Get Ready, 5G Is Coming – but Perhaps Not as Fast as You Think
April 16, 2020
By Jon Baldry
Director, Metro Networking
2019 was a great year for 5G, with network rollouts starting across the globe. Although important 5G showcase events, such as the 2020 Olympic and Paralympic Games in Tokyo, have been postponed due to the COVID-19 pandemic, 2020 still promises to be an even bigger step forward, as 5G network rollout continues to accelerate.
The world is now rapidly moving toward 5G, but true 5G, with new classes of service that weren’t possible with 4G, is still some way off, probably years for most of us. In this blog we’ll discuss the reasons for this staggered or phased approach to 5G services, the huge amount of work being done by network operators to prepare for these new advanced services, and the optical networking vendors that are supporting this migration.
First, let’s clarify what we mean by this phased approach to 5G.
There is a good chance you’ve seen the picture in Figure 1 below that explains the ITU-R’s view of 5G services – it was almost mandatory that any 5G-related presentation contain it for a while! It shows the three main focus areas of 5G services: enhanced mobile broadband (eMBB), which is essentially faster 4G; massive machine-type communications (mMTC) to support the ever growing Internet of Things (IoT); and ultra-reliable low latency communications (URLLC), which creates new service types with support for more demanding network performance in terms of low latency and reliability.
This diagram is sometimes criticized, as some of these focus areas, such as mMTC, are possible with 4G technology, although 5G extends their capabilities. Overall, though, the diagram is still a good representation of how mobile services will change from a single high-speed data service in 4G to a range of services with differing performance characteristics in 5G.
Figure 1: The International Telecommunications Union’s (ITU-R) view of 5G service types
The work to migrate to this range of 5G services is driven by the standardization bodies, and in this case the 3rd Generation Partnership Project (3GPP) in particular. Release 15 of the 3GPP’s series of standards in mid-2019 was the first 5G release, introducing non-standalone new radio capabilities and initial 5G functionality. This has enabled mobile operators to quickly introduce eMBB services, which are essentially faster 4G services.
This is why most of the mainstream media coverage of 5G launches usually involves a reporter using a speed test app to show how fast 5G is compared to 4G, but it’s nothing new from a service perspective other than faster speeds. 3GPP is now working on Release 16, which expands 5G with initial implementations of the full range of 5G services, and it is anticipated that R16 will be completed in June 2020. Going forward, 3GPP has already started to scope out the goals of Release 17, which will enhance these services further.
Figure 2: Phased introduction of 5G
The initial Phase 1 eMBB 5G services were comparatively quick to roll out as the bulk of the existing 4G network did not require any modification. The radio access network (RAN) in upgraded 5G cell sites needed to be updated, but the rest of the 4G network was not impacted. The backhaul transport network remined the same, as did all the complex processing within the core, hence the non-standalone name.
Some Phase 1 5G services to be showcased at the now-postponed Olympic and Paralympic Games include not only 5G coverage for spectators and the media teams covering the event but also 360-degree 8K video streams and virtual reality experiences for spectators, plus enhanced facial recognition-based security, all utilizing high-bandwidth 5G connectivity.
The move to Phase 2 5G services will see significant changes to the rest of the network, which will take some time to execute. Phase 2 requires a new standalone 5G core network and brings considerable architectural changes to the transport network that interconnects the RAN and the core.
Operators are currently upgrading and extending RAN capabilities to support Phase 1 capabilities over a larger service area and starting to upgrade transport networks in preparation for the additional bandwidth requirements and new architectures required to support Phase 2 services, including:
- Moving to new open and disaggregated networking architectures and initiatives. There are numerous open and/or disaggregated networking initiatives within the telecoms industry, with a range of goals that are usually associated with lowering the entry barrier for innovative new technology and lowering cost. A great example is the Telecom Infra Project (TIP), which is led by Facebook, Telefónica, Vodafone, and TIM, and their work on the Disaggregated Cell Site Gateway (DCSG) project. Infinera has been actively working with TIP and customers such as Telefónica to support this and other open disaggregation projects to support the rearchitecting of the transport network for the higher capacity demands of 5G.
- Initial migration to multi-access edge compute (MEC). URLLC services will deliver enhanced performance with lower round-trip latency than was achievable with 4G. In many cases, to achieve this lower latency, network operators need to move processing that was previously performed in the core of the network to locations closer to the end user via MEC, which essentially creates mini/macro data centers at midpoints in the network. This is a substantial network migration process, and a good example of the level of undertaking required to build an MEC-based network is Three in the U.K. Three has initiated a multi-year project that extends the Three UK core network to 20 major data centers as a first step toward a more distributed network to prepare for 5G services and to enhance existing 4G capabilities.
- Planning migration to xHaul. The revised transport architecture for 5G splits the previous 4G baseband unit (BBU) into two new components, the distributed unit (DU) and the centralized unit (CU). The vast majority of initial 5G deployments simply colocate both devices in the same location as the current 4G eNodeB or BBU for deployment speed and simplicity reasons. As 5G networks evolve to higher densities of cells and support for higher performance requirements, operators will start to migrate to the xHaul architecture outlined in Figure 3. This requires a more flexible transport network with better latency and synchronization performance than the previous 4G transport network. Network operators are now evaluating suitable products from optical networking vendors that support the range of capabilities required in these networks, such as Time-Sensitive Networking (TSN) and hardened equipment with hardened 100/200G optics for street cabinet deployments.
- Network migrations to free up resources. Networks are built with a finite set of resources – space, power, spectrum, and of course money. Operators always have to work within a constrained budget, but the demands of 5G are also driving network migration projects to free up the space, power, and radio spectrum resources that are needed.
For example, most countries have plenty of 3.5 GHz spectrum (C-band), which is often seen as a stepping stone to 5G, but the United States has issues as this band is already used by federal and satellite operators, pushing U.S. operators to jump directly to millimeter wave spectrum with reduced range performance. This has created a debate around migrating the federal and satellite users to new spectrum to free up the C-band for mobile operators.In other regions of the world, operators have started to sunset their 2G and/or 3G networks in order to free up the lower spectrum bands for 5G. Many operators, such as Verizon, are also migrating legacy transmission networks to more modern equipment with better density and lower power consumption in order to free up valuable space and power for 5G and other deployments.
- Increased use of network automation. The Verizon example above is a great example of the increased use of software automation tools to assist in the migration to 5G. The complexity of 5G networks themselves will almost certainly mandate the use of SDN-based network automation systems and apps. Beyond software automation, we are also seeing an increased demand for automation within networking hardware, as I discussed in my recent blog entitled “Jay-Z Says Auto-Tune Should Die. We Wholeheartedly Disagree!”
Figure 3: Mobile transport architecture migration from 4G to 5G
In summary, a lot of excellent work is being undertaken by network operators across the globe to roll out Phase 1 5G services and to prepare for the more advanced services that Phase 2 will enable. The move to widespread deployment of Phase 2 services will take time, but the end results will be worth the wait!