The application of network slicing in 5G

5G and network slicing

When 5G is widely mentioned, network slicing is the most discussed technology. such as KT, SK Telecom, China Mobile, DT, KDDI, NTT and other network operators, as well as Ericsson, Nokia, Huawei and other equipment vendors think network slicing is the ideal network architecture of the 5G era.

This new technology allows operators to cut off multiple virtual end-to-end networks in one hardware infrastructure, each of which is logically isolated from the device to the access network to the transmission network to the core network, adapting to the different characteristics of various types of services.

For each network slice, such as virtual server, network bandwidth, service quality and other exclusive resources are fully guaranteed. Because slices are isolated from each other, errors or failures in one slice do not affect the communication of other slices.

Why 5G requires network slicing

From the past to the current 4G network, mobile network main service mobile phone, generally only for mobile phone to do some optimization. In the 5G era, however, mobile networks need to serve devices of all types and needs. A number of applications include mobile broadband, large-scale internet of things, and mission-critical IoT. They all need different types of network, in terms of mobility, billing, security, policy control, latency, reliability and other aspects have different requirements.

For example, a large-scale IoT service connects fixed sensors to measure temperature, humidity, rainfall, and more. There is no need to switch, location update and other features of the mobile network for those major service phones. In addition, mission-critical IoT services such as autonomous driving and remote control robots require a few milliseconds of end-to-end latency, which is quite different from the mobile broadband business.

The main application scenario for 5G

Does this mean that we need to build a dedicated network for every service? For example, a service 5G mobile phone, a service 5G large-scale internet of things, a service 5G mission-critical Internet of things. Not really, because we can cut out multiple logical networks on a separate physical network through network slicing technology, which is a very cost-saving practice!

Application Requirements for network slicing

The 5G network slices illustrated in the 5G white paper published by NGMN are as follows:

NGMN 5G Network Slices How do we implement end-to-end network slicing? (1) 5G Wireless access Network and core network: NFV

In today's mobile network, the main device is the mobile phone. The Ran (Du and RU) and core functions are built by the dedicated network equipment provided by the RAN vendor. For network slicing, network function virtualization (NFV) is a prerequisite. Basically, the main idea of NFV is to deploy network function software (that is, MME,S/P-GW and PCRF in the packet core and Du in the RAN) to virtual machines on a commercial server, rather than deploying them on their own private network devices. In this way, ran is used as the edge cloud, and the core function as the core cloud. Connections between VMs located on the edge and in the core cloud are configured with Sdn. Then, create slices for each service (that is, phone slices, large-scale IoT slices, mission-critical IoT slices, and so on).

How to implement one of the network slices

Shows how each service-specific application can be virtualized and installed in each tile. For example, a slice can be configured as follows:
(1) UHD slicing: virtualizing the du,5g core (up) and cache servers in the edge cloud, as well as virtualizing the 5G core (CP) and MVO servers in the core cloud
(2) Phone slicing: virtualize 5G core (up and CP) and IMS servers with full mobility in the core cloud
(3) Large-scale IoT slicing (e.g. sensor networks): Virtualizing a simple, lightweight 5G kernel in the core cloud no mobility management capabilities
(4) Mission-critical IoT slicing: virtualize 5G cores (UP) and related servers (such as V2X servers) in the edge cloud to minimize transmission latency

So far, we need to create dedicated slices for services with different requirements. The virtual network functions are placed in different locations in each tile (that is, the edge cloud or the core cloud) based on different service characteristics. In addition, some network features such as billing, policy control, and so on, may be necessary in some slices, but are not required in other network slices. Operators can tailor their network slices in the way they want, and may be the most cost-effective way.

How to implement network slicing between the two (2) Edge and core cloud network slices: ip/mpls-sdn

The software-defined network, although a very simple concept when introduced first, is now becoming more and more complex. In the case of overlay, SDN technology is able to provide network connectivity between virtual machines on existing network infrastructures.

End-to-end network slicing

First we look at how to ensure that the network connection between the edge cloud and the virtual machine in the core cloud is secure, and that the network between the virtual machines needs to be based on IP/MPLS-SDN and Transport Sdn. In this paper, we mainly discuss the IP/MPLS-SDN provided by router manufacturers. Both Ericsson and Juniper offer IP/MPLS SDN network architecture products. The operations are slightly different, but the connectivity between SDN-based virtual machines is extremely similar.

In the core cloud is the virtualized server. In the server's hypervisor, run the built-in Vrouter/vswitch. The SDN controller provides a tunnel configuration between the virtualized server and the DC g/w router (the PE router that creates the MPLS L3 VPN in the cloud Datacenter), creating an SDN tunnel (that is, MPLS GRE or Vxlan) between each virtual machine in the core cloud (such as the 5G IoT core) and the DC g/w router.

Then, the SDN controller manages the mapping between these tunnels and the MPLS L3 VPN (for example, IoT VPN). The process is the same in the edge cloud, creating an IoT slice that connects from the edge Cloud to the IP/MPLS backbone and to the core cloud. This process can be done based on the technologies and standards that are available so far.

(3) network slicing between edge and core cloud: IP/MPLS-SDN

Now all that's left is the mobile forward network. How do we cut this mobile forward network between the Edge cloud and 5G ru? First, you must first define a 5G forward network. There are some options in the discussion (for example, by redefining Du and RU to introduce new packet-based forward networks), but no standard definitions have been made. is a diagram given in the ITU IMT 2020 Working Group, and gives an example of a virtualized pre-transmission network.

ITU organization 5G C-ran Network slicing example

5G era of network slicing technology is still evolving, some concerns and problems remain unresolved. As a result, we will track technical updates from operators, suppliers and standardization organizations in Korea and around the world, so you can get the latest information about the technology.

original link:http://www.netmanias.com/en/?m=view&id=blog&no=8325

• Site Statement : This site original article only represents the author's opinion, does not represent Sdnlab position. Reprint must indicate from Sdnlab and attach this article link.

The application of network slicing in 5G