The Broadband Forum (BBF), like other industry bodies previously confined to a wireline environment, has played an increasingly important role in 5G. Last year it submitted specifications for key elements of a converged wireless/wireline architecture to 3GPP, and now it has published a 5G Network Architecture Overview.
This focuses on bringing the same intelligence and automation that is envisaged for the virtualized 5G RAN and core, to the transport network, along with the high levels of capacity and efficiency that 5G or converged data traffic volumes will require.
Adding 5G New Radio to a 4G site means up to 10Gbps backhaul capacity will be needed in many cases, said BBF, and that’s before fronthaul, with its very low latency demands, is considered. The new document provides the Forum’s view of the ideal broadband network architecture to support 5G at the transport layer.
This is an approach that moves well beyond the old idea of backhaul and transport as just fat but fairly dumb pipes. Instead, BBF contends that 5G will need to be able to support intelligent, coordinated automation from end to end. Disaggregation of the network, which introduces fronthaul and midhaul between different elements, as well as the separation of control and user planes, and the addition of Massive MIMO, are all 5G-related changes that put new strains on the transport network.
“5G is driving mobile operators to take a holistic approach to transport network planning,” said Robin Mersh, CEO of BBF. “The technology brings a significant increase in capacity, requires an estimated doubling of radio sites deployed, and the need for a new architecture with new RAN and core interfaces.These new architectures and new interfaces each have specific requirements that must be met not only by the mobile equipment, but by the underlying transport network.”
BBF estimates that a typical 4G distributed RAN site, with 5G NR added, will require up to 10Gbps backhaul capacity, while dynamic spectrum sharing (DSS), an important tactic for many operators to adopt from 2020 to improve spectral efficiency, will drive demand for many 10GE ports to deliver RAN coordination. And for fronthaul, an interface like enhanced CPRI (eCPRI) will need up to 25Gbps connections.
“As such, 5G backhaul baseband interfaces will need 10 Gbps capacity and will need to scale efficiently up to 100GE,” the white paper argues.
The BBF architecture incorporates the modern generation of transport network technologies, such as IPv6 and MPLS, combined with Layer 2 Ethernet technology, as depicted in the diagram.
“5G cannot exist without a transport network to support it,” said Joel Halpern, editor of the paper and an engineer at Ericsson. “Technology and application performance requirements will lead to a future transport network that looks very different. The use of 5G network architecture requirements, such as those being created within Broadband Forum, will ensure operators can enhance and migrate their existing networks to support the new capabilities 5G needs in order to deliver on the promise it has while protecting their existing investment.”
The work highlights how the 5G network will be as much made up of wireline technologies as wireless, and some operators are moving towards fully converged access and core as well. This was the driver behind BBF’s submission, in early 2019, of proposals for 5G fixed/mobile convergence to the 3GPP.
“Two years ago, 10 of the world’s biggest operators came to us with their concerns that the 5G infrastructure which was being developed would not let them take advantage of both their wireless and wireline networks,” said Geoff Burke, CMO of the BBF, at the time. “Today, we are able to deliver a set of recommendations which will not only address how the 5G core can be evolved to support the fixed networks of these operators but provide them with the capabilities to launch new innovative combined subscriber offerings.”
Among the resulting submissions were a number of common interfaces between the access and 5G core networks to support convergence of wireline and wireless, along with other recommendations on signaling changes and other enhancements, which would help integrate wireless access fully into the core. Some are included in Release 16, which was recently finalized, and other submissions are likely to be evaluated for Release 17.
It is encouraging to see substantial work resulting from a close cooperation between the 3GPP and another group, given its patchy track record in the past on developing joint standards with the IEEE and others. The BBF even has a wireless and wireline convergence work area director, Dave Allan, also of Ericsson.
He said in a statement: “As Release 16 progresses and 5G further expands, our work with 3GPP will continue to ensure standardization of this new technology is successful. With 5G promising a variety of new and innovative applications such as autonomous driving and healthcare, this work is crucial to enabling converged operators to unlock these opportunities – empowering them to deliver a uniform experience to their customers irrespective of the access media type, technology, or appliance they are using.”
David Aders, in charge of converged control plane engineering at Australia’s Telstra – which contributed to the proposals – said: “The alignment and collaboration between Broadband Forum and 3GPP are essential to delivering an ecosystem that is truly game-changing and providing a globally supported framework for operators such as ourselves to develop many next generation services.”
Another aspect of the BBF’s fixed/mobile convergence activity is to develop a specification for a 5G Access Gateway Function (AGF) that adapts fixed access onto the 5G core; as well as specifications for 5G-capable CPE; and for interworking of existing fixed access subscribers and equipment with a 5G core, in various deployment scenarios.
Georg Mayer, chair of 3GPP’s core networks and terminals group, added: “The recommendations from Broadband Forum allow us to develop a truly access-agnostic and common 5G core network, which will allow seamless user and service mobility between 3GPP and non-3GPP accesses.”