5G network slicing: How to simplify and monetize through machine learning automation

Network slicing is a network architecture that enables the creation of multiple specialized and virtualized networks on a common shared infrastructure. Each of these network ‘slices’ can be monetized based on the specific needs and requirements of different sets of users and enterprise customers. As such, network slicing is critical to enabling the new revenue opportunities presented by 5G.

The challenges of 5G network slicing

While operators will indeed find many benefits and opportunities from network slicing, it’s important to note that it’s not without its challenges:  

• Slices that operate in a multi-vendor and multi-technology environment present great complexity.
• The slicing ecosystem is dynamic with slice requests often coming in on-demand.
• Heterogeneous multiband networks have different topologies in different areas — making configuring the slice, assigning resources, and assuring slices meet service level agreements (SLAs) an almost impossible task to do manually, especially in large-scale deployments.
  
  

How the Qualcomm Edgewise Suite can help manage network slicing

The Qualcomm Edgewise Suite presents the Radio Access Network (RAN) Slice Controller. This is an independent and vendor agnostic RAN Network Slice Subnet Management Function (NSSMF) solution, which interfaces with any Network Slice Management Function (NSMF) and any underlying RAN vendor through standardized open application programing interfaces (APIs). 

Qualcomm Edgewise Suite covers the full slice management lifecycle: from the requested order, to creation and activation (where Qualcomm Edgewise Suite rApps execute orchestration), service level agreement (SLA) monitoring, modification for performance optimization, and reporting. Furthermore, when the slice is no longer needed, Qualcomm Edgewise Suite also performs slice de-activation and termination.

As shown in the image below, when a new slice allocation is requested by the NSMF, the Qualcomm Edgewise Suite RAN Slice Controller receives the request (including the new slice characteristics), updates the slice inventory, and launches the slice creation flow. 

This flow incorporates a series of predefined rApps that are activated by the Recipe Orchestrator: 

• First, the Feasibility Check rApp evaluates whether there are enough RAN resources that can accommodate the new slice based on the slice request profile.
• Upon a successful feasibility check, the Recipe Orchestrator triggers the second rApp. 
• This is the Slice Configurator rApp, which then utilizes the slice profile to configure the network with the required slice parameters, as based on the operator’s business logic. 
• Once the slice is successfully configured, the Resource Allocator rApp is then executed.

With advanced machine learning, the Qualcomm Edgewise Suite RAN Slice Controller accurately translates slice throughput requirements into physical resource blocks and provisions them into the network. Once the slice configuration flow is complete, the Qualcomm Edgewise Suite Recipe Orchestrator activates the slice, and reports back to the NSMF. 

At this stage the Recipe Orchestrator allocates a monitoring instance that continuously tracks slice performance per slice, while regulating respective SLA commitments specific to each customer. This way, operators can be sure that no network element is overburdened with traffic, and subscribers are enjoying the services they want at the performance levels they expect.

The Qualcomm Edgewise Suite advantage

To achieve seamless, efficient and effective RAN slicing, enterprises can rely on Qualcomm Edgewise Suite to deliver these key capabilities:

• Multi-vendor network support.
• A hybrid approach to supporting the transition to Open RAN, as well as compatibility with any RAN — including traditional RAN, virtual RAN and O-RAN.
• Flexible Slice Designer that meets band, layer, macro/micro/indoor configuration requirements and more.
• Predictive slice admission that is driven by artificial intelligence and machine learning.
• Slice reconfiguration for existing slices in cases where new slices cannot be added.
• Root Cause Analyzer that enables the RAN NSSMF to identify the type of resolution apps that help regulate the customer’s SLA requirements. 
• Slice optimization for a multi-slice network with many different slice profile attributes, for improving performance of slice violations, while maintaining existing slices in the same cluster.

Now, you may be wondering:

Residential fixed wireless service: In suburban and rural areas that can’t be served by fiber, fixed wireless access (FWA) is great for delivering broadband internet access. However, for residential areas, FWA can place a great burden on the cellular network. With the Qualcomm Edgewise Suite RAN Slice Controller, operators can alleviate the burden with simplified network slicing that distributes the load and puts within reach customers that deserve high-quality service.

The mobile edge: The proliferation of autonomous vehicles and Internet of Things-powered devices, such as drones, has accelerated the need for delivering broadband services with continuous connectivity and minimized latency. But enabling the mobile edge with low latency can be very challenging. As such, edge computing is another prime case for network slicing with Qualcomm Edgewise Suite, which helps operators to bring the future of RAN technology today.

With a powerful combination of capabilities, the Qualcomm Edgewise Suite RAN Slice Controller simplifies 5G network slicing, while enabling operators to create slices for any RAN vendor and to connect to any slicing orchestrator, so they can accelerate time-to-market and capture the 5G monetization opportunity.