Disputes threaten US hopes that shared spectrum will close rural divide

In the past couple of years, the US has set the pace in seeking new sources of wireless broadband spectrum, and proposing creative ways to harness and share them. None of this comes without controversy, however, as seen last week in two areas where the US has pioneered flexible spectrum sharing – the 3.5 GHz CBRS (Citizens’ Broadband Radio Service) band, and the TV white spaces (TVWS).

The latter has not lived up to the high expectations which supporters like Google and Microsoft raised, when the FCC became the first regulator in the world, in 2008, to open up the fragmented spaces in the TV bands for open wireless use. The amount of spectrum available varied greatly between different locations, making it hard to plan services, and other countries were slow to follow suit, limiting global scale or roaming (though some, like the UK, have allowed wireless access and conducted trials to support use cases from rural access to the Internet of Things).

However, some of the work done to make TVWS usable has been very valuable in setting precedents for other, higher capacity bands like CBRS. In particular, Google and others devised geolocation database systems, to ensure that devices could seek out vacant channels before transmitting, and so avoid interference with the TV broadcasts. This work has fed into the geolocation databases for CBRS – Google and Federated Wireless are running the two major Spectrum Access Systems (SAS) for that band’s general access tier, which allows open access provided there is no interference with incumbent federal users, or priority access users (license holders).

Now the TVWS debate is back, following a proposal from Congress members of both parties, that the FCC should allocate three 6 MHz white spaces channels, in the 600 MHz band, for unlicensed use in every US market. This issue has become urgent following the incentive auction of 600 MHz broadcast spectrum to mobile operators, since supporters of dynamic and open spectrum want to have their own frequencies within that band too.

The main consideration is to improve rural broadband – 24m rural dwellers are still unserved, according to the FCC’s figures, out of 34m Americans who lack basic broadband access. Low cost wireless options are seen as a way to address that, though of course, the nature of unlicensed spectrum is that providers can innovate with a wide range of services and devices, and some players have argued that the primary value of TVWS will lie in the IoT, to support low power wide area (LPWA) connectivity.

However, TVWS have been somewhat superseded in this sector by other LPWANs in unlicensed sub-1 GHz spectrum, such as Sigfox, HaLow and LoRa, or in licensed bands, like NB-IoT.  That was made clear when an early technology for TVWS, Neul’s Weightless IoT protocol, was redeveloped to support the sub-1 GHz ISM spectrum (Neul is now part of Huawei).
Now the role of TVWS is seen as being mostly for rural consumer broadband, including streaming of video.

Both for rural access and LPWA, the TVWS have the advantage offered by any sub-1 GHz frequencies, enabling large cells which can cover big distances with sparse populations, at low cost (it can support ranges of 750 yards up to nine miles, depending on technology and power levels).

The FCC has already ruled that the duplex gap between 652-663 MHz and Channel 37 (a purposefully unused TV channel, which in the US is in 608-614 MHz) would be available on an unlicensed basis, rather than being included in the auction to MNOs. The regulator also has a draft NPRM (notice of proposed rulemaking) that would reserve an additional 6 MHz channel in each broadcast market for unlicensed use, at 54-608 MHz.

The latter band is the subject of the proposal by a group of 43 Congressional representatives, to reserve at least three TVWS in the 600 MHz band for rural broadband. This is based on an idea originally mooted by Microsoft in July this year, when it published a report commissioned from Boston Consulting Group (BCG), which outlined various options for bringing broadband to rural America, with their costs.

This concluded that, where population density is below two people per square mile, satellite remains the most cost-effective option; but when there are between two and 200 people per square mile, white spaces are the answer. Above 200 people, more conventional wireless broadband (WiFi or cellular) is best. It argued that the entire US rural population in that 2-200 population density sweet spot could be reached for just over $10bn, within five years, and at a speed of 25Mbps or more. The BCG paper estimates that this would cost $65bn using fiber to the home, or $45bn using satellite only, and between $25bn and 40bn using 4G or 5G fixed wireless.

Now the Congressional group has taken up the cause, saying in its letter to the FCC: “We believe that the television white spaces (TVWS) have strong potential to revolutionize broadband internet accessibility in rural areas.”

Support has grown among the hi-tech industry and consumer groups, with many of them petitioning the FCC ahead of a vote on the issue. Linda Moore, CEO of TechNet, a network of technology senior executives, wrote: “We urge the FCC to address the outstanding issues as soon as possible and ask that you take the final regulatory steps that are needed to provide regulatory certainty and allow operators to fulfill that potential of TV white spaces technology.”

However, there is opposition too. Broadcasters are concerned about interference with the transmissions which they still have in 600 MHz spectrum. Since the first TVWS consultations a decade ago, the National Association of Broadcasters (NAB) has consistently argued that the FCC’s measures to protect TV broadcasts and wireless microphones – including geolocation databases and mandatory approval of any white spaces devices – are not enough to ensure clear channels for TV stations.

“Using even the most wildly optimistic TVWS database numbers, TVWS advocates just need to connect 33,999,132 more devices to bring broadband Internet to 34m Americans without access,” said NAB EVP of communications Dennis Wharton, in a statement last week. “Despite sitting on the sidelines for years during the TVWS experiment, Microsoft now demands that the FCC oust television broadcasters and their viewers to pave the way for free spectrum for TVWS advocates. This would jeopardize local broadcast news, programming and lifeline emergency information forms of Americans. The FCC and members of Congress should not be fooled by Microsoft’s empty promises.”

And the National Translator Association used Hurricane Harvey disaster as an argument for keeping the white spaces clear for broadcasters, saying in a statement to the media:
“The response of free over-the-air television stations and their translators to the Harvey hurricane disaster with on-the-scene reporting, emergency information, and fund raising for disaster relief are ample indicators free-over-the-air television is America’s first choice for information. This should not be replaced with Microsoft’s ‘white space’.”

The NTA represents local governments, TV districts and not-for-profit organizations relaying free TV programming to rural areas beyond the major network coverage areas.

It went on: “Microsoft describes the spaces between full power TV stations as so-called vacant channels. NTA does not know anyone else who calls these channels ‘vacant’. Currently they are occupied by 3,776 licensed UHF and VHF translator stations, and 1,968 licensed UHF and VHF Low Power TV stations (Broadcast Station Totals as of June 30, 2017.”

Microsoft will not be giving up on its latest pet project to improve access to the Internet – and its services. In July, it created a Rural Airband Initiative and said it would invest in partnerships with telecoms companies to create 12 projects in 12 states over the next 12 months, to connect 2m people.

Such initiatives from large companies are important to build confidence, especially among the chipmakers which are essential to build an economic, large-scale device ecosystem. If Microsoft takes the lead and gets others investing – operators, ISPs and local authorities – this will make it worthwhile for WiFi chipmakers to adopt the underused 802.11af protocol (an implementation of WiFi which supports low power, long range links in the white spaces). This is one of a handful of TVWS protocols devised by the IEEE standards body, which under the right conditions, using up to 256 QAM modulation, OFDM signaling, and multiple MIMO streams, can deliver as much as 400Mbps per cell over multiple miles.

The PHY layer in 802.11af is based on the same scheme used in 802.11ac, the most recent WiFi version, and 802.11ah (brand-named HaLow), which supports LPWANs in the 900 MHz/868 MHz ISM bands. Because the propagation path loss is far smaller in the sub-1 GHz UHF and VHF bands, and because the FCC has allowed them to be operated at far higher EIRP levels, the range can be as long as 10 miles. If there is more than one channel not being used for broadcast, this allows up to four channels to be bonded. Then 400Mbps can be reached by adding 4×4 MIMO and Multiuser MIMO – which has already been launched in 802.11ac Wave 2 – plus higher modulation schemes and channel aggregation.

If the major chip companies see that there will be a market for these devices Microsoft believes low cost 802.11af-based devices could begin to enter the US market in about two years’ time. It cites work it has done with MediaTek, for example, on a triband 5 GHz/2.4 GHz/TVWS prototype based on 802.11af, which was used in a trial in Scotland, UK in 2015.

Microsoft made it clear that it was not getting into the telecoms business and it wants to act mostly as a stimulant for change. This is common practice for the company, whose overall objective is to boost usage of web services overall, and therefore its cloud services and apps. It has often worked to kickstart a new technology – it was active in the early days of WiFi, and less successfully when it invested in fixed wireless providers during the shortlived LMDS boom in the US at the start of the century.

The lessons from TVWS and CBRS have whetted the appetite of the US Internet industry for flexibly shared spectrum, as a middle ground between long term, rigid licences, and the wild west of entirely unlicensed bands. Shared schemes can make underused spectrum, in bands which have incumbents, usable for the first time, and they can encourage short term or dynamic usage, supporting a wide range of services and innovative providers.

Flexible usage, especially to improve rural access, is one of the topics at the center of the FCC’s latest enquiry, which it initiated at its August Open Meeting. This will focus on ‘mid-band spectrum’ between 3.7 GHz and 24 GHz, where it hopes to support the most efficient and flexible patterns of use, on both a licensed and unlicensed basis. The Notice of Inquiry (NOI) is looking for input on three bands (3.7-4.2 GHz, 5.925-6.425 GHz and 6.425-7.125 GHz). It is also asking respondents to identify other non-federal frequencies that may be suitable for expanded use for wireless broadband and other applications.

A section of this spectrum, in 3.7-4.2 GHz, has been attracting a high level of interest, and if Microsoft has anointed itself as chief flagwaver for TVWS (though Google is also a major player), the search giant is taking the lead in 3.7-4.2 GHz. This is considered a key target for shared or flexible usage because it is adjacent to the CBRS band in 3.5 GHz, and supporters are keen to use the rural broadband issue to justify their demands.

However, as in TVWS, there are incumbents (in this case broadcast and communications satellite operators, plus some fixed wireless operators) which will resist encroachment by wireless broadband providers.

Google’s parent company, Alphabet, is driving a new, as-yet unnamed, ad hoc coalition, whose founding members include AT&T, Verizon, T-Mobile, Google/Alphabet Access, Ericsson, Nokia, CTIA, Samsung, WiFi Alliance, Intel, Cisco, Comsearch, Broadcom, HPE and Information Technology Industry Council (ITI). They hope to build on some of the advances in spectrum sharing, and protection for incumbents, which are being proposed for the 3.5 GHz CBRS band, in order to have similar schemes in 3.7-4.2 GHz. They have yet to present their formal proposal, but their efforts have been praised by FCC commissioner Michael O’Reilly.

Alphabet is also supporting the Broadband Access Coalition (BAC), which is calling on the FCC to create a new licensed, point-to-multipoint (P2MP) fixed wireless service in the band. The law firm of Kelley, Warren & Drye wrote in a brief to support the idea: “Proponents smell an opportunity for a larger contiguous band which could bring equipment costs down and harmonize with international initiative in this same portion of the radio frequency spectrum.”

Alphabet wrote in a letter to the FCC: “Alphabet Access fully supports the larger goal of opening the 3700-4200 MHz band for licensed services and the 6 GHz band for unlicensed services, as contemplated in that NOI, while protecting incumbent systems from harmful interference. BAC’s proposals would allow limited deployments of point-to-multipoint systems, which would advance the Commission’s goal of expanding broadband service, especially in the remote and underserved areas where terrestrial service is critical and relatively few [fixed service] sites exist.”

BAC will be pleased with a recent FCC decision, to deny a request from CTIA – which represents the MNOs, always ambivalent about flexible spectrum as it may enable new competition. CTIA had asked the regulator to consider BAC’s petition along with all the other issues connected to the mid-band NOI, rather than potentially approving it separately before other topics were addressed.

A proposal from T-Mobile USA indicates the suspicion with which MNOs regard flexible spectrum. Although the MNOs have often managed to harness unlicensed public WiFi to enhance their own capacity and cost-effectiveness, for instance with WiFi offload or HetNets, they have also seen rivals like the cablecos using WiFi hotspots and homespots to build an extensive high speed wireless network without having to invest in licences.

TMO wants 3.5 GHz to be focused on 5G (as in other parts of the world including China) rather than LTE, and it has proposed that the FCC should auction all 150 MHz of the 3.5 GHz band as Priority Access Licences (PALs) with 10-year terms. That would favor large carriers and raise barriers to smaller service providers, and if approved, it would be likely to be extended to the 3.7-4.2 GHz band too.

The current rules for the 3.5 GHz CBRS spectrum establish three layers of access – top priority is for incumbent federal users, then there are the PALs, and the third layer is for unlicensed general access, with any provider able to use any unused channels, provided there is no interference to incumbents or PALs. The Qualcomm-inspired MulteFire technology, which allows LTE to run in unlicensed spectrum without a licensed anchor network, is targeting CBRS as well as 5 GHz.

Verizon is happy to move ahead with LTE usage in the CBRS band (also called Band 48). It claims it has conducted the first demonstration of licensed LTE-Advanced in the spectrum, working with Ericsson, Qualcomm and Federated Wireless. Federated provided its Spectrum Access System to “dynamically prioritize traffic” within the FCC’s spectrum sharing framework, the companies said. The demonstration was held in an Ericsson lab in Plano, Texas, and used 2×20 MHz LTE carriers with 256 QAM modulation on the downlink.

“The use of CBRS spectrum greatly advances our work in emerging spectrum bands,” said Nicola Palmer, Verizon’s wireless chief network officer, in a statement. “As industry leaders we work tirelessly to provide the best mobile experience available – that includes always innovating through new technologies and software platforms to better serve our customers.”

In February, Federated announced that its spectrum controller was granted conditional certification from the FCC, a step which opened the door to operator trials. The company has also, last year, demonstrated interoperability between its own SAS and the other major one, from Alphabet Access. Federated is also working on providing sensors for the Environmental Sensing Capability (ESC) which is also part of the FCC’s three-pronged sharing system.

Meanwhile, the CBRS Alliance, which promotes usage of the band and will certify compliant devices and equipment, has named CTIA to manage its LTE product certification program. CTIA joined the board of the CBRS Alliance earlier this year, represented by its VP for spectrum planning, Paul Anuszkiewicz, who also served with the Wireless Innovation Forum (WInnForum), which created technical specifications for CBRS. Certification programs, as the WiFi Alliance has clearly demonstrated, are essential for multivendor interoperability and to drive confidence and a broad ecosystem.

“CTIA shares our strong interest in the 3.5 GHz CBRS band and understands the importance of shared spectrum in the development of LTE-based technologies,” said Neville Meijers, VP of business development at Qualcomm Technologies and chairman of the board for the CBRS Alliance. “We look forward to collaborating with CTIA as we continue to pave the way for innovative solutions that will shape the future of wireless.”

The CBRS Alliance was founded a year ago by Federated Wireless, Google, Intel, Nokia, Qualcomm and Ruckus, and now has 70 members, including Ericsson and the four national MNOs.