专用网络生态系统：管理模式-新技术指南

May 2022 Contents Introduction04 Overall architecture of edge06native applications – componentsof ecosystem Value chain and operating model08 Resilience at the edge11 Ecosystem operations 13 Neutral Host Networks14 Market drivers15 Neutral host challenges16 Access to spectrum17 5G Delivery over radio networks18 Useful links and further reading21 Annex 1 22 Acknowledgements 23 Introduction This paper is targeted at organisations considering building out servicesusing 5G networking technology. It introduces and describes the principles ofneutral hosts, and then goes on to describe the architecture and ecosystemwhich supports the provision of shared services, particularly in the context ofhigh capacity/low latency applications, which will drive 5G deployment. Whilethis paper is focussed on 5G, many of the principles of neutral hosts, and thediscussion of edge versus core provision will apply to other technologies suchas Wi-Fi, including Wi-Fi 6. The full range of benefits that can be delivered over 5G requires other services to be provided inparallel. Devices specific to the application will need to be procured and deployed. Data will need to begathered and processed, either in the device itself, at the edge of the network or centrally, dependingon the specific use case. The provision of edge compute resources will enable a new generation of applications to bedeveloped. Theseedge-native applicationsdepend on both the high-speed and high-capacity featuresof 5G and the availability of low-latency compute without impacting backhaul networks. The valuechain for edge-native applications requires an ecosystem of providers and services, of which 5G is animportant component. Examples of such applications include: Services which require rapid location tracking, e.g., semi-autonomous robotics in a factoryenvironment or vehicle platooningServices which require very high bandwidths and low latency, e.g., Virtual Reality/AugmentedRealityServices that require processing very large amounts of data in time-limited applications, e.g.,high-definition video analytics for quality control, security, transaction generations, etc.Services where highly confidential data must be processed within a secure zone to ensure totalsecurity of information This paper focuses on 5G in the context of both private and public networks, recognising that someuse cases may cross both environments. For example, tracking a high value manufactured good (forexample an Electric Vehicle) through its creation in a factory (Private 5G), and then delivering servicesto it in-life (Public 5G). Many of the principles outlined in this paper can be equally applied to othercommunications technologies, in particular Wi-Fi, which is already widely used in private networks. The ecosystem consists of: Public network providers (MNOs)Fixed network providersData centre providers (including edge data centres)Private network providers (May be an MNO, a System Integrator, Neutral Host, equipment vendoror other provider of wireless telecommunications infrastructures)Wireless infrastructure operatorsDevice providersEdge cloud providersSpecialist application providers (specific to the use case)Provider(s) of ongoing support for the deployed solution This is not an exhaustive list, and specific providers may deliver in multiple aspects. For example, anMNO might offer a full solution. A System Integrator or Mobile Network deployment specialist mightprovide a solution bringing components together from various parties, bringing specific domainexpertise relevant to the customer. Overall architecture of edgenative applications – componentsof ecosystem The overall ecosystem that 5G is a vital partof provides a technical architecture andinfrastructure that can support edge nativeapplications on high-speed and high-capacitynetworks. The smartphones, vehicles, cameras,and devices used by consumers and businesseswill require access to a wide variety of cloudand edge services in order to fully exploit thepossibilities offered by edge native applications. In a distributed environment, system developersneed to make decisions on where to executefunctionality, and how to balance this betweenthe edge and the core elements. The edge cloud is likely to be most appropriatewhere latency is critical, or where very highvolumes of data needs to be analysed, but notfully stored. The core is likely to be most appropriate wheredata from multiple sources is aggregatedtogether over a period of time, or where staticdata is held in a system of record. The core willalso act as the central repository of data forcentralised processing of data from multiplesources, the majority of which may be at theedge. The core is also likely to be the venue fortransaction processing (e.g., payment systems,ledgers, etc.) for events originated in edgelocations – for example cashier-less retail. In amobility scenario, the core may also be used todeliver a subset of the services available at theed