Will closer links to IETF save 3GPP from the challenge of open source?

As the 5G standardization process moves on, the question is clear – is 3GPP really fit for purpose? The vision for 5G – and its only real commercial justification – is that it will not just be a fifth generation radio, but a complete architecture. Yet most of 3GPP’s work remains focused on the radio, and it has been slow to include organizations addressing standards in other parts of the network.

Yet 5G will need open standards at each layer of the network, and between physical and virtual, wireless and wireline technologies. For many operators, 5G will go hand-in-hand with virtualization of the network, and the introduction of SDN controllers to achieve full resource flexibility, plus orchestrators to tie all the elements together as a holistic entity.

The 3GPP’s work is addressing some of these areas, and certainly the 5G New Radio work takes account of emerging trends like network slicing. But despite the acceleration of the first wave of 5G NR, the pace is glacial compared to that of the open source community, whether the Linux Foundation-hosted initiatives like ONAP (Open Network Automation Platform) or vendor-driven activities like Facebook’s Telecom Infrastructure Project (TIP).

The major Internet standards body, IETF (Internet Engineering Taskforce), was represented at last week’s 3GPP meeting, which is a welcome sign, but rather late in the day, while overtures from the IEEE – the main body governing networks in unlicensed spectrum – have reportedly come to nothing.

The companies driving the 3GPP remain largely the same large operators and vendors which dominated the business in 2G and 3G days. In 5G, other organizations will have powerful positions as the carrier network morphs into a distributed cloud/IT platform and as unlicensed spectrum moves to the heart of deployment strategy. Too often, the new vendors – start-ups and incomers from the IT world – find 3GPP an alien and slow moving environment, and turn to more familiar standards bodies and open source agencies, or to initiatives like TIP or GE’s Industrial Internet Consortium, where they feel they will have more power.

As Wireless Watch commented on last week, initiatives like TIP are also starting to attract traditional mobile operators, which do not want to be outflanked by new challengers armed with these more rapidly evolving platforms. Orange is supporting start-ups such as Amarisoft, in association with TIP, with a specific focus on alternative and lower cost ways to build a RAN; AT&T and even the mighty China Mobile have contributed their management and orchestration efforts to the open ONAP effort. These are not all pure open source – TIP is increasingly working with companies which want to license their technology, not give it away – but they do provide an alternative ethos and pace of change, to 3GPP.

This is why the start-up Phluido has caused such a stir, with its ‘radio as a service’ approach to ease the path to virtualizing the RAN. There are plenty of innovative young companies with new approaches to radio technology or network deployment. Some get acquired by the big names, some fade from view, a few make a real mark on the landscape – but only with significant backing. So although Phluido gained some attention when it came out of stealth mode with its ‘radio as a service’ invention three years ago, it then looked as though it would become one of the long list of interesting but commercially non-viable companies in this difficult space.

But now it is back in the limelight, but not because it has managed to penetrate the hostile waters of the 3GPP community. Instead, it is driving a new workgroup within TIP, focused on virtualized RAN (vRAN). Surely this is exactly the sort of effort which should be run under the auspices of 3GPP, ETSI (which has NFV of course) or other established bodies? Phluido’s activity within TIP is the strongest sign yet that the Facebook project will provide an alternative structure to the traditional standards bodies.

And if it fails – because of poor processes, or mistrust of Facebook itself – there will be other open source bodies to take its place, and all the while operators, old and new, will be becoming more accustomed to working in this way, and may even feel liberated from the ways of the 3GPP and its key vendors.

At last week’s 5G World conference in London, several operators were calling for a new approach from their entrenched vendors. Deutsche Telekom’s VP of innovation and research, Uwe Janssen, said in his keynote presentation: “We need something really radical to happen. On the infrastructure side there is limited innovation in infrastructure and it is also very difficult for newcomers to enter. This is despite a lot of good ideas and innovations coming from universities and start-up companies, but it is hard for them to come to market because there are a few big suppliers with their own agendas. We need to see a change here.”

With DT being a supporter of TIP, it was no surprise that Janssen pointed to the project as one way to encourage that innovation and shake up the established supply chain.

So, like the disruptive RAN start-ups Orange is supporting (see Wireless Watch July 12 2017), Phluido will doubtless find TIP a valuable showcase to attract the attention of operators which want to do things differently next time, and do not believe 3GPP work alone can help them to slash their costs and transform their service agility, the critical aims of 5G.

The company emerged in 2014 with a new answer to the main challenge of virtualized RAN – the amount of high quality fiber required for the fronthaul links to connect the physical cells with the virtualized basebands in the data centre or cloud. Phluido claims its approach could make it as viable to virtualize the RAN in the cloud as the packet core. Founded by former Qualcomm engineers Dario Fertonani and Alan Barbieri (now CEO), it has developed a technique that allows the network to pre-process and compress the frequency data at the cell site, so that it can be realistically transmitted to the data center via a lower performance connection such as an Ethernet cable or microwave. Once there, the baseband processing would take place, running as an application on centralized servers.

Since it first mooted this idea, similar approaches have been developed, in different ways. Companies like Altiostar have advanced the viability of Ethernet cable for fronthaul and that is now the topic of a 3GPP standards effort. There are many applications where latency can be improved by processing much of the data and signalling at the edge of the network, in effect distributing the cloud and the baseband to many edge nodes – as in ETSI’s Multi-Access Edge Computing (MEC) or Cisco’s OpenFog.

But Phluido has its own particular solution to the fronthaul issue, and it may prove to be an alternative to the routes favored by ETSI and 3GPP, to bolster TIP and its supporters, and the push for an open platform. Its unique offering is the ‘as a service’ element, which would fit in with moves towards slicing, supporting large numbers of service providers from a pool of baseband capacity. This could be run on a wholesale basis or by a group of participating operators, which would build out slimline radio/antenna units at the macro or small cell sites and then tap into the baseband in the cloud.

Barbieri told LightReading that Phluido has had some success in licensing its alternative fronthaul technology and its vRAN software stack to vendors, but he believes that forming the TIP working group will be a turning point. It should be a way to attract further innovation, and to raise Phluido’s profile, more rapidly than a start-up could do alone. The firm is already on the radar of two large operators, the UK’s BT and Vodafone, which are the co-chairs of the new group, while Telecom Italia is also part of it.

The group aims to devise a set of open specifications to support standards-based vRAN deployments in 4G and 5G. “We want a lower PHY [physical layer] split of the RAN, which is something the 3GPP is not currently taking care of,” says Barbieri. “With the power of the companies backing up the group, the hope is this will become the de facto standard.”

The vRAN workgroup joins other TIP activities in key areas of the telecoms network, from the Voyager optical transponder to the OpenCellular open source small cell to the Aries Massive MIMO array. Many of the technologies are emerging from Facebook’s own R&D and then being placed into the open source process for others to adopt or enhance. Others are coming from third parties, which has led TIP to place greater emphasis on its less open licensing options – for companies fearful of the patent protection minefield of open source, it is offering RAND (reasonable and non-discriminatory) licensing terms too, despite criticisms that this is taking it down the rocky road to being an old-style standards body.

That, in turn, highlights the difficulty of achieving a good balance between open and controlled. Most open source platforms, once they have achieved commercial success, tend to become more dominated by a few vendors, more controlling over licensing and usage – Google’s Android is a good example, as are many technologies in the enterprise world.

In the RAN world, however, while Facebook or Canonical or Google may be able to afford to meet the high costs of developing radio network equipment and then giving it away, start-ups cannot. There will need to be a fine balance between opening interfaces to all, while allowing inventors to monetize their IPR. So the vRAN fronthaul group is looking entirely at a RAND licensing model, rather than full open source like OpenCellular.

“It is not about open source, but it is about unbundling and open APIs and the ability for small companies to be a part of the ecosystem,” Barbieri said in the interview. “The secret sauce remains 100% the property of Phluido.”

This echoes the views of Amarisoft, one of the start-ups being funded under Orange’s Telecom Track project, which also provides exposure to TIP. This firm is also focused on vRAN – its software virtualizes the base stations and network backbone on standard PC-class hardware. It could also run on open, low cost appliances designed more specifically for the challenges of the RAN, like OpenCellular itself. CEO Franck Spinelli said: “We don’t want to let our technology go for nothing. We want money for that.”

But despite the likely trade-offs, this type of organization does breed a culture of innovation and inclusiveness which is very important, especially at the dawn of a new platform such as 5G.

The old and the new need to work together – there are many risks to the open source approach when it comes to carrier-grade infrastructure, and most open initiatives become the domain of a clique of powerful vendors just as standards bodies do. So a combined approach would be ideal, but that will require the 3GPP and its core members to make some significant moves to embrace organizations with different cultures.

Perhaps the recent talks with the IETF are a sign of willingness to do this. Though a very traditional standards organization, the IETF comes from the world of the Internet – though this has been entwined with the mobile network from 3G onwards, there has often seemed to be a chasm between their standards efforts. Internet access and applications are a primary use of mobile connectivity; a rising number of cellular devices are IP-linked. Yet the common work between these two standards bodies has not been as rich and frequent as was often predicted when the cellular network started to become an IP one. There has been joint work in areas like IMS and SIP, but it has been, in the view of many stakeholders, too ad hoc and sporadic.

Alissa Cooper, chair of the IETF and a Cisco Fellow, was a guest at the 3GPP TSG plenary meeting in West Palm Beach last week. Georg Mayer, chairman of the CT (core network and terminals) group, claimed that 3GPP-IETF cooperation has already made a significant impact – something many mobile Internet players would dispute.

Often, the barriers in the way of closer cooperation are to do with process and culture rather than technology. For instance, as TelecomTV points out, the IETF creates standards as RFCs (requests for comments), a different process from that of the 3GPP. The responses to the RFC go through several phases of development and review by the community, before being published as standards.

But Cooper made it clear, in her address to the 3GPP delegates, that such differences had to be overcome as the worlds of many standards bodies collide in 5G. She said 5G will need to take account of the work of many standards bodies and open source projects.

She identified several work items within the IETF which could be relevant to the 3GPP work on 5G, both in use cases and technologies. She also proposed that two existing technologies – EAP (Extensible Authentication Protocol), which support network access authentication and is used in WiFi standards, among others; and HTTP/2, an update to the HTTP standard which could help reduce latency and make more efficient use of network resources, in 5G and IoT applications.

As for new projects within IETF, she suggested that a range of items, mainly router-related, could be of interest to support 3GPP work. These included:

Data models as abstractions for consuming topology or proximity
Path computation to meet constraints of 5G radio
Routing protocols for distributed networking
Encapsulations to provide abstractions and metadata (being developed by the MPLS, Network Virtualization Overlays, and Deterministic Networking working groups)
Segment routing as means to provide transport slices
Service chaining to transport traffic across virtualized functions

Some of these developments could be important in 3GPP’s second phase of 5G standards, Release 16, which is expected to intensify the focus on low latency IoT applications as well as additional spectrum options and air interfaces. For instance, Cooper’s references to deterministic networking (DetNet) will strike a chord, for instance. This is a way to reduce latency, loss and jitter and increase reliability, by distributing packets over time and space for redundancy, and reserving data plane resources on path for individual flows.

IETF is also working on a new transport protocol called QUIC – a UDP-based, stream-multiplexing, always-encrypted transport protocol focused on minimizing application latency. It is unclear how that might fit with ETSI’s Next Generation Protocol (NGP) work, which looks to drive a replacement for TCP/IP and alter the whole foundation of the Internet to support a new generation of services and behaviors.

Cooper was certainly a breath of fresh air for many attendees, calling for the kind of broad-based cooperation which the 3GPP has been slow to pursue in the past. She said the IETF’s existing alliances with other organizations – such as the IEEE, Broadband Forum, Open Platform for NFV and others – will help broaden collaboration, and achieve richer standards more quickly.

That is the key – to accelerate the process of delivering open, unified specifications without sacrificing their functionality and richness (just sacrificing some of the bureaucracy and politics). If 3GPP does not succeed in rising to that challenge – which will mean cooperation with other bodies and with open source projects – it will find itself supplanted by Facebook TIP and its like.

Apple seeks radio engineer to work with 3GPP:

Some of the change to the 3GPP is likely to come from within. Apple is one of the companies which wants to have a more prominent role in setting mobile standards, but it will certainly not favor an open source route, given its tight control over its own technologies and IPR. So it is no surprise that we see Apple becoming more involved in pushing 5G ahead. It is planning trials of millimeter wave radios, and now it is advertising for an engineer to represent it in 3GPP’s RAN Working Group 1.

The job advertisement is for a senior wireless standardization engineer to work with RAN WG1, who is “fluent in cellular communication standards such as 5G NR, LTE and UMTS/CDMA”.

The advert, spotted by LightReading, describes the role as being “responsible for representing Apple Inc. in wireless communication standard bodies, and drive Apple’s strategic interest in those organizations. With the exciting opportunity to create innovations that benefit Apple’s product lines and enrich company’s IP portfolio, he/she will be working with a fast-paced team in driving Apple’s global standardization activities in areas covering radio interfaces, protocols and signaling, system architectures, security, and end-to-end services.”