5G Network Slicing

What is network slicing in 5G?

It can be defined as a network configuration that allows multiple networks (virtualized and independent ) to be created on top of a common physical infrastructure. This configuration has become an essential component of the overall 5G architectural landscape. Each “slice” or portion of the network can be allocated based on the specific needs of the application, use case or customer.

While services like smart-parking meters value high reliability and security are more forgiving with respect to latency, others (like driver-less cars) may need ultra-low latency (URLLC) and high data speeds. Network slicing in 5G supports these diverse services and facilitates the efficient reassignment of resources from one virtual network slice to another. 

The applications that are enabled or enhanced by the 5G need greater bandwidth, more connections and lower latency than was achievable with previous generations. Each use case will have its own unique performance requirements, making the one-size-fits-all approach to service delivery obsolete. 

Architecture 

Network slicing architecture in 5G is somewhat analogous to a complex public transportation system. Rather than rows of identical lanes and automobiles, some transportation elements (such as roads and bridges) are universal. However, other modes and vehicles are tailored to the speed, budget and volume requirements of the user. E2E network slicing (end-to-end) and logical isolation from other slices are central tenants of the architecture, although each unique slice transverses many common network elements. 

A network slicing SDN (software defined network) is an essential ingredient of the architecture used to manage traffic flows through the application program interfaces (APIs) of a central control plane. The control plane configures resources to deliver tailored services to the client through the application layer. SDN also includes an infrastructure layer, which contains basic network services and is responsible for data forwarding and rule processing from the control plane. The network slice controller (orchestrator) maps services and monitors the functionality between other layers.

It is through SDN virtualization that each client instance can unlock and orchestrate the specific resources that create a slice with the requisite service(s) included. Fulfilling these demands is a dynamic function that requires continuous monitoring of performance and isolation between slices. Recursion is another key feature of network slicing SDN that allows the control plane to create multiple sub controllers to support slice composition.

NFV (network function virtualization) is another prerequisite for slicing. The strategy behind NFV is to install network functionality onto virtual machines (VMs) on a virtualized server to provide services that traditionally ran on proprietary hardware.

NFV can also be used to manage the lifecycle of network slices and their infrastructure resources. The SDN is used to control the provisioning of VMs located in either edge or core clouds. These technologies working in concert can provide a sound foundation for efficient SDN and NFV network slicing utilization of virtual and physical resources including RAN’s.

3GPP Network Slicing

The 3GPP has recognized network slicing to be an essential overall component of 5G. This has made slicing an ongoing focus for working groups developing 5G core architecture with network slicing as an integral feature. 3GPP technical specification (TS) 23.501 defines stage 2 with slicing included, while TS 22.261 specifies the provisioning of network slices, association of devices to slices and performance isolation during normal and elastic slice operation. 

Release 16 of the 3GPP 5G specification includes 5G opportunities like low latency industrial IoT and autonomous driving. Among the study, items included in 3GPP release 16 are the 5G core solutions to enable cellular IoT and the bandwidth and cost implications of unlicensed NR spectrum.

Challenges and Opportunities

The benefits are self-evident, since a single network can be divided to cover diverse use cases based on customer demand and segmentation. Operators can then allocate resources to each slice, utilizing the necessary speed, throughput and latency to cover the breadth of network slicing in 5G. 

The ability to offer a network slice as a service minimizes operating expenses (OPEX) as well as capital expenditures (CAPEX). It can allow critical public entities, such as first responders and medical emergency teams, to be prioritized with respect to coverage, capacity and connectivity.

While network slicing is a requisite component of 5G, the opposite is not true. Unlike other fundamental elements of 5G, it can be deployed on existing 4G/LTE, so the benefits can be realized in the short term while preparing for the forthcoming transition. 

Through AI-powered orchestration, it also provides a secure and efficient alternative for the testing and deployment of new services. With the network divided, it is no longer necessary to implement changes that disrupt existing services in order to assess new ones. Fewer functions need to be deployed with each new slice. 

The Challenges
Despite the tremendous upside, there are still many challenges that remain for operators and developers. Full E2E network slicing includes implementation in the radio access network (RAN), but these RAN’s will need to be redesigned to accommodate network slicing. Although progress towards standardization continues, full industry consensus on network slicing deployment on 5G networks with other architectural elements remains ambiguous. 

The addition of more networks over the same physical infrastructure creates added stress for operators. Difficulty maintaining SLA, QoS and security assurance for each individual slice and managing the spectrum slicing and allocation for highly dynamic scenarios are among the common challenges.

Security

The complexity introduced also extends to network security considerations. Each slice will have unique security requirements commensurate with the use case it has been designed to provide and will require its own device authentication to validate users. Much like the IoT, the scaling factor associated with network slicing introduces billions of potential new attack vectors. A successful attack from a central 5G network management point could infiltrate many slices and/or network domains simultaneously.

To effectively address network slicing security concerns, the responsibilities between operators and enterprises utilizing the slice should be clearly established. Innovative strategies for security, such as micro-segmentation, continue to evolve to meet this need. Prevention and resolution of security issues will be an ongoing priority for 5G operators with substantial investment in security solutions expected, both before and after commercial launch. This provides opportunities for security-as-a-service (SECaaS) vendors with comprehensive solutions that support all stakeholders.

Use Cases 

Although the list of 5G applications seems limitless, most cases fit into one of three categories:

1. Enhanced Mobile Broadband (eMBB)
2. Massive Machine Type Communications (mMTC)
3. Ultra-Reliable Low Latency Communication (uRRLC)

Advanced applications such as Vehicle-to-Everything (V2X) combine elements of all three use case types because more features and functionality are being integrated each day. 

While the IoT is sometimes perceived as a single use case with a uniform set of requirements, the variety of potential IoT applications will drive surge of diversity into 5G network slicing. Some IoT devices will have flexible latency requirements, whereas remotely operated surgical or medical monitoring equipment can place a life-or-death premium on latency reduction. Public safety devices, such as smoke detectors and cameras, will require prioritized traffic routing and higher levels of security that can be enabled within the network slice.

The diversity of use cases highlights the need for E2E test practices. Essential practices include:

1. 5G core network emulation
2. network slice functionality verification and node selection
3. simulation of the variety and volume of devices

The TeraVM is an ideal software based tool for 5G security validation and application emulation. The TeraVM is 100% virtual and can emulate a city’s worth of 5G network subscriber activity and easily pinpoint flow problems and bottlenecks.

The TM500 test system is another potent solution for test challenges that can validate the network performance experienced by end users. The TM500 can support millimeter wave frequencies and large numbers of 5G UEs in either standalone or non-standalone modes. Both the TeraVM and TM500 are recognized as industry leading 5G tools and have been selected to perform the first wrap-around testing of 5G standalone (SA) base station equipment.

As commercial deployment of 5G is now a reality, network slicing is now center stage as a vital architectural lynchpin. What separates 5G from past wireless generations is not only the revolutionary speed and bandwidth improvement, but the various services that will drive wholesale changes in medicine, transportation, manufacturing and many other industries.

For each use case, network slicing can enable the ideal conditions as part of the network big picture. The new 5G highway will soon be overloaded with traffic in all shapes and sizes, so utilizing test tools to proactively verify functionality is a prudent course of action.  
 

Resources