5G边云协同实时数据计算架构的研究

A study on architecture of edge-cloud collaborative real-time data computing in 5G

实时计算在通信技术(CT)领域被广泛应用在运营支撑系统(OSS)域位置类计算、网络(关键业绩指标)KPI指标、业务支撑系统(BSS)域实时订单归集以及实时指标计算等方面;而5G网络架构的变革对实时计算提出了新的要求,部分超低延时的计算将从数据中心下推至边缘;受限于边缘计算中心的计算资源,实时数据中台系统需要合理地将实时计算类型在边缘和中心协同管理与编排;同时,实时计算也需要与5G网元等系统进行相互操作,充分利用网络能力的演进提升实时计算能力而增强对更多场景的适用性。基于这样的背景,提出了一种面向5G特性的边云协同实时计算架构,将传统中心化的实时处理扩展为边缘、云端协同实时处理,从而适应5G业务快速响应、实时决策的需求,提升实时处理的效率。

Real time computing is widely leveraged in location computing,wireless network key performance indicator(KPI)computing,bill converging and business support system(BSS)KPI analyzing.The reform of 5G network architecture raised new requirements for real-time computing.The part of ultra-low latency computing will be pushed down from data center to edge.Constrained by resource on the edge,real-time’mid-end’data system needs manage and orchestrate the real-time computing in a collaborative approach between edge and cloud.Meanwhile,real-time’mid-end’data system also needs to interoperate with 5G network elements and other systems,making full use of the evolutive network capabilities to promote real-time computing capabilities and enhance applicability to more scenarios.Based on this background,an architecture of 5G oriented collaborating real-time computing between edge and clouds was put forward,extending the traditional centralized real-time processing to edge and cloud collaborative real-time processing in order to adapting more rapid response and real-time decision features of 5 G service,and improving the efficiency of real-time processing.