Industry Voices—Lowenstein’s View: The next year is critical for fixed wireless access

For the wireless historians among us, we’ve been talking about fixed wireless access (FWA) since AT&T’s "Project Angel" back in the 1990s. Up until now, it’s been mainly a niche product, used for "broadband as a last resort" in areas underserved by cable or DSL. In the U.S., it’s used mainly by WISPS in rural areas, while in some developing countries, wireless is a primary means of internet access in some locations.

However, FWA is now getting a serious look as a potentially viable mainstream broadband service in North America and other developed economies. It has two flavors:

1. Using existing LTE spectrum, mainly in rural areas, as sort of a "next gen" version of what the WISPs have been providing for some time.
2. In cities, it can be a competitor to broadband, as one of the initial use cases for pre-5G/5G service, using the millimeter wave (MMW) spectrum bands.

FWA has a large number of tests and trials underway and planned for the near future by leading operators such as Verizon and AT&T, as well as from exciting startups such as Starry and Mimosa.

It’s always been a bit amazing to me that there’s been so much focus on whether the wireless market is competitive, given the fact that some 50% of urban and suburban areas have only one ISP option for 25 Mbps or better internet, and that 39% of rural Americans lack access to 25/3 broadband (the FCC’s current requirement). So there’s definitely an addressable market out there.

The main question is, can FWA can be a viable solution for broadband?

Before tackling this, it’s important to define minimal performance requirements and a baseline use case. For urban areas, I’d say it’s minimum 25 Mbps download and 3 Mbps upload (50-100 Mbps more likely required), supporting capacity of ~100 GB per month per user or ~200 GB per household, for circa $50 per month. It’s gotta pass the Netflix test. In rural areas or other spots underserved by broadband, a minimum of 10 Mbps would be required, with a capacity of ~100 GB per household.

Some important market developments over the past year could also potentially move the needle on FWA viability. First, more spectrum is becoming available, with new bands, carrier aggregation techniques and new capacity becoming available from LAA, CBRS and so on. In the mmWave bands, the wide 200 MHz to 1 GHz channel swaths support, at least in theory, game-changing speed and capacity improvements. There is a gap in mid-band spectrum options—something the FCC is looking at—which could open up new opportunities for FWA in ex-urban type areas.

Second, significant technology advances are becoming available as well, including beamforming capabilities, smaller and more advanced antennas and improvements in the size and cost of the CPE (such as transceivers).

With all these stars aligning, there’s been a lot of action on FWA:

• AT&T is selling LTE-based fixed wireless internet in Georgia in rural and underserved areas using WCS spectrum, with plans to reach 400,000 locations in 17 states by the end of the year and 1.1 million locations by 2020. AT&T is also testing FWA on millimeter wave in two cities.
• Verizon is piloting FWA in 11 cities using the 28 GHz band, although the operator has said little about how extensive those service areas are and has not published any results to date.
• Starry, a well-funded startup founded by the Aereo team, has launched a beta test of its Starry internet service beta in Boston, using the 37 GHz band. The service offers broadband-esque capabilities of ~200 Mbps download and 5 ms latency, at $50/month for unlimited data.
• Webpass, now owned by Google, is focused on the MDU market in cities with its point-to-point service. It is expanding service in Denver, Seattle and select additional markets.
• Mimosa, an FWA MVNO, is focused mainly on the sub-6 GHz band, offering service mainly through WISPs such as Triad Wireless, and starting to do some interesting work with utility companies. Their focus is on rural and suburban areas.
• Several WISPs, such as Redzone Wireless in Maine and First Wireless in Ohio, are offering revamped FWA services using LTE Advanced or pre-5G spectrum. 

The two big questions for FWA are the technology and the business case

Much of the trial activity today is focused on the mmWave bands, which in theory could support the speed and capacity for a competitive FWA service. The major challenge in the higher bands is propagation, where there is far greater sensitivity to foliage, glass and building construction materials. Starry has been doing some pioneering work here, having built its Starry Beam network node and Starry Station transceiver in-house with new advances in beamforming technology on the MIMO antenna.

The business case is the other challenge. I am still not convinced that cellular offers a viable broadband alternative in cities and dense suburbs, particular where there is already good broadband infrastructure. The economics, in terms of cost to get to a home, plus the equipment, don’t really add up, given the speed expectations and capacity requirements.

The urban opportunity is more likely in three specifically defined segments:

1. In the MDU/MTU market, which is where most of the pilot and trial action is right now.
2. As a cut-rate broadband service, a la Metro PCS/Cricket in their day, circa $25-30/month for a 20-25 MB service and some capacity limits.
3. As a solution for the last ½ mile or so from a fiber node, since building fiber directly to homes has proven too expensive.

In rural areas, historically the most challenging from a broadband perspective, FWA is starting to become a more viable solution, and we are seeing somewhat of a resurgence in the WISP market.

One emerging segment of market potential is the ex-urban or less densely populated suburban market, which has been sort of a broadband vacuum—sometimes beyond the reach of cable and generally underserved by DSL. Here, we are seeing increased interest from Tier 2 wireless companies, telcos and even utilities, all of whom have assets that can be leveraged for FWA. Availability of more mid-band spectrum could be a catalyst for FWA in these areas.   

This is a peak time for FWA activity, with a new wave of technology and products, and many pilots and trials. There are still a lot of questions with regard to the technology, the business case and how FWA can be deployed at scale. We will know a lot more in a year. 

Mark Lowenstein, a leading industry analyst, consultant and commentator, is Managing Director of Mobile Ecosystem. Click here to subscribe to his free Lens on Wireless monthly newsletter, or follow him on Twitter at @marklowenstein.