You are here
Does 5G Slicing Impact Transport Network Equipment?
The fifth generation of mobile communications promises to deliver ten-fold improvements in multiple parameters to address various new use cases. Such parameters include higher data rates, lower end-to-end latencies, support for connection densification, better timing accuracy, and more.
The good news is that most intended applications do not need the highest performance in all parameters and 5G is therefore expected to rely heavily on network slicing. That is the dynamic setting up and tearing down of fully isolated virtual networks, with each such network slice tailored to a specific application.
Various standards developing organizations (SDOs), such as the BBF and NGMN, are attempting to address the concept of network slicing in general, and its impact, if any, on the equipment within the transport network in particular. In this blog post, we will review various SDO perspectives of this issue.
What is network slicing? – NGMN’s approach
The term was initially coined by the Next Generation Mobile Networks (NGMN) Association, who defined the concept as “multiple logical self-contained networks on the top of a common physical infrastructure…” Typically, more than one device or user equipment (UE) may connect to a single slice, while a single device or user equipment may connect to multiple slices at the same time.
The NGMN proposes to address the implementation of network slicing by three main layers, (i) the service instance layer, (ii) the network slice instance (NSI) layer, and (iii) the resource layer. Each service instance reflects a service, while the network instance represents a set of abstracted resources customized to accommodate the performance requirements of the particular service.
The 3GPP SA5 work group sees network slicing as a means to transform the static “one size fits all” paradigm to a new one, where logical systems are created on-demand. This means that logical networks or partitions are created with appropriate isolation, resources and optimized topology to serve a particular service category or individual customer(s). The 3GPP also follows the NGMN network slice architecture and further defines the notion of instances:
- Network slice instance: A set of network functions and their resources, which are arranged and configured to form a complete logical network addressing certain network characteristics.
- Network slice subnet instance: A set of network functions and their resources for these network functions, which are arranged and configured to form a logical network.
What about the Broadband Forum (BBF)?
The BBF’s SD-406 on-going project sees network slicing as built according to six main principles that shape the concept and related operations:
- Automation enables on-demand configuration of network slicing without the need for manual intervention. Automation relies on signaling to allow third-parties to request a slice with the desired capacity, latency, jitter, security, etc. They also provide additional scheduling information, including start and end times, duration or frequency.
- Isolation facilitates performance guarantees and addresses security aspects within and between slices. However, isolation may reduce multiplexing gain, depending on the means of resource separation. Isolation relates not only to the data plane but also to the control plane and controller/orchestration systems.
- Customization allows the resources to be allocated to a particular tenant to meet specific service requirements. Again, this can be done on the network level, as well as on the data and control planes. Customization can be achieved via service-tailored network functions and data forwarding mechanisms, programmable policies and protocols, big data and context awareness, as well as specific management, controller/orchestration systems.
- Adaptability allows the resources associated with a particular slice to be modified in response to changing network conditions, user load, or geographical coverage. This can be achieved by scaling up/down or even relocating VNFs, or by adjusting policies. In addition, data and control plane elements can be re-program med to modify their functionality.
- Programmability uses application programming interfaces (APIs) to enable third-party players to control allocated network resources either via the infrastructure provider or by direct on-demand customization and adaptability.
- Multi-domain allows network slicing to stretch across different administrative domains, such as multiple infrastructure provides, as well as to unify heterogeneous resources, e.g., the radio access network (RAN), core network, transport domain, home networks and the cloud. In particular, network slicing consolidates diverse resources to enable an overlaid service layer, which provides new opportunities for fixed-mobile convergence.
Let’s now look what these principles mean for the transport network:
|Principle||Transport Network Requirements|
|Automation||Management stations and controllers should include a northbound interface and support multitenancy. In addition, each physical device need to be treated as several logical devices with network paths built using different devices.|
|Isolation||Support for multiple control planes with different entries, as well as for data isolation tunneling, e.g., VRF and VLAN.|
|Customization||Usually, network devices are forwarding traffic on the basis of IP or MAC addresses. Should the network slice require a different forwarding mechanism, then an SDN controller might offer a solution.|
|Adaptability||Where the data plane is concerned, this is mostly a management function, rather than a device function. Network function virtualization infrastructure (NFVI) enables efficient relocation of VM.|
|Programmability||Applies to network management|
|Multi-domain||Applies to network management|
In 5G, it is typically the UE that prompts the request to set up a network slice. But how would the transport network identify the slice? Should the slice information be part of the data packets, or transmitted via signaling between the transport and mobile networks?
Due to different slice granularities, the mechanism by which the mobile network identifies the slice is different from that of the transport network. In addition, mobile data consists of several hierarchies, such as user data and tunnels. It is therefore not recommended to mark slice information within the data. A better alternative would be for mobile network controllers to pass the policy/classifiers for each slice in each location to the transport network’s controller, which then passes it on to the devices. This requires the network elements to feature appropriate SDN capabilities, such as NETCONF/YANG support.
The bottom line
As we saw above, the concept of network slicing will definitely require changes in the management of transport equipment, but might not require significant changes in the handling of data by network devices, especially those in the access network. That said, there’s quite a number of SDOs – in addition to the ones mentioned above – that are currently hard at work in developing relevant frameworks. These include O-RAN, IETF, ONF, GSMA, TIP and the new 5G Slicing Association. So, all these aspects are still under development in the evolving 5G transport networks. Stay tuned!
If you’re in MWC Barcelona this week, stop by RAD’s booth to learn about 5G early adopters’ xHaul strategies.
About RAD's Blog
We’ll be blogging on a wide range of hot topics affecting service providers and critical infrastructure network operators. Our resident experts will be discussing vCPE, Cyber Security, 5G, Industrial IoT and much, much more.