Road to 5G will finally make the neutral host model inevitable

Projects like the Flat Distributed Cloud (see above) are still somewhat futuristic, despite FDC being demonstrated in an LTE testbed. But others are working on more readily deployable mobile edge cloud platforms. One interesting start-up is Vapor IO. It recently announced its ‘data center as a platform’ offering, Project Volutus, which is powered by its Vapor Edge Computing technology.

Vapor IO’s proposition is to offer fully managed micro data centers at the base of cell towers and other cell sites, to allow MNOs, and cloud or webscale providers, to deliver edge cloud applications efficiently, and in a manner that can harness intelligence about the radio network and the location and context of users in a cell.

This is interesting because it points to the very different pattern of infrastructure ownership which may emerge along with the very dense, very distributed telco network. The balance of central and local, physical and virtual, on-premise and cloud will vary with each operator and will decide its requirement for assets like sites, servers and fiber. But it will have another big decision – which of those elements to own and which to use on an ‘as-a-service’ basis.

Leasing fiber backhaul and towers is nothing new of course, but for dense edge-oriented networks, there will be an explosion In the number of sites, wireline connections and microservers required. It will rarely be economically or logistically viable to have more than one of these dense networks per location, so the model which, currently, is seen mainly in specialized environments like stadium DAS is likely to extend into the mainstream. Neutral host networks of passive infrastructure and backhaul will spring up for small cells as they have in the macro network, where MNOs increasingly rely on towercos rather than owning their own sites.

Those towercos may be the companies which finally make the neutral host model work for small cells, and hence open floodgates to real densification. It is notable that Crown Castle, the most small cell-friendly of the major US tower operators, has taken a minority stake in Vapor IO, and is no doubt weighing up the potential to add value to site and fiber deals by including cloud computing resource as well.

Crown Castle has made a string of acquisitions recently to strengthen its ability to offer a full small cell platform including fiber backhaul or fronthaul. This will be important to enable towercos to extend their macrosite model, rather than losing out to another logical category of neutral host providers, fiber or cable network owners.

The most recent of these deals was announced last week, with Crown Castle saying it would pay $1.7bn in cash for fiber owner Lightower. The towerco said this would make it the largest provider of shared wireless infrastructure in the US, with about 40,000 towers, 50,000 small cell nodes on air or under development, and 60,000 route miles of fiber. Earlier this year, the company paid $600m for another fiber owner, Wilcon, and that followed a $1.5bn deal for FiberNet. Together, those two added about 11,600 route miles of fiber in Florida, Texas and southern California. Crown Castle’s fiber expansion started in 2015 with the $1bn takeover of Sunesys, part of a gold rush for fiber assets which has been going on for the past two years and has also seen Verizon and AT&T snapping up providers.

Crown Castle said its latest purchase “significantly increases opportunities for small cell network deployments in top metro markets in the Northeast including Boston, New York and Philadelphia.” The dowry includes 32,000 route miles of fiber, which will double its fiber footprint. The company said it plans to combine those fiber assets with its small cell platform to meet new demands from mobile operators. It expects the acquisition to add between $850m and $870m to its site rental revenues in the first full year after completion.

Just after announcing the deal, Crown Castle reported its second quarter figures, which saw revenues rise 8% year-on-year to $869m, with a 30% rise in net income, to $112m.

“Among the three publicly traded tower companies, Crown Castle has uniquely chosen to focus on small cell development to supplement its macro site tower business,” wrote Guggenheim analyst Robert Gutnam in a client note recently. “Currently, this segment comprises 16% of total revenue, and generates 6% ROI—intended to ultimately reach 10% as the company adds second tenants (colocation) to small cell installations.”

Gutnam highlighted the critical success factor for the neutral host model – as in macro towers, it is important to have multiple customers per site. “While 85% of this capital is estimated to be spent on new systems, 15% of the total is expected to be allocated to colocation installations,” Gutnam wrote. “We believe colocation is crucial to the success of the strategy as incremental tenancy is necessary to generate double-digit ROI.”

Verizon and AT&T are investing heavily in their own fiber and site assets to support small cells and Cloud-RAN, but they have also acknowledged that they will use third party services including dark fiber as they expand the footprint of the new architectures. And the smaller operators are likely to keep capex costs down by relying heavily on outsourcing from day one. T-Mobile USA plans to make heavy use of dark fiber, Karri Kuoppamaki – VP of radio network technology and strategy – told a recent Wells Fargo 5G Forum. He said TMO works with a variety of partners which provide the fiber, real estate and labor while the MNO still procures the active equipment and also takes the lead on negotiating site deals with municipalities. That has resulted in a cost structure for densification which is at a “manageable level”, he said.

He added: “Ultimately small cell deployments, or successful small cell deployments, depend on the cost structure, especially the backhaul piece. If you can do that by partnering up with the right people, and bring that cost down a fraction of the cost of a macro then it makes sense.” He says TMO currently has 15,000 small cells including 13,000 DAS nodes and will add several thousand more this year, and a further 25,000 over the next few years, mainly to add targeted capacity in areas of dense population or usage.

Other infrastructure providers are starting to believe that the long-promised small cell ramp-up will finally happen, to their benefit, ahead of 5G. ExteNet, for instance, predicts that the number of active small cell sites in the US will grow from between 80,000 and 100,000 today to roughly 500,000 by 2021. And according to Wells Fargo, another fiber company, Zayo, estimates that each national MNO will need 150,000 to 400,000 small cell sites each over time, which will mean new form factors such as antennas on every street corner, fronthauled to the local or central core using fiber.

Low energy will put new pressures on 5G infrastructure:

As well as capacity, coverage and low latency, the 5G network will need to deliver significantly lower levels of energy consumption, and of overall resource to deploy a new cell. The Chinese vendors have been leading the way in many of these areas – both Huawei and ZTE have shown some imaginative approaches to lowering the cost and power levels of macro and small cells. At the recent Mobile World Congress Shanghai show, Huawei’s Network Energy unit presented its smart site solutions, which claims to improve cost of ownership and network energy efficiency of site. The company claims its products and services will increase carriers’ management efficiency by 150%, their equipment energy efficiency from 90% to 98%, site energy efficiency from 60% to 95% and network efficiency from 55% to 75%.

Fang Liangzhou, VP of Huawei Network Energy, explained that Huawei had combined network and electronic technology and incorporated distributed power systems, intelligent power distribution, intelligent energy storage and intelligent temperature control. “These solutions will make carrier’s site facilities and components visible, manageable and controllable, and then continuously improve their network quality,” he said.

Meanwhile, China Mobile and ZTE have signed a collaboration focused on 5G network energy efficiency. The vendor says it has been developing energy efficiency solutions for its pre5G line, which includes small cells and Massive MIMO antenna arrays, and is introducing them into its Access Cloud Engine (ACE). This supports mobile edge computing and flexible service deployment as well as multi-site collaboration and multi-system integration, and wireless resources can be enabled and disabled depending on the services in use, reducing energy consumption.

ZTE has also proposed a range of energy efficiency techniques for Massive MIMO base stations. The company’s planning director for TDD products, Zhang Shizhuang, said: “ZTE is willing to work with the partners in the industry chain to build more energy-efficient networks from the six perspectives of architecture evolution, networking, energy-saving technologies, application scenarios, new base stations, and chip evolution, laying a solid foundation for large scale commercial use of 5G networks.”