If 5G is to deliver on any of its large promises, it will have to signal the end of ‘one size fits all’ for wireless networks. The latest in the long string of international 5G projects, launched by the European Union’s PPP (public-private partnership), has this firmly in its sights, which gives cause for optimism that it will deliver something useful.
The initiative has one of the tongue-twisting names beloved of the EU – 5G Novel Radio Multiservice adaptive network Architecture – but at least this comes with a neat acronym, NORMA – although the lower case ‘adaptive’ sadly downplays the most important world in the title. A network that can adapt to many services and usage patterns, some of them as-yet undefined and even unforeseen, is the goal, and will be essential to 5G.
The 5G PPP, launched in December 2013 to support and coordinate R&D efforts in this area across Europe, said in a statement: “Not following the ‘one system fits all services’ paradigm of current architectures, 5G NORMA will allow for adapting the mechanisms executed for a given service to the specific service requirements, resulting in a novel service and context–dependent adaptation of network functions.”
If any proof is needed that a single-minded network is no longer relevant to modern needs, we only have to look at the way M2M applications were sidelined in the LTE process, with participants wholly obsessed with the rise in broadband usage. That has forced the 3GPP community to revisit M2M in order to apply cellular to the internet of things (IoT), hastily reviving dormant specifications like Category-1 and throwing together new low power specs for Release 13.
There are 13 organizations involved in NORMA’s quest to avoid these pitfalls this time around, and to develop an end-to-end network architecture which can meet an extremely diverse and unpredictable range of requirements and use cases. The lead vendors are local giants Nokia and Alcatel-Lucent, plus NEC Europe, and the group includes three European operators – Deutsche Telekom, Orange and Vodafone. Other participants include UK consultancy Real Wireless, which will model the socioeconomic benefits of 5G for current, and potential future, users; as well as academic partners, the Kaiserslautern University of Technology in Germany, and the Carlos III University of Madrid. The project begins immediately and is due to be completed in 30 months’ time.
“5G NORMA will ensure economic sustainability of network operation and open opportunities for new players, while leveraging the efficiency of the architecture to do so in a cost and energy-efficient way,” said the group.
Dr Werner Mohr, chairman of the 5GPPP Association, echoed the thoughts of many early contributors to the 5G process – that it needs to look far beyond the air interface or the traditional remit of the 3GPP, and consider wider architectures, including virtualization, in the context of supporting a large number of business models, devices, usage patterns and service providers. Mohr said in a statement: “5G is not only about new radio access technology, network architecture will play an important role as well. 5G networks will have to be programmable, software driven and managed holistically to enable a diverse range of services in a profitable way.”
All this is far easier said than done, and undoubtedly some of the most advanced R&D on virtualization and SDN for the cellular world is being done in Asia and the US rather than Europe. But the EU and its homegrown wireless powerhouses – Nokia, perhaps, more than any – are determined to seize back the initiative in 5G and restore Europe to its old place at the heart of mobile standards.
Many of NORMA’s assumptions about the 5G network are, unsurprisingly, common to most similar initiatives, and heavily influenced work on operator requirements by bodies like the NGMN Alliance. So the resulting architecture will need to support higher bandwidth, lower latency, vast numbers of connected ‘things’, as well as self-optimization, HetNet and full duplex transmission, said the participants. But virtualization is at the heart of it all, or, as the official statements put it, “the innovative concept of adaptive (de)composition and allocation of network functions, which flexibly decomposes the network functions and places the resulting functions in the most appropriate location.”
NORMA may be resolutely flying the EU flag (apart from the involvement of NEC), but increasingly the major European vendors are earning their 5G chops by embedding themselves with Asian partners, particularly the operators in Korea and Japan which are set to deploy ‘pre-standard 5G’ at the earliest stage and so, as in previous generations, lay down some important markers for the rest of the industry. Ericsson, which is notably absent from NORMA, had its own 5G news in the past week, supporting a demonstration by SK Telecom of Korea, of an experimental small cell architecture with minimal interference.
The operator said the technology, piloted at the two companies’ joint 5G testbed at SK Telecom’s Bundang Building in Seoul, had overcome some long standing challenges of small cell deployment by using a new radio frame architecture with advanced interference control technology. This more than doubled user performance at the cell edge, compared to LTE (and this is SKT’s LTE, some of the fastest in the world).
Alex Jinsung Choi, the carrier’s CTO, said: “SK Telecom is delighted to announce the successful development of a 5G small cell system that tackles the root cause of inter-cell interference. We will continue to work closely with Ericsson to develop key 5G technologies so as to demonstrate a 5G trial service in 2018.”
Aside from this interesting display from Ericsson, the week in 5G has belonged to Nokia, especially when it comes to Asian cooperations. It shared the stage with both SK Telecom and Japan’s NTT Docomo in showing off advances that the operators hope will feed into 5G standards in time. Its latest breakthrough with SKT was to verify a ‘hybrid core network architecture’, which the operator claims will be a key enabling technology for 5G.
This builds on the industry’s current moves towards network virtualization and SDN, creating a technique to separate the user plane and control plane, a fundamental aspect of SDN. The two companies have co-developed a method to separate the two planes of an LTE core network physically, and say this can be applied to future 5G architectures (whatever they may be).
This will allow operators to build hybrid core networks which can combine centralized and distributed network models, allowing for the flexibility which many believe will be the hallmark of 5G, rather than a single rigidly defined network design. The SKT/Nokia technology restructures the core network around distributed user plane functions, but controlled by a central control plane, which handles services such as authentication, voice and internet connectivity. This is the optimal way to support and transmit massive amounts of data, said the partners.
SK Telecom said in a statement: “As end user traffic must go through not only the RAN but also the network before reaching the internet, radical renovations on core network architecture is one of the keys to the successful implementation and commercialization of 5G.”
The Korean MNO also announced the opening of a 5G R&D center, also a joint effort with Nokia, and based at the Finnish firm’s offices in Korea. The center will allow researchers from both firms to study and develop core 5G technologies, including “gigabit-level data transmission technology and cloud-based virtualized base stations”. A testbed will be opened there by the end of this year to verify and demonstrate 5G technologies.
“SK Telecom is spearheading efforts to develop 5G enabling technologies and architecture to flawlessly support high speed transmission of massive data, including immersive multimedia contents,” said Choi.
Meanwhile, Nokia hosted a session at the LTE World Summit in Amsterdam recently, at which another of its 5G R&D partners, NTT Docomo, gave updates on its efforts. The carrier took the almost inevitable step of coining the ‘5G+’ phrase, before anyone even has base-level 5G, to reflect its planned two-stage launch – in 2020, and then in 2022-23.
In fact, the initial deployments will be largely focused on traditional spectrum and evolutions of the LTE architecture, said Takehiro Nakamura, VP and director of RAN for the operator, and on traditional use cases, mainly high speed mobile broadband. Ultra-low latency networks and high frequency bands, both expected to be important for the internet of things (IoT), will come later.
“Our primary requirement is mobile broadband, so in 2020 deployment we will mainly focus on the higher data rate and higher capacity. But in the later stage we should focus on the other aspects, such as the low latency,” Nakamura said.
He sees high frequency bands as an important element of next generation networks, but perhaps not until the ‘5G+’ stage, after 2020. At the turn of the decade, research into harnessing these bands will still be immature, and there are unlikely to be major spectrum allocations until after the World Radiocommunication Conference of 2019, which will be charged with examining 5G spectrum.
Nakamura was realistic that, in order to launch commercial services in 2020, Docomo would have to work with early versions of standards, and with whatever spectrum might be available. The operator needs a stable specification by the end of 2018, for a 2020 roll-out, he said, which means the ITU-R IMT2020 specs, scheduled for late 2019, will be “too late for Docomo”. So, as in previous wireless generations, the Japanese firm will go ahead with pre-standard platforms, with the proviso that these will have a forward path to ‘5G+’ (or in other words, officially sanctioned platforms).
Among the technical advances which Docomo is making in its labs, and which it believes should be part of 5G, are massive MIMO; carrier aggregation between macro and small cells; phantom cells to create small cell overlays for Cloud-RANs; and new radio access technologies such as the carrier’s own NOMA (non-orthogonal multiple access).
After all this talk of early deadlines and 5G+, it was left to a UK carrier, EE, to bring the discussion in Amsterdam back down to earth. Mansoor Hansif, the director of radio access networks, spoke for many when he said: “There’s too much focus on the Gs. The real global race is between customer expectation and customer experience … We don’t necessarily want to be first with 5G, but we want to be aligned with industry thinking in the future.”