通算存智一体协同的未来网络模型OA北大核心CSTPCD
Future network model for integrated collaboration of communication,computing,caching and intelligence
针对智能时代多样化多形态业务对未来网络按需、精准、高效服务的诉求,提出了一种通算存智一体协同的网络模型(3CI-CoNet).首先以通信技术连接计算设备与存储设备,构建网内计算和网内存储的硬件底座,然后在底座上部署智能网络模型及算法,为网内业务构建智能的网络环境,同时链接与行业/应用相关的智能设备、平台及算法,为新兴应用提供智能的网络服务,保障各业务的有序高效运行.设计了3CI-CoNet的层次化功能架构,包含基础资源层、智能接入层、按需服务层和业务表达层,各层次间相辅相成,协同构建了智能的网络环境和网络服务.进一步,结合自动驾驶和智能制造中的典型场景,论述了3CI-CoNet及其功能架构、运营机制对多形态业务高效运行的支撑作用,阐述了3CI-CoNet在智能时代的推广和应用价值.以智能制造中多AGV按需调度业务场景为原型,通过将3CI-CoNet与其他网络模型相对比,定性定量论证了3CI-CoNet对提升业务效能具有积极作用.最后,总结了3CI-CoNet的未来发展趋势和面临的技术挑战.
In order to adapt to the demands of diverse businesses for on-demand,accurate and efficient network services,the communication-computing-caching-intelligence cooperative network(3CI-CoNet)was proposed.The hardware bases of 3CI-CoNet was built by connecting computing devices and storage devices with communication technologies.Intelligent models and algorithms were deployed on the bases to build an intelligent network environment.Simultane-ously,3CI-CoNet connected with intelligent devices,platforms,and algorithms relevant to various industries and applica-tions to provide intelligent network services for emerging applications,ensuring the orderly and efficient operation of various services.3CI-CoNet was designed with a hierarchical functional architecture,including the foundational resource layer,intelligent access layer,on-demand service layer,and business expression layer.These layers worked together to construct an intelligent network environment and network services.Furthermore,in conjunction with typical scenarios in autonomous driving and smart manufacturing,the functional architecture and operational mechanisms of 3CI-CoNet were discussed.Using the on-demand scheduling of multiple automated guided vehicle(AGV)in intelligent manufactur-ing as a prototype,a qualitative and quantitative analysis was performed to demonstrate the positive impacts of 3CI-CoNet on enhancing business efficiency.Finally,the future development trends and the technological challenges of 3CI-CoNet were summarized.
胡玉姣;黄韬;贾庆民;谢人超;刘韵洁
紫金山实验室未来网络研究中心,江苏 南京 211111紫金山实验室未来网络研究中心,江苏 南京 211111||北京邮电大学网络与交换国家重点实验室,北京 100876
电子信息工程
通算存智一体协同网络未来网络智能应用新型架构
3CI-CoNetfuture networkintelligent applicationsnovel architecture
《通信学报》 2024 (005)
12-28 / 17
国家自然科学基金资助项目(No.92367104,No.92267301);江宁百家湖计划基金资助项目(No.74072203-3) The National Natural Science Foundation of China(No.92367104,No.92267301),The Jiangning Baijia Lake Plan Program(No.74072203-3)
评论