-networking/
[5]ZTE.人工智能前沿[R].2019
[6]KEPHART J O,CHESS D M.The Vision of Autonomic Computing[J].IEEE Computer,36(1):41-50,2003
[7]IGLESIA D G D L.MAPE-K Formal Templates for Self-Adaptive Systems:Specifications and Descriptions[R].USA:ACM,2015
[8]Cognet.An NFV/SDN Based Architecture for Autonomic 5G Network Management using Machine Learning[EB/OL].(2018-01-10)[2019-01-15].http://www.cognet.5g-ppp.eu/wp-content/uploads/2015/08/ETSI-CogNetA0_Poster_final.pdf
[2]KOLEY B.The Zero Touch Network.CNSM2016[EB/OL].(2018-01-10)[2019-01-15].https://research.google.com/pubs/pub45687.html
[3]Gartner.Gartner Inc.2017 Market Guide for AIOps Platforms[EB/OL].(2018-01-05)[2019-01-15].https://www.moogsoft.com/resources/aiops/guide/gartner-2017-aiops-marketguide/
[4]Cisco.2018 Intent-Based Networking Building the bridge between business and IT[EB/OL].(2018-08-31)[2019-01-15].https://securenetworkers.com/2018/08/31/intentbased-networking/
[5]ZTE.人工智能前沿[R].2019
[6]KEPHART J O,CHESS D M.The Vision of Autonomic Computing[J].IEEE Computer,36(1):41-50,2003
[7]IGLESIA D G D L.MAPE-K Formal Templates for Self-Adaptive Systems:Specifications and Descriptions[R].USA:ACM,2015
[8]Cognet.An NFV/SDN Based Architecture for Autonomic 5G Network Management using Machine Learning[EB/OL].(2018-01-10)[2019-01-15].http://www.cognet.5g-ppp.eu/wp-content/uploads/2015/08/ETSI-CogNetA0_Poster_final.pdf