How to Orchestrate & Automate 5G Network Slicing?

We discuss the following topics in this blog:

1. Significance of Network Slicing
2. Radio Access Network (RAN) Support
3. Cloud-Native Architecture Support
4. What do Service Providers Need to do to Deliver Network Slicing?
5. Essential Attributes of Network Slice Orchestration Solutions

In addition to these topics, we shall also be answering the following FAQs:

1. What is 5G NR?
2. What is WiFi?

Overview

Leveraging the monetization opportunities made available by network slicing requires handling the 5G network as a single composite object. The network needs to be augmented by appropriate slice management and operations technology to disperse, designate, and lifecycle manage the network slices for enterprise use.

Automating cross-domain network slice management and operations is significant for economies of scale. Network automation is a central concern for maintaining 5G networks. The following provisions arise for the management and orchestration (MANO) layer:

What is the Significance of Network Slicing?

End-to-end network slicing necessitates a communication service management function (CSMF) that links to the customer service and produces allocation requests for a slice instance in addition to the network slice management function (NSMF) that takes care of the slice end-to-end.

The NSMF interacts with various network slice subnet management functions (NSSMF) for a particular domain such as RAN, core, or transport. The presence of these layers within or outside the MANO software is a structural choice.

Radio Access Network (RAN) Support

5G needs a few radio network components to be deployed through the Physical Network Functions (PNFs). It means that a 5G network service has to maintain PNFs and cloud-native network functions such as VNFs or CNFs3.

Furthermore, the MANO layer needs to maintain functionalities including Non-RealTime RAN Intelligent Controller (RIC) as stipulated by the O-RAN Alliance.

Cloud-Native Architecture Support

Conventional network services and their component VNFs have been set up in large data centers. However, in 5G networks, the 5G core and virtualized RAN will be highly dispersed. 

These elements are progressively veering towards a cloud-native architecture that demands Kubernetes as an NFVI.

Real-Time Analytics

It is imperative 5G networks optimize themselves in real-time. Keeping the volume in mind, this will be particularly useful in handling subscriber inquiries and network behavior.

Real-time analytics will be needed to impact day 1 and 2 configuration and lifecycle management actions such as scaling, fault supervision, productivity optimization, and others.

Network slicing enables communication service providers to provide differentiated slice-based services on a per-use-case basis. This offers enterprises control at the slice level and the required tools to self-provision connectivity services.

What do Service Providers Need to do to Deliver Network Slicing?

Implement end-to-end Network Slicing Orchestration:

Let us list out the basic features before we take a deep dive. Here are the basic features of an optimal network slicing solution:

• Cross-domain
• Linked end-to-end slice spanning the RAN, edge, transport, and core
• Support a model-driven multi-layer architecture
• Open APIs to provide required abstraction at the domain and cross-domain levels
• Multi-vendor support to service providers

For instance, a network slice using a CUPS-based topology will need:

• The instantiation of particular control plane network functions
  • SMF and PCF in the 5G core domain
  • Remote user plane functions (UPF) in the edge cloud data center domain
• Optimized transport network resources have to be set up in the backhaul domain

The Essential Attributes of Network Slice Orchestration Solutions

Apart from possessing an end-to-end cross-domain capability, the network slice orchestration solution must be constructed based on the following design principles.

Multi-layer architecture:

An elementary competence of this solution is the groundwork for multi-layer control at the domain and cross-domain levels. This will guarantee abstraction and operational distinction in a highly distributed mobile network architecture. This includes multiple networking data centers and cloud domains, including the RAN, WAN, edge, and core.

The lower-layer domain orchestrators will carry out the respective domain-level orchestration and software-defined supervision of these individual domains. The cross-domain orchestrator, however, will be the ‘orchestrator of orchestrators.’ It will render the higher-layer abstraction and end-to-end cross-domain service orchestration.

Model (or intent)-driven:

The blend of network framework in the mobile network ecosystem ranges from elements and conventional physical to VM-based, and the new container network functions for 5G standalone (SA).

These individual components will display a distinct level of sophistication. The orchestration solution should be capable of maintaining and promoting a model-based strategy. Such an approach will offer a high degree of personalization to meet diverse network function alterations and specifications. It will facilitate the development of new services and use cases by communication service providers without hinging on the singular units of the underlying network.

Platform-oriented and open API-driven:

Embracing a platform-oriented strategy will permit the solution to be developed in a functional form, be microservices-based, and simplify the inclusion of new capabilities. This will make the solution composable while admitting service augmentations with the slightest disturbance to the run-time environment.

The solution must adhere to open, industry-standard APIs for smooth solution composition and straightforward integration with third-party software elements to enable best-of-breed orchestration.

Furthermore, building the platform using selective tools for handling the DevOps and CI/CD pipelines to automate the development, testing, and deployment processes. Additionally, this will counter ‘tool sprawl’ and enhance the automation competence of solution evolution and deployment.

Multi-vendor support:

The cross-domain network slice orchestration (NSMF) solution must maintain and strengthen multivendor capabilities. It must be capable of effortless integration with any third-party domain-level slice orchestrators (NSSMF).

Vendor-agnostic domain-level orchestrators will be able to orchestrate a network comprising multi-vendor network equipment, VNFs, and CNFs.

As the networks grow increasingly virtualized, it is anticipated that few of the network roles will be hosted in a hybrid-cloud environment, either private or public.

It will necessitate multi-cloud slice orchestration competence from the private cloud settings within service providers’ data centers and public clouds such as AWS, Azure, and Google.

Automate network slicing orchestration

As more and more enterprises utilize network connectivity through network slices to promote new use cases, service providers must manage the incremental cost of generating fresh slices.

It can be accomplished by keeping a record of slice templates for the most sought-after use cases. It will ensure they can be rapidly and smoothly instantiated and provisioned, whether at the domain or cross-domain level.

Increasing adoption will require an extraordinarily skilled and cost-effective way to handle the lifecycle of adoption. Whether enterprises or communication service providers, there needs to be an innovative, intuitive way of directing, instantiating, provisioning, tracking, and handling the lifecycle of network slices.

Communication service providers must consider the economics of deploying a slice and how far they can regulate the accumulating expense of implementing tens or hundreds of more slices.

Enterprises must examine their agility in managing their slices and strengthen them before rapidly proposing new service capabilities. High levels of automation of network slice management and operations is a decisive factor in reaching the essential economies of scale.

The Final word

With particular services, separate network slices need to be developed to meet their strict SLAs. STL orchestrator creates multiple logically independent network slices to maintain these specific services and handles them effectively.

STL has developed a model-driven orchestrator to manage the lifecycle of the 5G network functions along with provisioning services and provide closed-loop control. It will perform various functions ranging from mapping service-related requirements and managing the multiple lifecycles of the slice to creating independent slices within the network and more.

Reach out to us for more information.

FAQs

What is 5G NR?

5G typically refers to the fifth generation of wireless technology. NR, commonly known as New Radio, is a standard developed by the 3GPP Group (Release 15 being the first version introduced back in 2018) outlining the technology required to harness the newly-available millimeter-wave frequencies. The two frequency bands in which 5GNR operates are Frequency Range 1, i.e., Sub 6GHz band (410 MHz to 7125 MHz), and Frequency Range 2, i.e., millimeter-wave (24.25 to 52.6 GHz). Over 4G LTE, 5G NR provides better spectrum utilization, faster data rates, hardware efficiency, and improved signal processing.

From a deployment standpoint, we have Non-Standalone Mode(NSA), Dynamic Spectrum Sharing(DSS), and Standalone Mode (SA). The initial deployments of 5G NR are based on NSA standards, meaning the existing 4G LTE network will operate on the control plane, and 5G NR will be introduced to the user plane. This particular standard was introduced by 3GPP, keeping in mind the industry’s push to faster 5G services rollout while utilizing the existing 4G LTE infrastructure currently in place. On the other hand, operators are also implementing Dynamic Spectrum Sharing (DSS) to accelerate the deployment cycle, reducing costs and improving spectrum utilization. In this standard, the same spectrum is shared between the 5G NR and 4G LTE, multiplexing over time per user demands. Lastly, we have the Standalone Mode (SA), which moves towards a complete 5G based network where both signaling and the information transfer are driven by a 5G cell.

In the future, 5G will enable new services, connect new industries and devices, empower new experiences, and much more, providing mission-critical services, enhanced mobile broadband, and various other things.

a) Enhanced mobile broadband (eMBB) Applications: High device connectivity, High mobile data rates, and Mobile AR & VR applications
b) Ultra-reliable, low-latency communications (uRLLC)Applications: Autonomous vehicles, Drones, Data monitoring, Smart mfg.
c) Massive machine-type communications (mMTC)Applications: Healthcare, Industry 4.0, Logistics, Environmental monitoring, Smart farming, Smart grids

What is WiFi?

Put simply, WiFi is a technology that uses radio waves to create a wireless network through which devices like mobile phones, computers, printers, etc., connect to the internet. A wireless router is needed to establish a WiFi hotspot that people in its vicinity may use to access internet services. You’re sure to have encountered such a WiFi hotspot in houses, offices, restaurants, etc.

To get a little more technical, WiFi works by enabling a Wireless Local Area Network or WLAN that allows devices connected to it to exchange signals with the internet via a router. The frequencies of these signals are either 2.4 GHz or 5 GHz bandwidths. These frequencies are much higher than those transmitted to or by radios, mobile phones, and televisions since WiFi signals need to carry significantly higher amounts of data. The networking standards are variants of 802.11, of which there are several (802.11a, 802.11b, 801.11g, etc.).