Editor's Corner—The economics of fixed wireless, from LTE to 5G, and what it means for Verizon

verizon tower carolinas
Fixed wireless services typically rely on towers broadcasting to outdoor receivers.

Verizon has indicated it will launch fixed wireless services using its 5G network technology standard sometime next year, after tests in 11 markets this year. The company hasn’t provided many exact details about how it plans to go to market—such as planned pricing, speeds and coverage—but a review of the current fixed wireless landscape indicates Verizon will need to overcome a number of hurdles to successfully challenge its wired cable and telco rivals with a wireless service.

First, it’s worth noting that fixed wireless services have been around for years, and have been transmitted through a variety of technologies including WiMAX, 3G and proprietary protocols. Fixed wireless deployments usually rely on a central tower (connected to some kind of high-speed backhaul) that broadcasts a signal to outdoor receivers, usually those installed on the rooftop of a residential or business location. Ergo: Fixed.

As wireless technologies have improved (think 2G to 4G), so have the speeds available through fixed wireless services. Australia’s NBN (PDF) wholesale TD-LTE network (currently run by former Clearwire chief Bill Morrow) provides a great example of this wireless progression: The service initially launched in 2012 with 25 Mbps downloads, and by 2018 that’s expected to increase to 100 Mbps downloads thanks to NetComm Wireless’ modems, Qualcomm’s chips and kit from Ericsson. Current NBN monthly data allotments range from 6 GB for $27 per month to 2.6 TB for $66 per month.

Fixed wireless in rural areas

Closer to home, Rise Broadband offers an even clearer look at the economics of LTE-powered fixed wireless services. Specifically, the company’s cofounder and chief development officer, Jeff Kohler, said recently that fixed deployments typically cost a fifth to a tenth what it would cost to build a comparable wired service. Rise operates fixed wireless services in rural locations in Colorado, Idaho, Nevada, Oklahoma, Texas, Utah and elsewhere in the West. The company’s data allotments range from 250 GB per month to 500 GB per month.

Although Rise currently operates various nonstandards-based network systems, like those from vendors Cambium Network and Mimosa, the company today is mostly building out LTE network technology in 2.5 GHz and 3.65 GHz. Kohler explained that LTE equipment typically costs roughly twice as much as nonstandard equipment, but fixed LTE services can generally provide speeds up to 100 Mbps—twice the speed of most nonstandard equipment—as well as support for more customers. (While Kohler wouldn’t give a specific price for Rise’s LTE tower equipment, he said that “these things do cost tens of thousands of dollars.”)

Kohler cautioned that actual, real-world speeds are often determined by receivers’ distance from Rise towers. He said most Rise customers live 3 to 4 miles away from the company’s towers. As for customers’ receivers, often called customer premises equipment (CPE) Kohler said that an LTE receiver typically costs $100 to $200, or around half the cost of most nonstandard receivers.

As for the backhaul on its towers, Kohler said Rise typically pays pennies per gigabyte, or around 5-10% of its revenues (he noted though that those prices are often based on total volume, which means that smaller providers might pay more per gigabyte than Rise, which counts around 200,000 customers).

As a result of these economics, Kohler said today Rise usually deploys LTE in locations where it expects to service more than 100 customers per tower. He said that, depending on the weight restrictions on the tower it’s using, Rise can increase its capacity by adding additional access points on its towers. He said one Rise tower serves roughly 800 customers, mainly thanks to the installation of a handful of additional access points on the tower. Rise customers usually pay $50-$60 per month for service.

Rise and other, smaller providers aren’t the only ones playing in the fixed LTE game. AT&T recently said its “Fixed Wireless Internet” offering for rural and underserved areas is now available through 70,000 locations across nine states (using outdoor TD-LTE antennas from NetComm). The company said it plans to serve more than 400,000 locations with the service by the end of the year, growing to 1.1 million locations by 2020.

As CCS Insight noted, AT&T's fixed service provides download speeds of up to 10 Mbps and 160 GB of data per month at a price of $70 per month or $60 on a contract. Each additional gigabyte costs $10. Users must have an external fixed antenna installed outside their home or business; the installation fee is $99.

“The offering may be more expensive than traditional fixed-line services such as cable and DSL, which usually allow unmetered broadband access, but it's a bargain when compared with other LTE broadband services or satellite, which cost about $60 for 10 GB of data,” CCS added.

Fixed wireless in urban areas

But there’s a big difference between Verizon’s fixed 5G plans and Australia’s NBN, Rise Broadband and AT&T’s fixed wireless adventures: Verizon is targeting urban areas with fixed wireless services running in millimeter-wave spectrum, not rural locations with lower-band spectrum.

Rise’s Kohler noted that high-band fixed wireless services running in the 28 GHz, 39 GHz and even 60 GHz blocks are “great for high-capacity, high-speed services. [But] the stuff travels about two blocks. So that’s your problem,” he said. “So in a dense, urban situation, it will take a lot of small cells to make that work because you’ll have to have it travel over some distances.”

Rise isn’t alone in noting such concerns. “Despite millimeter wave bandwidth capacity, the propagation remains challenging,” noted the analysts at Wells Fargo in a recent report on 5G. “We heard estimates that mmWave could cover possibly 0.2 to 0.25 square miles (essentially several hundred feet), but that something as simple as a hand or any physical barrier could block signals for mobile devices.”

Indeed, high-band signals can be affected by rain, trees or vehicles parked between a transmitter and a receiver. For example, an AT&T executive told RCR Wireless News recently that the company’s 5G tests in Austin found that AT&T could bounce 28 GHz signals around obstructing foliage in order to reach a receiver almost a mile away. But the executive added that indoor receivers were stymied by reflective window coating, forcing the carrier to test an indoor system through more traditional dual-pane glass.

And signal reliability is just one of the challenges that Verizon will have to face in the market for urban fixed wireless services. Unlike rural areas, urban locations in the United States already likely have some kind of internet service, be it DSL or DOCSIS. In order to combat existing, incumbent wireline internet providers, Verizon’s fixed 5G service will have to provide faster speeds at a lower costs than the alternative.

In evaluating Verizon’s fixed 5G economics, it’s actually worth looking at a tiny startup that’s attempting essentially the same model in the same sorts of locations. Starry, founded by the serial entrepreneur Chet Kanojia, is currently testing fixed wireless services in Boston.

“The key to Starry’s technology, which uses beamforming to deliver symmetrical broadband speeds as high as 1 Gbps using millimeter wave spectrum over distances as far as 1.5 kilometers, is cost,” wrote the Wall Street analysts at MoffettNathanson in a detailed report on Starry issued in April. “With all-in costs for a base station of perhaps $25K, they have already driven their total cost to ‘pass’ a home to as low as $10 in a dense city like Boston. Their next target is lowering the costs of their CPE equipment, which is currently more than $1,000, to perhaps $200 per home.”

Continued MoffettNathanson: “With costs that low, they could offer competitive broadband speeds for a fraction of the current price of wired broadband.”

Importantly, the firm added that Starry is looking at deployments in areas with more than 1,000 homes per square mile in order to make its service economical, and that Starry hasn’t yet settled on a price for its service but discussed options ranging from $20 to $50 per month.

Tweaking the economics

Verizon is likely looking at similar economics, and is surely looking to drive the costs out of its own model through its massive purchasing power, its growing war chest of millimeter-wave spectrum and potentially its own fiber backhaul network, like the one its building in Boston.

Although the final economics of Verizon’s planned 5G fixed wireless service remain unclear, it’s telling that T-Mobile has specifically spoken out against using 5G for fixed scenarios. T-Mobile of course doesn’t have the millimeter-wave spectrum holdings or fiber assets that Verizon does, but it’s reasonable to assume T-Mobile would at least consider the business if it offered exceptional economics.

That said, Verizon’s existing wired internet business has set a relatively low bar in terms of selling internet services. As Leichtman Research Group reported, the nation’s cable operators—led by the likes of Comcast, Charter and Altice—continue to make significant gains on telco internet providers like Verizon, CenturyLink and AT&T. The firm noted that the nation’s cable operators collectively accounted for fully 90% of the total wired broadband internet net additions the industry recorded in the first quarter of 2017. For its part, Verizon lost 27,000 customers during the period.

It’s certainly possible that next year Verizon could offer a compelling bundle combining its nationwide cellphone service with a high-speed, fixed wireless internet offering for residential and business customers in select markets—a bundle that might combat cable’s gains in the broadband internet market. But Verizon is also the company behind the Go90 video app, which clearly hasn’t lived up to expectations. The jury is definitely still out on the economics of a fixed 5G service. – Mike | @mikeddano