The basic proposal of precise cloud-network is to focus on scenario-based, value-driven, and precise services of converged digital infrastructure. The core idea is to identify the pain points and key problems in 2B scenarios, and help telecom operators build brand-new cloud-network platforms to facilitate the digital transformation of industry customers through rapid iteration.
With the development of mobile internet, big data, artificial intelligence, cloud computing, IoT, edge computing, and other new technologies, human society is entering the fourth industrial revolution characterized by digitalization, 5G, automation and intelligence. The characteristics of 5G make it destined to go beyond the scope of connectivity, integrate with the cloud computing, AI, and big data technologies behind it, and become the cornerstone of the fourth industrial revolution. With the commercial launch of 5G and the acceleration of network construction, digital transformation implemented by the integration of 5G in vertical industries will become the key to future society. According to KPMG, the potential value of the 5G technology in the global vertical industry is estimated to be US$4.3 trillion.

The digital transformation of vertical industries differ greatly in network requirements covering network coverage, latency, bandwidth, reliability and security. The private networks currently used by vertical industries are limited and cannot meet their diversified requirements. At the same time, different industries have very different requirements for the cloud. The existing public cloud cannot meet the digital transformation requirements of industry customers. Based on this, ZTE proposes a package of cloud-network integration solutions named "precise cloud-network solution" to help telecom operators meet the huge market opportunities in the digital transformation of the industry.

The precise cloud-network uses "distributed precise cloud" and "deterministic precise network" to realize cloud on-demand and network scaling with the cloud, and provides comprehensive support with "end-to-end coordination", "simplified maintenance" and "endogenous security".

5G Deterministic Precise Network

5G deterministic precise network refers to the use of 5G end-to-end networks to provide deterministic, accurate, and SLA guaranteed mobile network capabilities to meet the needs of vertically differentiated services. Before 5G, the 2G/3G/4G network was a best-effort network, and the main target services were best-effort IT applications, lacking a precise SLA guarantee mechanism, and unable to undertake the requirements of interconnection of everything. In 5G deterministic precise network, deterministic SLA guarantees are achieved by wireless, transport, and core networks, and end-to-end orchestration (Fig. 1).

Precise Wireless Network 
The 5G wireless network guarantees precise network SLA through slice-level wireless physical resource blocks (PRBs). The wireless access network (RAN) can customize the physical private network (exclusive), hybrid private network (priority) or virtual private network (shared) for vertical industries. It integrates various advanced technologies such as multiple resource scheduling, QoS guarantee, delay&jitter control and access control to guarantee the bandwidth and service availability, meeting the different requirements of vertical industries precisely.

Precise Transport Network
Based on the existing SPN technical architecture, the TSN+ by adding new hierarchical TDM multiplexing small particles and hard isolation in the L1 implements the precise transport network. On this basis, the FlexE technology perfectly isolates the forwarding plane to ensure the high security of users, and the telemetry and in-band OAM technologies are introduced to provide real-time perception of network traffic and quality.

Precise Core Network
As the network brain for network perception, user perception, and business perception, the core network adopts a micro-service architecture, supports 2G/3G/4G/5G access, and supports 5G NSA/SA architecture, and uses software acceleration and hardware acceleration technologies to improve its performance. AI and machine learning (ML) are introduced to enhance the intelligence of operation and maintenance, so as to realize a precise core network.

Precise End-to-End Orchestration
The end-to-end orchestrator uses the big data engine to predict KPIs such as user behaviors and traffic within the specific slices, and network resources are dynamically scaled-in/out to ensure end-to-end network KPIs. At the same time, the root cause analysis (RCA) based on the AI engine can locate and handle the network faults automatically, so as to ensure that when some nodes in the network are faulty, the active/standby switchover is automatically triggered, and KPIs are guaranteed through remote disaster recovery and alternative routing.
In a specific industry application, deterministic precise networks do not always require the highest KPIs of all scenarios, but optimal costs while meeting the right requirements for industry application. For example, industrial control slices such as power grids require extremely high reliability and low latency; live HD video broadcast slices have high bandwidth requirements but general reliability requirements; and financial slices have the highest requirements for isolation and security. Precise networks provide three networking modes of "shared", "priority" and "exclusive" for all industry application scenarios, providing deterministic service quality, independent intelligent evolution capabilities, end-to-end network slicing, all-round security assurance, and full-process self-service.

Distributed Precise Cloud
Compared with the existing central public cloud, the distributed precise cloud has two distinct features. The first feature is distribution. By using operators' central office and sites all over the country and the 5G edge computing (MEC) capability, the precise cloud can be deployed in geographical areas near industrial customers. This achieves low latency processing and avoids the security and bandwidth usage problems caused by long-distance data transmission to central public cloud. ZTE proposes the site-level computing power engine solution named NodeEngine, which can implement site-level cloud deployment, so as to perform real-time edge cloud processing for local services in the place nearest to the enterprise campus. The NodeEngine solution ensures low-latency, reliable transmission of enterprise data without leaving the campus, avoiding the risk of leaking enterprise information, and helps the enterprise localize the digital application with the minimum cost and time.
The second feature of precise cloud is to locate industry needs precisely. The platform based on the dual virtualization technologies of virtual machines and containers is fully compatible with applications on mainstream cloud platforms. IaaS and PaaS can be customized and tailored according to different industry applications. At the IaaS level, the precise cloud can use different hardware acceleration cards (FPGA, GPU, ASIC, etc.) to achieve customized acceleration for different industry applications. Take e-game clouds for an example. GPU resource pools used for 3D rendering of e-games can be implemented in the edge cloud, so that even on ordinary mobile phones, you can experience high frame rates and extremely low delay control. At the PaaS level, the precise cloud integrates the capabilities of multiple general platforms such as AI engine, real-time database and big data engine. Based on the image recognition capabilities of the AI engine, it can be applied in general occasions such as face recognition for access control, and can also be customized into the automatic defect detection in manufacturing. Based on the ZTE GoldenDB distributed database engine, ZTE brings computing and expansion capabilities to the financial industry that traditional stand-alone databases cannot provide. ZTE GoldenDB provides database services featuring high availability, high reliability and flexible resource scheduling, and supports financial industry in business innovation and comprehensive digital transformation.

End-to-End Service Chain Coordination for Cloud-Network Convergence
Compared with the uncertainty brought by the traditional "best-effort network + best-effort cloud," the "5G deterministic precise network + distributed precise cloud" can guarantee end-to-end deterministic service experience. The precise cloud-network solution provides three capabilities: end-to-end service coordination, minimal maintenance, and endogenous security. End-to-end service coordination including cloud-network synergy, cloud-edge synergy, and edge-edge synergy realizes dynamic scheduling of overall resources on demand, elastic scaling, and delivers a senseless experience to customers. Minimal maintenance can be operated with one-click, so that the enterprise has full customization and management rights of their own. Endogenous security means constructing an active defense and customized security network. Through the introduction of a zero-trust architecture, the ultimate enterprise security guarantee is created in the three dimensions of network, data and users, and end-to-end security solutions are customized according to the characteristics of different industry customers.

ZTE: To Be an Ultimate Cloud Company 
ZTE aims to build itself into an ultimate cloud company. ZTE has started digital transformation of R&D a few years ago, and now it is currently using the concept of precise cloud network to upgrade and transform the entire R&D, sales, logistics, manufacturing and after-sales system. Through such components as the private cloud, public cloud, AI engine, big data engine, and orchestration center, we can build an agile customer oriented service chain to improve end-to-end working efficiency. We also use big data to provide insight into business opportunities, and accelerate online services to respond to customers in a timely manner. Nowadays, an all-in one digital application named iCenter enables mobile office, paperless office, office-free office and client-side office.
At the Nanjing Binjiang manufacturing base that was put into operation in early 2020, ZTE built a flexible 5G manufacturing line independently, which adopts the digital twin technology. A large number of robots/AGVs based on precise 5G network connection are used, which realize the fully automated manufacturing and packaging of 5G base stations. The AI-based visual inspection system on the precise edge cloud can automatically identify various product defects such as false welding, lap welding, and inadequate assembly, and truly realize the 24/7 "lights-off factory" of "manufacturing 5G with 5G".
ZTE has carried out comprehensive digital transformation exploration in many industries such as education, media, industrial manufacturing, mining, agriculture, health care, logistics and entertainment, enhancing the vertical industry's own value. Since 2019, ZTE has assisted 500 leading companies across 15 industries in 5G+ digital transformation.

The precise cloud-network is to help operators meet the huge market opportunities in digital transformation of vertical industries, give full play to the advantages of operators' networks, and build a new cloud-network integration platform. In the wave of digital economy, ZTE has positioned itself as a road builder of the digital economy, providing products and technologies for the industry, and working with industry partners to create a bright future of digital transformation.