CN119383083A

CN119383083A - A three-layer decoupling method, system, device and medium based on 5GC technology

Info

Publication number: CN119383083A
Application number: CN202411440960.XA
Authority: CN
Inventors: 张磊; 杨经纬; 陈相如; 刘变红; 蒋彪; 史亚威
Original assignee: China Telecom Digital Intelligence Technology Co Ltd
Current assignee: China Telecom Digital Intelligence Technology Co Ltd
Priority date: 2024-10-16
Filing date: 2024-10-16
Publication date: 2025-01-28

Abstract

The invention discloses a three-layer decoupling method, a system, equipment and a medium based on a 5GC technology, which comprises the steps of forming the three-layer decoupling method by designing a physical hardware layer, a virtualization layer and an application bearing layer, wherein the three-layer decoupling method comprises decoupling of the hardware layer, decoupling of the virtualization layer and configuration of hardware resources according to requirements, the physical hardware layer comprises physical machines which are connected to form different resource pools under a network technology, the virtualization layer decouples the physical hardware layer through the network technology, the physical machines in the physical hardware layer are uniformly integrated with resources of the virtual machines according to requirements and then distributed to the virtual machines to complete decoupling of the virtualization layer, the virtual machines butt joint virtual network requests from the service bearing layer to configure the hardware resources according to the requirements, and the application bearing layer provides software application services for 5GC technical services through an application management system. The invention can provide technical support for the virtual machine with the optimal allocation performance according to the service, effectively solve the speed problem when the virtualization layer schedules the hardware resource, and improve the service efficiency of the virtual machine.

Description

Three-layer decoupling method, system, equipment and medium based on 5GC technology

Technical Field

The invention relates to the technical field of three-layer decoupling, in particular to a three-layer decoupling method, a system, equipment and a medium based on a 5GC technology.

Background

The traditional decoupling technology is mainly focused on two-layer decoupling, is used by operators and large-scale internet companies worldwide, but when the cloud computing technology is used, the performance release of a CPU (Central processing Unit), a GPU (graphics processing Unit) and storage can only reach 80%, the design of virtual decoupling is lacking, and for the scheduling of the traditional physical machine resource scheduling technology, the hardware resource scheduling efficiency is reduced due to the lack of the design of a resource pool. In addition, in actual use, the service related to cloud computing is real-time, and according to the importance of the service, the service is changed every moment, if only a cloud computing technology which is not flexible enough is used for bearing the service requirement, it is difficult to really allocate computing resources according to the requirement to meet the university support of the cloud computing technology on the service, and high-load and high-concurrency high-flow service can cause a plurality of blocked services, so that the linkage collapse of the system is caused. In order to solve the problem of high-efficiency utilization of resources, a three-layer decoupling design of a bottom architecture is adopted for a cloud computing technology, so that the performance of a cloud computing platform can be fundamentally improved, the loading capacity of the platform can be improved, and the safety of the platform can be improved by adopting self-grinding virtualization.

The three-layer decoupling technology of the 5GC technology is internationally an industry standard designed by European telecommunication Union, only China telecom is commercially landed in the world at present, and most of the 5G technology used in the market at present is a two-layer decoupling technology which is only a hardware layer and a software layer. The cloud computing three-layer decoupling framework based on the operator self-grinding virtualization mainly comprises a hardware layer, a virtualization layer and an application bearing layer. At present, the main stream processing mode in the industry is to encapsulate a hardware layer and a virtualization layer, and an application bearing layer is a single layer to realize double-layer decoupling, but in the continuous development of a 5GC technology, different hardware needs to release all performances, and the configuration customized for the virtualization layer is needed, so that the fitting degree of the hardware and the software is greatly increased. Because the virtualization layer is added on the basis of the two-layer decoupling technology, the complete virtualization of hardware is realized, the customized free configuration of hardware resources is realized, the technology is developed, the flexibility and the safety of the cloud platform are greatly enhanced, the safety of the hardware is also greatly improved, the cloud computing technology is always challenged in terms of safety, and if the cloud platform with high safety is provided, the cloud service is also developed for protecting the user.

Disclosure of Invention

The invention aims to solve the problem of inflexible hardware scheduling by adopting a two-layer decoupling architecture, and provides a three-layer decoupling method, a system, equipment and a medium based on a 5GC technology, which provide technical support for a virtual machine with optimal allocation performance according to services, effectively solve the speed problem of the virtual machine when the virtual machine schedules hardware resources in a virtualization layer, and improve the use efficiency of the virtual machine.

In order to achieve the above purpose, the technical scheme provided by the invention is as follows:

the first aspect of the application provides a three-layer decoupling method based on a 5GC technology, which comprises the following steps:

forming a three-layer decoupling method by designing a physical hardware layer, a virtualization layer and an application bearing layer, wherein the three-layer decoupling method comprises decoupling of the hardware layer, decoupling of the virtualization layer and configuration of hardware resources according to requirements;

The physical hardware layer comprises physical machines which are connected to form different resource pools under the network technology;

the virtual layer decouples the physical hardware layer through a network technology, and after the resources of the physical machines in the physical hardware layer are unified and integrated, the resources are distributed to the virtual machines according to requirements, so that the decoupling of the virtual layer is completed; the virtual machine interfaces with the virtual network request from the service bearing layer to configure hardware resources according to the requirements;

the application bearing layer provides software application service for 5GC technical service by an application management system.

In order to optimize the technical scheme, the specific measures adopted further comprise:

the virtual layer comprises a service gateway constructed by a VIM gateway and a PIM technology based on a 5GC technology and used for decoupling the virtual layer, the VIM gateway divides resources of a physical machine into a plurality of virtual resources and integrates the resources into a unified resource pool, and the PIM gateway provides support for the stability of the system by managing and scheduling the virtual resources.

The decoupling of the physical hardware layer is that the hardware resources of the bottom physical machine are virtualized into virtual resources through a service gateway and an Openstack technology, the Openstack technology configures the virtual resources into a plurality of virtual machines, and the virtual machines can be dynamically adjusted and deployed according to the service requirements of the application bearing layer.

The virtualization layer comprises virtualization software, wherein the virtualization software comprises a virtual driver, hardware simulation and Linux-Kernel, the Linux-Kernel comprises Kvm technology and SDN technology, the SDN technology provides network services for a software control network, the Kvm technology is used for interacting with the bottom physical machine hardware of the physical hardware layer, and single server resources obtained by Kvm are integrated in all directions through a service gateway and distributed according to requirements.

Further, the resources of the physical machine comprise CPU resources, GPU resources and network resources, and the physical hardware layer divides the hardware resources by a nano-tube technology to construct a resource pool.

The application bearing layer controls the VNM through the NFVO, the VNM can control the VIM gateway of the virtualization layer, the VIM gateway of the virtualization layer can be connected with virtual hardware with 5GC technology, the virtual hardware is in butt joint with Kvm technology through virtual drive and Linux-Kernel technology, and Kvm technology is connected with the bottom physical machine hardware resources in the physical hardware layer.

Further, the services in the application bearer layer are supported by virtual machines provided by Openstack, and the virtual machines can use resources in a customized manner without being limited by regions.

The second aspect of the application provides a three-layer decoupling system based on a 5GC technology, which comprises a physical hardware layer, a virtualization layer and an application bearing layer;

A third aspect of the present application provides an electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing a three-layer decoupling method based on 5GC technology as described above when executing the computer program.

A fourth aspect of the present application provides a computer-readable storage medium storing a computer program for causing a computer to execute the three-layer decoupling method based on the 5GC technique as described above.

Compared with the prior art, the invention has the beneficial effects that:

The invention provides a three-layer decoupling method, a system, equipment and a medium based on a 5GC technology, which are based on the three-layer decoupling technology, and a business system based on the 5GC technology and under a three-layer decoupling architecture is built by taking commercial landing as a scheme designed by an application scene.

The traditional method has the defects that the whole performance is difficult to truly and flexibly develop by using a two-layer decoupling technology, the high-flow application scene is not flexible enough when triggered, and a large amount of hardware resources are wasted because hardware cannot be fully utilized when a task is loaded. The three-layer decoupling technology can well solve the problem, breaks the region limitation in the hardware layer decoupling, can divide the uniform resource pools of multiple servers, solves the performance problem caused by continuous high load of the virtual machine when the virtual machine is used for dispatching the virtualized CPU resources, reduces the operation and maintenance cost, can be popularized in a plurality of resource pools, improves the physical machine efficiency, and solves the safety problem through the virtualization technology.

The invention adopts a three-layer decoupling technology to design a bottom layer framework of a cloud computing platform, adopts a network technology of a 5GC technology as a support for the whole system operation, designs a virtualization layer decoupling, adopts a plurality of different network technologies on the 5GC technology for adapting to a service, realizes seamless connection of a staggered and complex 5GC technology and a VIM, can truly realize resource pooling of hardware resources, can efficiently call all resources in the pool, adopts the 5GC technology to efficiently design a network structure, designs corresponding access configuration according to the service requirement, and safely, stably and efficiently solves the problems such as performance, efficiency and the like caused by two-layer decoupling.

The scheme of the invention can be applied to the performance improvement of a system under three-layer decoupling of a 5GC technology, can be applied to the performance improvement of cloud computing server equipment for cloud change and digital development by searching intermediate points in the conditions of three-layer decoupling and service limit operation, realizes balance and can maximize the value exerted by a cloud platform.

Drawings

FIG. 1 is a schematic diagram of a three-layer decoupling method based on 5GC technology according to the present invention.

Fig. 2 is a diagram showing a comparison of a three-layer decoupling method based on a 5GC technique of the present invention and a conventional method.

FIG. 3 is a block diagram of a three-layer decoupling system based on 5GC technology of the present invention.

Detailed Description

The above-described matters of the present invention will be further described in detail by way of examples, but it should not be construed that the scope of the above-described subject matter of the present invention is limited to the following examples, and all techniques realized based on the above-described matters of the present invention are within the scope of the present invention.

The application provides a three-layer decoupling method based on a 5GC technology, which comprises the following steps:

Forming a three-layer decoupling method by designing a physical hardware layer, a virtualization layer and an application bearing layer;

The virtualization layer decouples the physical hardware layer through a network technology, uniformly integrates the resources of the physical machines in the physical hardware layer, distributes the resources to the virtual machines according to requirements, and the virtual machines butt-joint virtual network requests from the service bearing layer;

The application bearing layer provides software application service for 5GC technical business by an application management system.

Some of the terms in the present invention are explained as follows:

GC technology is a network node configuration of needle virtualization hardware designed based on 5G technology.

In the application, the CPU obtained by the virtual machine is called virtual CPU, and the real calculation power is provided by the hardware at the bottom layer.

The 5GC technical business in the application comprises the service for 5G business, edge cloud, color ring and other applications.

The virtualization software consisting of the virtual drive, the hardware simulation and the Linux-Kernel and the VIM gateway together form a virtualization layer.

The VIM gateway is a virtual machine manager responsible for virtualizing physical resources into virtual resources.

The virtualization layer in the application decouples the physical hardware layer through the network technology, and distributes the resources of the physical machines in the physical hardware layer to the virtual machines according to the requirements after unified integration, the virtual machines can use the resources in a customized way without being limited by regions, and can realize resource scheduling from an upper application layer to a lowest server layer by interfacing with the NFVO (network function virtualization) technology and the VNM (virtual network function manager) technology of the application bearing layer.

SDN is a software controlled network, similar to a power system, the transmission of power from province to province relies on cables, while the process of sending packets between two ports controlled by SDN is similar to the transmission process of cables.

Kvm is a virtualization technology that interacts with hardware, and has wide application in VM ware and other software.

The application bearing layer controls the VNM through the NFVO, the VNM can control the VIM of the virtualization layer, the VIM of the virtualization layer can be connected with virtual hardware with a 5GC technology, and the virtual hardware is directly connected with hardware resources in a physical hardware layer through virtual drive and Linux-Kernel technology butt joint Kvm and Kvm, and can be hardware resources such as a server CPU, a GPU and the like which are far away from thousands of a user.

The application bearer layer is the top layer of the 5GC architecture and is responsible for running various 5G services and applications. This layer includes components such as cloud-native computing platforms, edge cloud services, and Network Function Virtualization (NFV) frameworks.

Compared with the traditional Kvm technology, the technology of the application is characterized in that a large amount of 5GC technology gateways are used, the resource utilization rights of hardware such as a CPU, a GPU and a network are obtained through Kvm technology, the gateway has the function of comprehensively integrating the single server resource obtained by Kvm and distributing the single server resource according to the requirement, each network can flexibly cooperate with actual production service, and the Kvm technology and the 5GC technology are combined to generate greater benefit.

The invention is further described in detail below with reference to specific examples:

The embodiment of the invention provides a three-layer decoupling technology based on a 5GC technology, which mainly comprises the steps of resource division of a hardware layer, self-grinding virtualization software, VIM gateway design, virtualization layer decoupling design and application bearing layer design as shown in figure 1.

(1) Resource partitioning at hardware layer

Before the method related by the application is started to be used, the technology of the application needs to be clearly known about what kind of hardware resource pool can be met.

The dependence on the network is mainly based on a network of a 5G technology, the delay-free access of servers between provinces is realized, the real servers are deployed in Shanghai in Sichuan in the current commercial perspective, and the unified nano-tube of a resource pool is realized.

The system has the advantages that the support is provided by the nano-tube technology, the servers are deployed in multi-provinces and multi-cities such as Guizhou, sichuan, shanghai, changsha, wuhan, zheng, hangzhou, etc., the nano-tube technology is adopted to uniformly bring the servers into a unified network, the coordination of server resources across provinces is realized, the remote business can be supported, the commercial server cluster is built in a remote mode, and the real-time effective and stable server support is provided for the business of the national level.

The construction of the resource pool realizes the construction of the cross-territory, and the application can support the nationwide nano-tube service by dividing different resource pools and different resource pools in the later period for facilitating the simultaneous support of different nationwide businesses, namely, the nationwide assembly server cluster can be flexibly selected as required, thereby providing a hardware foundation for the flexible construction of the businesses. The three-layer decoupling technology related to the technology of the application is firstly hardware layer decoupling, which mainly selects a machine with proper performance from the nationwide range according to business requirements of merchants, and adopts a nano tube mode to construct the machine into a resource pool.

In the aspect of cost, the method only needs to meet the smoothness of a 5G network and the uniform distribution of servers nationwide, so that the phenomenon that one city deploys millions of servers can cause network blocking is avoided, the servers are uniformly distributed to various places, network resources are balanced, and meanwhile, the cloud computing technology is provided with services.

The technology of the application also adopts the hardware to construct the resource pool, the traditional large-scale internet company comprises operators except telecom at present, after optimizing the hardware, the virtualization technology is directly adopted, the hardware and the virtualization technology are classified into one layer, and two layers of decoupling are constructed, so that resources are difficult to be freely allocated for sub-services of the follow-up service, only resources fixed by a physical machine can be used, and the release rate of the hardware performance is only 70%.

(2) Virtualization layer and virtualization layer decoupling

The three-layer decoupling technology mainly comprises hardware layer decoupling, then virtualization layer decoupling, wherein the hardware layer decoupling and the virtualization layer decoupling are separated, hardware resources can be freely and flexibly configured for project requirements of different provinces according to a 5GC technology after the virtualization decoupling, and the three-layer decoupling technology is constructed.

The self-grinding virtualization software mainly comprises three parts of virtual drive, hardware simulation and Linux-Kernel, and forms a virtualization layer together with the VIM gateway, and meanwhile, the self-grinding virtualization software is opposite to a two-layer decoupling technology.

The Kvm technology is to directly access physical machine hardware through Bios, particularly for control of a CPU, the core technology of Linux-Kernel is Kvm, SDN and the like, SDN is used as a software control network, a 1-3-layer network is provided for the SDN, and a connection service is provided for the physical hardware layer, the data link layer and the network layer by the SDN.

The self-grinding virtualization is carried out on Linux-Kernel, and technologies such as Kvm, hardware driving, SDN and the like are adopted to connect physical hardware, namely servers in all places of the country can be connected through the SDN technology, then the configuration of Bios is opened through the hardware driving, finally hardware such as a CPU is accessed through KVM, and hardware resources are obtained through a network mode.

The servers are connected with each other through a network, and service functions provided by different servers are different;

In some embodiments, the application provides network service through a server of the level of the springboard machine, the springboard machine can construct a resource pool according to the service, the technology such as Linux-Kernel can be obtained one by one, the application realizes the Linux-Kernel technology through the springboard server, then hardware simulation is developed on the springboard machine, the hardware simulation is to quantize the resource quantity of each physical machine connected through the Linux-Kernel through the SDN technology, and the quantized resource quantity is abstracted into virtual hardware.

If the virtual hardware needs to be configured by the 5GC technology, the virtual CPU, the GPU, the network and other virtual hardware can be better adapted to increase the compatibility of the hardware, the technology is called hardware simulation, and the hardware simulation needs Linux-Kerne to independently design corresponding interface components and then to be in butt joint with Kvm after Libvirt, so that the functions of the virtual hardware are realized.

The 5G core network (5 GC) three-layer virtualization architecture comprises a hardware layer, a virtualization layer and an application bearing layer, each layer has unique functions and responsibilities, and high-efficiency, flexible and extensible 5G network service is realized through close cooperative work.

Hardware layer: foundation stone of infrastructure

The hardware layer is the basis of the whole 5GC architecture and consists of physical devices, wherein the physical devices comprise a plurality of physical machines distributed in different geographic positions, such as Shanghai, sichuan and other areas, each physical machine is provided with rich hardware resources, including a CPU (Central processing Unit), a GPU (graphics processing Unit), a network interface, a storage device and other computing resources, and the hardware resources are the physical basis for the operation of a virtualization layer and an application bearing layer and provide computing, storage and network capabilities.

The physical devices are connected through a high-speed network, so that the rapid transmission of data and the high-efficiency utilization of resources are ensured, the physical machines in different cities can cooperatively work to jointly complete calculation tasks, the processing capacity and fault tolerance capacity of the whole system are improved, and when one physical machine fails, other physical machines can rapidly take over the work of the physical machines, so that the stability and the continuity of the system are ensured.

Virtualization layer resource abstraction and dynamic management

The virtualization layer is the core of the 5GC architecture, abstracts hardware resources into virtual resources through a virtualization technology, so that the resource utilization rate and flexibility are improved, and mainly comprises components such as virtual machine management, software Defined Networking (SDN), virtual resource management and the like.

A virtual machine manager (VIM) is responsible for virtualizing physical resources into virtual resources, a physical CPU can be divided into a plurality of virtual CPUs through Kernel-based Virtual Machine (KVM) technology, physical memory is distributed into a plurality of virtual memory blocks, physical storage devices are divided into a plurality of virtual storage units, and each Virtual Machine (VM) runs an operating system and corresponding application programs like a separate physical computer.

The virtual machines are mutually isolated, so that the respective safety and stability are ensured, the VIM allows the resource allocation of the virtual machines to be dynamically adjusted, virtual CPU, memory or storage capacity is increased or reduced according to actual needs so as to cope with different workloads, and the flexibility enables the resource utilization to be more efficient, and the waste of physical resources is avoided.

The SDN technology realizes the virtualization and dynamic management of network resources by separating a network control plane and a data plane, the SDN controller is responsible for managing the flow in a virtual network, dynamically adjusts the network configuration through software instructions to meet different service requirements, and when a certain application needs more bandwidth, the SDN controller can adjust the network bandwidth allocation in real time to ensure the performance of the application.

SDN also supports Network Function Virtualization (NFV), so that traditional network functions (such as routing, switching, firewall and the like) can be operated on general hardware in a software mode.

The virtual resource management is realized through cloud computing platforms such as Openstack, a Nova component of the Openstack is responsible for creating, starting and destroying a virtual machine, a Neutron component manages a virtual network, and a Cinder component provides virtual storage service. Through the components, the virtualization layer can realize unified management and scheduling of virtual resources, and high-efficiency utilization and dynamic adjustment of the resources are ensured.

Application bearer layer-realization and delivery of services

The application bearing layer is the top layer of the 5GC architecture and is responsible for running various 5G services and application programs, and comprises components such as a cloud primary computing platform, an edge cloud service, a Network Function Virtualization (NFV) framework and the like.

The cloud primary computing platform adopts a container technology and a micro-service architecture, so that development, deployment and operation of applications are more efficient and flexible, containers are of lighter weight than virtual machines, each container encapsulates one application and an operating environment thereof, so that the applications can operate uniformly in any place, and Kubernetes serves as a container arranging system and is responsible for managing creation, destruction, expansion and reduction of the containers, and high availability and elasticity of the applications are guaranteed.

The cloud native platform allows applications to be developed and deployed in the form of micro-services, each running independently and communicating with each other. The architecture improves maintainability and expandability of the application, so that a development team can quickly iterate and deploy new functions, and rapidly-changing market demands are met.

The edge cloud service deploys computing and storage capacity at the network edge close to the user, thereby reducing network delay and improving user experience, and video monitoring service, 5G Radio Access Network (RAN) and other applications requiring low delay and high bandwidth can be deployed on the edge cloud to provide quick response and high quality service.

The edge cloud service also supports local data processing and caching, reduces delay and bandwidth consumption of data transmission, and improves efficiency and performance of the whole system.

The NFV framework is enabled to run on a general hardware platform by software of network functions, the NFV comprises a Virtual Network Function Manager (VNFM), a VNFM instance manager (VNFMi), a Virtualization Infrastructure Manager (VIM) and other components, and the VNFM is responsible for managing and coordinating the lifecycle of the virtual network functions, including creation, configuration, monitoring and destruction.

The NFV flexibly deploys and manages network functions in a software-defined mode, when a certain network function needs to be upgraded, the software can be updated under the condition of no service interruption, the maintainability and the flexibility of the system are improved, meanwhile, the NFV reduces the dependence on special hardware equipment, and reduces the operation cost.

In the 5GC three-layer virtualization architecture, high-efficiency, flexible and extensible network service is realized through tight interaction and cooperative work among layers, a hardware layer provides physical computing, storage and network resources, a virtualization layer abstracts the physical resources into virtual resources through a virtualization technology, and a VIM component of the virtualization layer directly manages the virtualization of the hardware resources, for example, virtual machines are created and managed through a KVM technology.

When a certain virtual machine needs more calculation resources, the VIM can allocate more CPU and memory from the physical machine, thereby ensuring the normal operation of the virtual machine.

The cloud primary computing platform deploys and runs micro-service applications depending on virtual machines and container environments provided by the virtualization layer, the edge cloud service and the NFV framework also depend on virtual resources provided by the virtualization layer, the NFV needs virtual networks and computing resources provided by the virtualization layer so as to deploy and manage virtual network functions, and service requirements of the application bearing layer can be responded and met rapidly through resource management and scheduling mechanisms of the virtualization layer.

The service deployment and management of the application bearing layer comprise the steps that the application bearing layer realizes the rapid deployment and management of the service through a cloud native technology and an NFV framework, the video monitoring service can deploy a plurality of container instances on an edge cloud, kubernetes is responsible for the arrangement and management of the instances, high availability and load balancing are guaranteed, the NFV framework dynamically manages virtual network functions through the cooperative work of VNM and VIM, when the load of a certain network function is increased, the processing capacity of the virtual network functions can be expanded through the mode of increasing the virtual machine instances or the container instances, and meanwhile, an SDN controller dynamically adjusts network configuration, and high-efficiency transmission of data traffic is guaranteed.

Important role of virtualization layer

The virtualization layer plays a critical role in the overall 5GC architecture. The method not only realizes the abstraction and high-efficiency utilization of hardware resources, but also ensures the flexible allocation and optimization of the resources through a dynamic management and scheduling mechanism. The main roles of the virtualization layer are represented in the following aspects:

The virtualization layer abstracts physical resources into virtual resources through virtualization technology, so that multiple virtual machines or containers can share the same physical resources. Thus, the utilization rate of resources is improved, and the number and cost of physical devices are reduced; the virtual layer can dynamically adjust the allocation of virtual resources according to the actual service demands, and when a certain service load is increased, the processing capacity of the virtual machine can be expanded by increasing the virtual machine instance or the container instance.

The virtualization technique ensures isolation between different virtual machines or containers, preventing resource contention and security issues. Each virtual machine operates in an independent virtual environment, and even if one virtual machine fails, the normal operation of other virtual machines is not affected. The virtualization layer realizes high availability and fault tolerance of the system through dynamic adjustment and migration of resources. When a certain physical machine fails, the virtual machine can be automatically migrated to other physical machines, and continuity and stability of service are ensured.

The virtualization layer provides infrastructure support for the cloud native computing platform and the NFV framework, through the virtualization layer, cloud native applications can be rapidly deployed and expanded, NFV can flexibly manage and adjust network functions, and these characteristics enable the 5GC system to rapidly adapt to continuously changing business requirements and network environments.

The 5GC three-layer virtualization architecture realizes efficient, flexible and extensible 5G network service through the close collaboration of a hardware layer, a virtualization layer and an application bearing layer. The hardware layer provides physical resource support, the virtualization layer realizes the abstraction and dynamic management of resources, and the application bearing layer is responsible for realizing and delivering concrete services. The virtualization layer plays a key role in the whole architecture, and ensures the efficient operation and flexible adaptability of the 5GC system through resource abstraction, dynamic management, isolation and security, high availability and fault tolerance capability and support to cloud protogenesis and NFV. This three-layer architecture provides a solid foundation and broad prospects for the development of modern communication networks.

(3) VIM gateway design

The virtualization underlying technology related by the application is mainly a butt joint kvm technology, on the basis of the kvm technology, hardware resources such as a CPU (Central processing Unit), a GPU (graphics processing Unit) and the like on a physical machine far beyond kilowatt are abstracted into virtual hardware by a hardware simulation technology and a virtual driving technology, and the virtual hardware can serve on a trigger jumper beyond kilowatt.

The 5GC (5G core network) technology realizes network power saving and resource efficient utilization through a plurality of advanced technical means. The interrelationship between the components in 5GC technology will be detailed herein, revealing how they work in concert to provide efficient, flexible and reliable 5G network services.

The hardware layer is the basis of 5GC technology, providing physical computing, storage, and network resources. The virtualization layer abstracts the physical resources into virtual resources through VIM (Virtualized Infrastructure Manager) technology, so that efficient utilization and flexible management of the resources are realized.

The hardware layer includes a plurality of physical servers, each equipped with rich computing, storage, and network resources. Through VIM techniques, these physical resources are abstracted as virtual resources. The physical CPU is partitioned into multiple virtual CPUs and the physical storage is partitioned into multiple virtual storage units, each of which may be independently assigned to a different virtual machine or container. The VIM is responsible for managing the creation, allocation and destruction of these virtual resources, ensuring efficient utilization of the resources.

The virtualization layer integrates virtual resources into a unified resource pool through an Openstack platform. The Nova component of Openstack is responsible for managing the lifecycle of the virtual machine, including creation, startup, and destruction, the neutral component manages the virtual network, and the Cinder component provides virtual storage services. Through the components, the virtualization layer can realize unified management and scheduling of virtual resources, and flexible allocation and optimal utilization of the resources are ensured.

Interaction of virtualization layer with application bearer layer

The virtualization layer provides the underlying computing and storage resources that the application bearer layer uses to run various 5G services and applications. The interaction between the two is mainly reflected in aspects of resource scheduling, service deployment and dynamic adjustment.

The application bearing layer deploys and manages 5G business applications through the cloud native computing platform and NFV (Network Function Virtualization) framework. The cloud primary computing platform relies on a virtual machine and a container environment provided by a virtualization layer, and a Kubernetes and other container orchestration system is utilized to realize rapid deployment and expansion of services. The NFV framework manages the lifecycle of the virtual network functions through the collaborative work of VNFM (Virtual Network Function Manager) and VIM, ensuring efficient operation of the service applications.

When the service demand changes, the application bearing layer can adjust dynamic resources through the virtualization layer. When a certain application load increases, its processing power can be extended by adding virtual machines or container instances. The VIM component of the virtualization layer dynamically adjusts the allocation of virtual resources according to actual requirements, and ensures the continuity and high performance of the service.

Interaction of C-plane network and U-plane network

The C-plane network (control plane network) and the U-plane network (user plane network) are two important components in 5GC technology, and are respectively responsible for control and data transmission of the network, and the two components perform close cooperative work through interfaces and protocols.

The C-plane network is mainly responsible for managing and controlling various functional modules in the network, including AMF (access management function), SMF (session management function), PCF (policy control function), UDM (unified data management function), etc., and these functional modules communicate with UPF (user plane function) in the U-plane network through interfaces and protocols to cooperate.

The U-plane network is responsible for the transmission and routing of data. The UPF is a main processing unit of user data and is responsible for packet routing forwarding, policy implementation, flow reporting and QoS processing, the SMF in the C-plane network is responsible for managing session and tunnel establishment and maintenance to ensure efficient transmission of data, and the C-plane network and the U-plane network communicate through interfaces (such as N1, N2 and N3 interfaces) to ensure synchronization and coordination of control instructions and data flows.

The VIM gateway is a key component in 5GC technology, through which integration and management of hardware resources are achieved. The VIM gateway is similar to a socket, and can access different virtual hardware resources (similar to a plug), so that flexible management and scheduling of the resources are realized.

The resource nano-tube refers to a process of uniformly placing hardware resources packaged one by one into the same resource pool, can uniformly manage and schedule different types of hardware resources through the VIM gateway, ensures the efficient utilization of the resources, and can quickly access new hardware resources through the VIM gateway when the network capacity needs to be increased so as to realize the dynamic expansion of the resources.

The VIM gateway realizes the deployment and management of network functions through interaction with the VNM and other network function modules. The VNFM is responsible for managing the lifecycle of virtual network functions, including creation, configuration, monitoring, and destruction. The VIM gateway provides the resource support of the bottom layer, and ensures that the network function module can operate efficiently and be flexibly adjusted.

The cloud native computing platform and the NFV framework are two important components of an application bearing layer, and the cloud native computing platform and the NFV framework realize rapid deployment and efficient management of 5G services through close cooperative work.

The cloud native computing platform adopts container technology and a micro-service architecture, and manages the creation, destruction, expansion and reduction of the containers through a Kubernetes and other container orchestration system. The NFV framework manages the lifecycle of the virtual network functions through VNFM, ensuring efficient operation of the service applications. The two are communicated through the API and the control interface, so that unified scheduling and management of resources are realized.

The NFV framework allows it to run on a generic hardware platform by software-ing the network functions. The cloud-native computing platform provides the underlying computing and storage resources that the NFV framework utilizes to deploy and manage virtual network functions. When new network functions need to be deployed, corresponding containers or virtual machines can be quickly created through the cloud native platform, and the VNFM is responsible for managing its lifecycle and configuration.

The 5GC technology ensures the safety and reliability of the system through a multi-level safety mechanism and a high availability design. The components are communicated safely through an authentication and authorization mechanism, so that the integrity and confidentiality of data are ensured.

The AMFs and AUSF in the C-plane network are responsible for authentication and authorization of the user equipment, ensuring that only legitimate devices and users can access the network. The UDM provides unified data management and user identity authentication, and ensures the security and privacy protection of user data.

The UPF in the U-plane network is responsible for the transmission and routing of user data, and the transmission safety of the data is protected through an encryption and authentication mechanism. An SDN (software defined network) controller manages traffic in a virtual network, ensuring secure transmission of data and high availability of the network by dynamically adjusting network configuration.

The virtualization layer and the application bearing layer realize high availability of the system through dynamic adjustment of resources and a fault tolerance mechanism. When a certain physical device or virtual machine fails, the system can automatically transfer the load to other devices or virtual machines, so that the continuity and stability of the service are ensured.

The 5GC technology realizes network power saving and resource efficient utilization through the tight cooperative work of a hardware layer, a virtualization layer and an application bearing layer. The components are closely interacted through interfaces and protocols, so that the efficient operation and flexible expansion of the system are ensured. The virtualization layer integrates and manages resources through VIM technology and an Openstack platform, and the application bearing layer deploys and manages 5G services through the cloud native computing platform and the NFV framework. The C-plane network and the U-plane network work cooperatively through interfaces and protocols to ensure the synchronization and coordination of control instructions and data streams. Through a multi-level security mechanism and a high availability design, the 5GC technology guarantees the security and reliability of the system. The multi-level and tightly coordinated architecture design enables the 5GC technology to meet the requirements of high efficiency, flexibility and reliability of modern communication networks.

The 5GC technology is suitable for the C-plane network and the U-plane network, different servers are matched with different base stations, hardware under different networks can be timely and effectively scheduled, a VIM gateway is served, each virtualized hardware can be bound with one 5GC technology, data transmission is carried out through a 5G meganetwork card, and scheduling of hardware resources is achieved. The cloud computing is a technology for uniformly scheduling huge hardware resources, the packaging technology of the hardware resources of a plurality of physical machines is realized through the technology, then hardware is uniformly planned into a resource pool through a VIM (virtual memory) nano-tube technology, and 1000 servers provide CPU resources of 40000 cores and are close to the resource allocation of 80000 threads. The unified planning of the large-scale hardware is the most core technology of cloud computing, and the decoupling of hardware resources of a virtualization layer is almost completed by the time of the process.

(4) Virtualization layer decoupling design

The decoupling of the virtualization layer in the 5GC technology is a complex and systematic process, and the efficient utilization and flexible management of hardware resources are realized through a series of technical means and components. Specific steps of decoupling the virtualization layer will be detailed herein, revealing how it achieves virtualization decoupling and provides support for the traffic.

Before the decoupling of the virtualization layer, the hardware-level decoupling needs to be completed first. The method is realized through VIM (Virtualized Infrastructure Manager) gateway and Openstack technology, physical hardware resources are virtualized into virtual resources, and a foundation is laid for decoupling of a subsequent virtualization layer.

The VIM gateway is a key of decoupling of a virtualization layer, and implements virtualization and abstraction of resources by managing and scheduling physical hardware resources. The VIM gateway may partition CPU, GPU, storage, and network resources on the physical server into multiple virtual resources and integrate these resources into a unified resource pool.

Openstack provides the creation and management functions of virtual machines as an open-source cloud computing management platform. Through Openstack, virtual resources can be configured into multiple virtual machines, which can be dynamically adjusted and deployed according to business requirements. The Nova component of Openstack is responsible for lifecycle management of virtual machines, the Neutron component manages virtual networks, and the Cinder component provides virtual storage services.

After the hardware decoupling is completed, the decoupling of the virtualization layer follows. The core of the decoupling of the virtualization layer is that a virtual machine is used as an independent computing unit, and dynamic management and efficient scheduling of resources are realized through the cooperative work of the VIM and PIM (Platform Infrastructure Manager) gateway.

The PIM gateway is a special gateway which complements the VIM gateway and helps to share the problems of calculation force surge and insufficient performance caused by decoupling. The PIM gateway provides support for the stability of the system by managing and scheduling virtual resources. The introduction of PIM gateways enables the virtualization layer to manage virtual resources more flexibly and efficiently.

Through the cooperative work of the VIM gateway and the PIM gateway, unified management of virtual resources can be realized. When a new virtual machine needs to be created, openstack acquires corresponding hardware resources from the VIM and PIM gateway by configuring virtual machine parameters, and finally creates the virtual machine. These virtual machines serve as the main carrier of the service, running various 5G service applications.

Libvirt is a management platform of KVM (Kernel-based Virtual Machine) technology, which is adapted to the network requirements of 5GC technology by calling a KVM interface for secondary development. Libvirt manages hundreds of interfaces, can be well adapted to network modes of different areas, and realizes unified management of virtual machines in the whole country and free allocation of hardware resources.

The virtualization layer is constructed mainly based on 5GC technology, and provides connection and distribution ports for different hardware. Hardware resources far beyond thousand miles can be aggregated into a unified resource pool for scheduling via high-speed network connections. Libvirt manages these virtual resources, and through the API of secondary development, efficient management and scheduling of the C-plane network and the U-plane network are achieved.

The ultimate goal of the virtualization layer decoupling is to achieve the interfacing of the services and creation of virtual machines. Through the cooperation of Openstack, VIM and PIM gateways, virtual machines can be efficiently created and managed, and support is provided for 5G service applications.

Openstack enables seamless interfacing with VIM and PIM gateways. And configuring virtual machine parameters at the Openstack side, and acquiring corresponding hardware resources through the VIM and PIM gateway to realize the creation of the virtual machine. The virtual machine is used as a main carrier of the service, and can flexibly cope with different service demands.

After the virtual machine is created, dynamic adjustment and deployment can be performed according to service requirements. As traffic load increases, its processing power may be extended by increasing the number of virtual machine instances. The Nova component of Openstack is responsible for lifecycle management of virtual machines, ensuring high availability and stability of virtual machines.

After the virtual machine is created, various 5G business applications can be run on the virtual machine. Through the cloud native computing platform and NFV (Network Function Virtualization) framework, rapid deployment and efficient management of service applications can be realized. The edge computing service and the network slicing technology can be flexibly deployed on the virtual machine, and customized services can be provided for different business scenes.

In the virtualization layer decoupling process, security and high availability are critical. The stability and reliability of the system are ensured by a multi-level security mechanism and a high availability design, and a 5GC technology.

The AMFs and AUSF in the C-plane network are responsible for security authentication and access control of the user equipment, ensuring that only legitimate devices and users can access the network. The UDM provides unified data management and user identity authentication, and ensures the safety and privacy protection of user data.

The UPF in the U-plane network is responsible for the transmission and routing of user data, and the security of the data in the transmission process is protected through a data encryption and transmission security mechanism. An SDN (software defined network) controller dynamically adjusts network configuration, ensuring secure transmission of data traffic and high availability of the network.

The virtualization layer realizes high availability of the system through dynamic adjustment of resources and a fault tolerance mechanism. When a certain physical device or virtual machine fails, the system can automatically transfer the load to other devices or virtual machines, so that the continuity and stability of the service are ensured.

The decoupling of the virtualization layer in the 5GC technology is a complex and systematic process, and the efficient utilization of hardware resources and the flexible management of virtual resources are realized through the cooperative work of the VIM and PIM gateway. The Openstack is used as a core management platform, and through secondary development and Libvirt management, the virtual machine is created and managed, and a powerful support is provided for 5G service application. Through a multi-level security mechanism and a high availability design, the 5GC technology guarantees the security and reliability of the system. The decoupling of the virtualization layer not only improves the utilization rate of resources and the flexibility of the system, but also lays a solid foundation for rapid deployment and expansion of the 5G network.

(5) Design of application bearing layer

The technology of the application can obtain a plurality of virtual machines through the design of the hardware layer at the beginning and the structural decoupling design of the virtualization layer after the realization. The original double-layer decoupling technology is to create a virtual machine directly by virtualization on a local server. The hardware use is not flexible enough.

When the hardware is used, a locally created virtual machine is adopted, and at the moment, the service needs to use the 16 cores of the physical machine, the total of 20 cores are arranged in the physical machine, and the remaining 4 cores are not supported by three layers of decoupling, so that the remote connection of the hardware is not supported, the waste is caused when the user can only see that the 4 cores are open and run the service again, the three layers of decoupling technology can release the performance of hardware resources such as a CPU (Central processing Unit), a GPU (graphics processing Unit) and the like to 100%, almost all the cores of the CPU can be used, and the components of a resource pool are not influenced by the cores.

The services of the upper layer application mainly comprise video color ring, 5G message, edge cloud, voLTE viMS service, and the like, the services are supported by a virtual machine provided by Openstack, the middle is supported by adopting a VNM technology and an NFVO technology, the VNM technology can realize complete control of a VIM gateway, because some services are directly assisted by hardware, a plurality of VNM resources can be controlled by an NFVO scheduler, and then the VIM is controlled to directly use virtual hardware according to the 5GC technology, so that the application of real cloud computing level is completed. The number of CPUs in the resources of the resource pool can even reach 100000 cores. There are enough resources to create tens of thousands of virtual machines of different sizes and different performances for services to apply, and meanwhile, services with easy performance priority are needed, which can be directly scheduled through VNFM.

The NFVO orchestrator provides a cloud network operation system, a cloud resource operation subsystem, a resource and data sharing platform and a cloud acquisition and control platform. Taking hardware scheduling as consideration, the stability of the system is monitored at any moment, and meanwhile, the supporting condition of the cloud platform can be directly and freely adjusted according to different businesses.

In some embodiments, the application also provides a three-layer decoupling system based on the 5GC technology, which comprises a physical hardware layer, a virtualization layer and an application bearing layer;

The virtualization layer decouples the physical hardware layer through a network technology, and after the resources of the physical machines in the physical hardware layer are unified and integrated, the resources are distributed to the virtual machines according to requirements, so that the decoupling of the virtualization layer is completed; the virtual machine interfaces with the virtual network request from the service bearing layer to configure hardware resources according to the requirements;

In some embodiments, the application provides an electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the three-layer decoupling method based on the 5GC technology.

In some embodiments, the present application provides a computer-readable storage medium storing a computer program that causes a computer to perform the three-tier decoupling method based on the 5GC technique as above.

In the disclosed embodiments, a computer storage medium may be a tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. The computer storage medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of a computer storage medium would include one or more wire-based electrical connections, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

The pair of the three-layer decoupling method based on the 5GC technology and the traditional method is shown in fig. 2, the application scene of the method is a service system built under a three-layer decoupling architecture based on the 5GC technology, and the application scene of the method is a service scene of honest land, and has originality in the service scene.

The three-layer decoupling design breaks the region limitation by decoupling at the hardware layer, and can divide the uniform resource pools of multiple servers. Particularly, decoupling of a virtualization layer adopts a plurality of different network technologies on the 5GC technology for adapting to services, so that seamless connection of the complicated 5GC technology and VIM is realized, bottom virtualization is smoothly connected, and three-layer decoupling is truly realized.

The application solves the performance problem caused by the continuous high load of the virtual machine from the virtual machine to the virtualized CPU resource scheduling, reduces the operation and maintenance cost, is popularized in a plurality of resource pools in the whole country, improves the efficiency of the physical machine, and solves the safety problem by the virtualization technology.

The application provides a three-layer decoupling technology designed based on a traditional two-layer decoupling technology by using a KVM technology, and a network technology of a 5GC technology is adopted as a support for the operation of the whole system.

The present invention is not limited to the preferred embodiments, and any simple modification, equivalent replacement, and improvement made to the above embodiments by those skilled in the art without departing from the technical scope of the present invention, will fall within the scope of the present invention.

Claims

1. The three-layer decoupling method based on the 5GC technology is characterized by comprising the following steps of:

2. The method for three-layer decoupling based on 5GC technology according to claim 1, wherein the virtualization layer comprises a service gateway constructed by a VIM gateway and a PIM technology based on 5GC technology for performing the virtualization layer decoupling, the VIM gateway divides resources of a physical machine into a plurality of virtual resources and integrates the resources into a unified resource pool, and the PIM gateway provides support for stability of a system by managing and scheduling the virtual resources.

3. The method for three-layer decoupling based on 5GC technology according to claim 2, wherein the decoupling of the physical hardware layer is to virtualize the hardware resources of the underlying physical machine into virtual resources through a service gateway and an Openstack technology, the Openstack technology configures the virtual resources into a plurality of virtual machines, and the virtual machines can be dynamically adjusted and deployed according to the service requirements of the application bearer layer.

4. The three-layer decoupling method based on the 5GC technology according to claim 2, wherein the virtualization layer comprises virtualization software, the virtualization software comprises virtual drive, hardware simulation and Linux-Kernel, the Linux-Kernel comprises Kvm technology and SDN technology, the SDN technology provides network services for a software control network, the Kvm technology is used for interacting with underlying physical machine hardware of a physical hardware layer, and single server resources obtained by Kvm are distributed as required after being integrated in all directions through a service gateway.

5. The three-layer decoupling method based on the 5GC technology according to claim 1, wherein the resources of the physical machine comprise CPU resources, GPU resources and network resources, and the physical hardware layer uses a nano-tube technology to divide hardware resources to construct a resource pool.

6. The method of claim 4, wherein the application bearer layer controls the VNM through the NFVO, the VNM can control the VIM gateway of the virtualization layer, the VIM gateway of the virtualization layer can be connected with virtual hardware with 5GC technology, the virtual hardware is in butt joint Kvm technology through virtual drive and Linux-Kernel technology, and the Kvm technology is connected with the bottom physical machine hardware resources in the physical hardware layer.

7. The method for three-layer decoupling based on 5GC technology as recited in claim 1, wherein the services in the application bearer layer are supported by virtual machines provided by Openstack, and the virtual machines can use resources in a customized manner without being limited by regions.

8. The three-layer decoupling system based on the 5GC technology is characterized by comprising a physical hardware layer, a virtualization layer and an application bearing layer;

9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the 5GC technique-based three-layer decoupling method of any one of claims 1-7 when the computer program is executed by the processor.

10. A computer-readable storage medium storing a computer program for causing a computer to perform the three-layer decoupling method based on 5GC technique according to any one of claims 1-7.