I’m going to make a sweeping assumption here that most readers are involved in the telecommunications industry in some way and understand the migration to 5G to varying degrees. But it’s clear that many end users are very confused by 5G. Initially there was the promise of world-changing services such as augmented reality gaming, remote surgery, or 5G-connected sensors in our clothing giving us early warning of impending health concerns. But so far, while end users are impressed with the speed of 5G rollouts and the higher speeds achieved over 4G, they haven’t really seen anything beyond faster 4G service yet.
Most readers will probably know that this is because initial 5G services were built around the non-standalone (NSA) specification that was part of the 3GPP Release 15 specifications that added new 5G radios but kept the existing 4G core and transport network. The initial rollout of 5G involved a lot of work with new 5G radio/cell site deployments and sometimes new spectrum. But as I described in my most recent article for The Fast Mode, there is considerable work underway to prepare networks for those more advanced services that 5G has the potential to offer.
Progress with 5G SA
Key to understanding how 5G services are progressing is to understand where network operators are in the migration to full 5G standalone (SA) specifications that added the new 5G core. Data published by the Global mobile Suppliers Association (GSA) from March 2022 states that 201 operators across the globe had already commercially launched 5G services. Within this group of operators, 20 operators in 16 countries had already deployed 5G SA, with a further 25 in trials. So, only about 10 percent of the 5G operators were running 5G SA and around 12 percent were in its trial stages of 5G SA, but these numbers are expected to grow rapidly.
Initial 5G standardization coming to conclusion
And how have the 5G standards progressed since the freezing of the 3GPP’s initial Release 15 NSA and SA specification at the end of 2018? Release 16 continued to develop the 5G specifications by starting to add capabilities around low-latency communications, higher connection density, increased reliability, and private networks. Release 16 was completed and frozen in June 2020. Release 17, which is currently underway and scheduled to be frozen in September 2022, completes these additional capabilities and concludes the initial set of 3GPP 5G specifications, enabling operators to deploy a broad range of services that meet the initial goals of 5G.
Moving forward with 5G Advanced
Release 18 is already underway in the early scoping stages and will be the first 5G Advanced release when it is frozen in March 2024. Of course, release dates for standards are important, but it takes time for this functionality to get developed, tested, and ultimately deployed. We are therefore a few years away from any operator running a full 5G network with multiple classes of service, let alone a 5G Advanced network.
Does this mean that nothing is happening, and everyone is sitting back and waiting for Release 17 to appear in the feature lists from the vendors? Absolutely not! Pushing out the footprint of 5G networks with new 5G radios that can support Release 16 and 17 functionalities via software upgrades continues at pace. In addition, most operators today are upgrading the underlying optical transport network in preparation for more advanced 5G service offerings in Releases 16 and 17.
5G-driven transport network upgrades
Network upgrades vary considerably as no two transport networks are the same. Existing traffic levels vary considerably between operators as well as between rural, urban, and dense urban sites. Strategies around the placement of radio unit (RU), distributed unit (DU), and centralized unit (CU) components of the 5G radio access network (RAN) also vary between operators and within a network based on node type (rural/urban/dense urban), available spectrum, and available space and power, leading to a mix of front-, mid-, and backhaul domains.
Figure 1: 5G xHaul strategies and transport domains
There are some common threads that can be observed across these 5G xHaul transport network upgrades:
- Capacity upgrades. At the lowest-capacity end of the spectrum, network operators are upgrading from N x Gigabit Ethernet to 10G per macro cell site in preparation for 5G bandwidth growth. At the highest-capacity end of the spectrum, operators are evaluating new architectures that can deliver N x 25G per macro cell site. In all cases, the economics are very sensitive in terms of CapEx and also OpEx factors such as space and power costs.
- Higher performance. For various security and geopolitical reasons, many operators are also migrating to using the transport network as a primary or backup distribution network for timing and synchronization. 5G also drives considerably tougher synchronization performance requirements into the network. This is a particularly challenging requirement for optical networks, but one that can be resolved if the network is designed correctly.
- Virtualization, cloudification, and MEC. The trend towards virtualization of RAN components such as the DU and CU into software components, the move of this software into the cloud, and the increased use of multi-access edge compute (MEC) all have a significant impact on the transport network. MEC resources to support virtualized DU/CU and low-latency services require additional space and power in already constrained locations. In addition to dealing with the increased capacity outlined above, operators need a transport solution that actually frees up space and power for new MEC resources.
5G xHaul transport network options
To address the range of challenges and variety of traffic demands, network operators need a range of transport option choices. All must address the challenges of economically increasing capacity while providing any necessary additional features, such as synchronization. They ideally do not just minimize space and power requirements, but also free up space and power for new MEC hardware.
One optical networking trend that addresses these concerns is pluggable DWDM optics that are directly hosted in third-party systems such as routers, switches, and, potentially in the future, RAN equipment or MEC servers. Usually, these optics are used as part of the transport system, but advanced pluggable optics enable direct hosting in third-party devices, which reduces costs and frees up that all important space and power. Examples include lower-speed 10G autotuneable optics that are already being deployed in 5G xHaul networks today. In the very near future, these capabilities will be extended to higher-speed coherent XR optics running at rates between 25G and 400G. These new XR optics also bring a new point-to-multipoint architecture to the optical domain, which has the potential to revolutionize optical xHaul networks.
Figure 2: XR optics in action
So, returning to those folks from outside our industry who don’t think 5G has achieved much yet. From an end-services perspective, they aren’t wrong. Most 5G today is faster than 4G but does little else yet to offer next-generation services. What they don’t see is the huge amount of work currently underway to prepare transport networks across the globe for the full range of 5G services and ultimately upgrades to 5G Advanced. Network operators need a toolkit of transport options and should consider some of the advanced pluggable optics options if they aren’t already.
