The leading specialist vendors which have gained new prominence from the interest in O-RAN have been announcing a string of deals to support open specifications, and to combine their respective offerings. Some of these, like an agreement by base station vendor Airspan to support Altiostar RAN VNFs (virtual network functions), or a similar agreement between Altiostar and NEC, represent commercialization of work done for the Rakuten super-project.
Others are arising from other trials and deployments, or just from the logic of joining forces to prove that multivendor vRANs are realistic, and to gain a bit of scale against the big vendors.
The latest pact is between Mavenir and Altiostar, which plan to cooperate to specify open RAN radio units which could work with either of their vRANs, or those of others. This would result in an open reference design, which could be developed by third party OEMs.
Such designs reduce cost and time to market for the wider ecosystem and can kickstart adoption of new categories of equipment, as seen most dramatically in WiFi in the mid-2000s. The Facebook-initiated Telecom Infra Project (TIP) has seeded some reference designs for radio units (RUs) and other vRAN elements, and TIP may well be a home for the Mavenir/Altiostar design, since both are active members.
It is not clear how the new initiative will fit in with TIP’s own Evenstar project, announced earlier this year. This aims to establish open specifications and reference designs for a range of vRAN units, starting with the RU. This project is led by Deutsche Telekom and Vodafone, together with Facebook’s own Connectivity division, plus Mavenir itself; another vRAN provider, Parallel Wireless; and two radio specialists, MTI Mobile and AceAxis.
Vodafone’s Santiago Tenorio, the new chair of TIP, said the RU would be a “killer radio” because it is so important to operators. It should be available in mid-2020, initially targeting the 1.8 GHz band and a target price point of $1,000 per unit.
Seeding a broad ecosystem for the physical elements of a vRAN, such as the baseband hardware and the RUs for relatively small or simple base stations, could help to stimulate a segment where there will be plenty of new demand – in private networks, smart cities, in-building systems and so on – but where the giant vendors are not highly active. This will be easier for a new platform with a WiFi-style supplier base to address, rather than taking on the majors in the macro RAN, where Ericsson and Nokia are incumbent, and where development and integration challenges are beyond the capabilities of a low cost ecosystem.
There is certainly a need for a wider choice of RUs to work with the vRAN baseband software from companies like Mavenir and Altiostar. The highest performing RUs in the market come from the major vendors, but those suppliers are rarely willing to open their interfaces so their radios can be used with third party basebands. Nokia did it for the Rakuten Mobile deal, of course, but has not made that open interface a standard offering. NEC did the same, and should be in a strong position to improve its tiny RAN market position by offering its advanced 5G RU, co-developed with Rakuten, to work with O-RAN base stations.
And of course, there is the need to foster non-Chinese vendors since it is more than possible that the USA – and some other governments – will increasingly clamp down on any Chinese equipment in 5G networks. The hardening line was demonstrated when the US administration recently said that any chip manufacturer using any US technology in its fabs would be unable to supply Huawei without a special licence.
By placing the world’s largest semiconductor foundry, Taiwan’s TSMC, in a difficult dilemma, this ruling also highlighted the awkward position of Taiwan’s hi-tech industry – usually the cornerstone of an open, low cost technology base. Though politically aligned to the USA, it is not recognized as a separate state from China by the USA or many other countries, and its large manufacturers often have factories on the mainland, and large numbers of Chinese customers.
To get around such geopolitical complexities, the US-based open RAN contingent dream of a supply chain free of any elements from Greater China – some politicians talk of a US-only platform, but operators and most vendors, of course, realize it would be counter-productive to exclude Huawei and Ericsson, or indeed suppliers from more US-friendly parts of Asia, such as Samsung and NEC.
Steve Papa, CEO of Parallel Wireless, recently pointed out one of the flies in that ointment – many 5G deployments will rely on Massive MIMO antenna arrays to extend range and capacity in midband and millimeter wave spectrum. But Huawei, in Papa’s view, has taken a clear lead in Massive MIMO active antennas, because it has invested in the advanced silicon technology (gallium nitride) to support them optimally.
He is calling for increased US investment in such foundational silicon technologies so that, in future, operators will not have to opt for second-rate technology if they exclude Chinese vendors. But he is also highlighting a major challenge for O-RAN – open RUs and low cost antennas are very important for the smaller networks, but the emerging ecosystem lacks the technology to match the performance of the Massive MIMO products developed by major suppliers over many years.
Parallel Wireless is working with its own radio partner, GigaTera, formerly known as Korea Micro Wave, but now adopting a less Asian name and market focus, presumably to tap into US 5G business. It designs radios for the north American market from a base in Los Angeles, and has been a Parallel partner for over five years.
Papa wrote in recent comments to the press: “The US rural operators are looking for 4G-like radios versus the 5G active antennas that make 5G so compelling. These 5G active antennas (Massive MIMO) are where Huawei has their lead and developing this line of radios is not going to change that lead. … The biggest challenge for operators is future-proofing their investments. They need to know that if they deploy O-RAN, that beyond just solving the rural problem of today that they can also ultimately solve the high end 5G Massive MIMO problem. In order for operators to make big O-RAN investments they need confidence that ORAN can solve their entire problem and it is critical O-RAN has an answer for 5G Massive MIMO.”
Even the most gung-ho US politicians calling for a brand new, America-first O-RAN platform are realizing that this will not be ready for the biggest network deployments overnight, and some interim solutions are needed. Attorney General William Barr came up with the odd idea of the US government taking a financial stake in Nokia or Ericsson. Earlier this month, Secretary of State Mike Pompeo called for what he called an Integrated Open Network (ION), which would include a mixture of traditional integrated networks and new open, modular architectures.
This reportedly came during a roundtable featuring the CTOs of Dell, Ericsson, Google, Microsoft, Nokia and Samsung last month. Mung Chiang, director of the State Department’s Office of the Science and Technology Adviser, said 5G ION would be “a new, practical kind of ‘open’ that can start today … With a broad, inclusive tent of what ‘open’ means, a nuanced appreciation of network deployment reality, and a more solid view on architectural choices, ION becomes one of the areas where the United States and partner countries can lead in 5G innovation.”
Details of the proposal remain unclear but the terminology echoed the call by Nokia CTO Marcus Weldon for an approach that was “integrated but open” – in his view, like that of Ericsson (see separate item), a fully multivendor RAN would risk sub-optimal performance in a primary network but large vendors could still support open interfaces and integrate at a deep level with third party elements (as Nokia has done with Altiostar and others in the Rakuten 4G network).