While many operators are interested in the high capacity of the millimeter wave bands from 26 GHz upwards, the early trials are often focused on a fixed wireless access (FWA) model which looks unattractive in most markets. Verizon and AT&T may see FWA as a means to expand their fixed/mobile footprint, but for most operators, the level of investment required to come close to the performance of fibre would be prohibitive.
Speaking at the broadcasters’ annual jamboree, the IBC conference in Amsterdam, Balan Nair – CTO of cable giant Liberty Global – insisted 5G would not be a replacement for fixed broadband.
“The economics just aren’t there,” he said. “You’re talking about buying hundreds of towers and all of that spectrum. And on the residential end, putting a device outside the window and wiring it back into the home. It’s a question of business model and if you plan on making any money. The economics benefit fixed.”
One factor which may improve those economics is a new project within the Facebook-led Telecom Infra Project (TIP), called Millimeter Wave Networks Project Group. Like other TIP efforts, this aims to develop advanced telecoms technology which is open, commoditized and affordable. Large operators are increasingly active in an initiative which could disrupt their traditional supply chains and cost bases, and Deutsche Telekom is taking the lead role in the mmWave project.
Initially, the focus is on the 60 GHz band, which is already in use for the WiFi-based WiGig technology, and has the advantage of being unlicensed in most markets. Whatever the relationship between 3GPP 5G and evolutions of WiFi turn out to be – a fully multi-access platform or merely an offload scenario, like today’s – it is likely that WiGig will play a major role in 5G. Most of the pioneering work in using mmWave for broadband wireless, including getting the original expensive gallium arsenide chips into CMOS, has been done in 60 GHz so there is an ecosystem maturity there which is currently lacking in the likely ITU 5G bands such as 28 GHz, or the US-designated 26 GHz and 39 GHz – let alone the very high frequencies above 100 GHz, where companies like Samsung have been carrying out R&D.
Facebook has already developed equipment for 60 GHz with its Terragraph solution for dense urban hotzones. Along with its other RAN-focused developments, like the Aries antenna array, this is intended to be turned into a low cost white box offering within the TIP process, which now allows for open source and licensed approaches (the latter so as not to deter innovators from taking part).
Facebook will co-chair the new mmWave group with DT, and this will also focus heavily on metro area networks for smart cities, as well as other use cases such as residential access. The German operator is opening a TIP Community Lab, also heavily focused on mmWave, which will be located in Bonn.
The German firm will also open an Ecosystem Acceleration Center (EAC) – these facilities are to be a primary way for TIP to foster start-ups and extend its reach, and several carriers are already participating, including BT in the UK.
“Our goal is to deliver the best network services to our customers, and at the same time deliver a superior capital efficiency,” said Bruno Jacobfeuerborn, CTO of DT. “This requires more radical approaches. We value Facebook’s contributions to this space and enjoy the excellent partnership we have as part of TIP and beyond.”
The goal of the mmWave group is to develop very high performance but very cost-efficient access and metro networks, initially in 60GHz. While TIP may be highly disruptive to telco suppliers, it is, currently at least, very geared to conventional telcos and their use cases. Like Google, Facebook’s overriding aim is to extend affordable internet access to the whole planet, in order to drive usage of its own services and adverts. In future, that is likely to involve support for platforms that will enable a far wider variety of operators – such as Google’s idea of a national, wholesale, virtualized network which could be accessed, on an on-demand basis, by large numbers of service providers.
Virtualization, cloud networks, dynamic spectrum and network slicing will all make this vision achievable, but they will not converge and mature for years. In the meantime, the webscale giants know that telcos are the biggest controllers of broadband access, and they need to be wooed as partners, and given the tools to accelerate their own deployments to reach new users and support new use cases.
So the mmWave group’s list of target use cases is entirely telco-focused – gigabit wireless fiber services for residential users, mobile backhaul, and dense hotzones for smart cities.
“We are aiming to deliver cost-efficient solutions, truly innovating within the market,” said Andreas Gladisch, the group’s DT co-chair. “We expect our work to create a vibrant open, community-driven ecosystem, that will meet the needs of telecom and city operators.”
DT was a founder member of TIP in February 2016, along with Facebook itself plus SK Telecom, Intel and Nokia – the latter the most progressive of the large network OEMs in adapting its model to a virtualized, software-driven world. Nokia’s work on white box platforms and network slicing enables it to remain calm when it hears CTOs like Jacobfeuerborn talking about the death of the current RAN business.
“Telco is generally a market with high entry barriers and is dominated by established vendors,” said Jacobfeuerborn. “Small start-ups have problems entering the telco market, even for proof of concepts. As a result, the venture capital community has been very hesitant to make infrastructure-related investments. The TEAC centre in Berlin will help to break this deadlock. Our work on 5G has shown Deutsche Telekom’s leadership in driving telco innovation, and I am very confident that we will also be successful in helping push the broader telecom infrastructure ecosystem forward.”
The project will draw on the lessons of the Terragraph concept platform, which claimed to have overcome many limitations of signal range and absorption which are associated with 60 GHz. Now the new group will work on designing small nodes, suitable for deployment on light poles and other city infrastructure, which will combine radio transmitters and receivers and will work in mesh configurations to route traffic to and from a small cell, WiFi access point or building.
“This network architecture requires only a handful of nodes to be connected to fiber in order to provide city-wide coverage, which will help minimize capital expenditures,” said Gladisch and Salil Sawhney, Facebook’s co-chair, writing on the TIP blog. “In addition, the hardware will use a commercial off-the-shelf WiGig chipset components to contain costs.”
The group will also work on a set of software tools and best practice blueprints to simplify deployment for telcos and municipalities. Its non-hardware activities will focus in particular on a template for cost economics modelling; test and validation tools; network planning; and best practices for deployments in FWA, smart city or mobile backhaul environments.
Yael Maguire, director of connectivity programs at Facebook, told Mobile World Congress Americas: “For us, the unvalidated hypothesis is whether or not 60 GHz unlicensed spectrum can be used in the context of providing reliable broadband connectivity in dense urban environments. We contend if you’re willing to make your mesh the number of links relatively short, oxygen is actually your friend. That is our claim and we have to demonstrate this and prove this.” His idea is to optimize the interference so that the oxygen absorption becomes a benefit.