Rural wireless broadband has never been blessed with an embarrassment of riches, namely radio-frequency spectrum at 900 MHz, 2.4 GHz and 4.9 GHz, and most recently, 3.5 GHz. Users of each frequency band are either unlicensed or lightly-regulated, and all of the bands are densely populated. Only the 900-MHz accommodates non-line-of sight (NLOS) propagation.
Fortunately, rural broadband service providers can now exploit TV white space (TVWS) between 470 MHz and 698 MHz in the United States and between 470 MHz and 786 MHz elsewhere. At these frequencies, radio wave propagation is much more favorable to NLOS communications, so things are indeed looking up.
Nearly 1 million U.S. residences lack wireless service fast enough to fit the FCC’s definition of broadband speed, which is 25 Mbps downstream and 3 Mbps upstream. Most of these homes are in rural areas where laying fiber is too expensive and customers are too few for major carriers to make a profit.
In the early 2000s, a group of technically astute individuals took matters into their own hands, creating point-to-multipoint wireless services. They became known as wireless internet service providers (WISPs). Their number has grown from a few small companies to about 1,500 companies. A few have several hundred thousand customers. Among WISPs, downstream and upstream wireless service speeds vary, but the speeds have consistently increased over the years as technologies advanced and WISPs took advantage of them.
One of the biggest challenges faced by the WISP industry is its reliance on frequencies unsuitable for NLOS operation. The inability to provide wireless internet service where obstructions block line-of-sight signals limits the number of serviceable customers and restricts the maximum distance between a WISP’s main transmitter and the customers it serves. The NLOS problem can be mitigated to some degree by adding equipment, but this step is generally not cost-effective. Although recent actions by the FCC, including creating the Citizens Broadband Radio Service (CBRS) at 3.5 GHz, provide more spectrum for WISPs, all operate at higher frequencies that do not solve the NLOS propagation problem as well as the TVWS spectrum does.
TVWS frequencies consist of unused spectrum between 470 MHz and 790 MHz assigned for over-the-air TV channels that have traditionally been used as guard bands to mitigate interference. The guard band TV spectrum is either not used at all or is used in only a few places throughout the country. In its quest to make more spectrum available, the FCC and its Canadian counterpart, Industry Canada, adopted rules to allow unlicensed services to operate in the guard bands, also known as white spaces. Most of the spectrum lies unused in rural areas, precisely where high-speed broadband wireless service is most needed. In addition, as remaining analog low-power U.S. TV broadcasters will eventually go off the air, interference issues should decline with time.
Although the newly available spectrum is for unlicensed use, it is nevertheless regulated and users must adhere to FCC guidelines, the most significant of which mandates that unlicensed users must protect incumbents (i.e., mostly TV stations) from interference. The ingredients required to make this work are geolocation capability, user devices with spectrum-sensing ability and spectrum management administered by parties approved by the FCC that identifies vacant TV channels at specific locations at specific times.
If this sounds familiar, it’s because CBRS uses the same approach. It has taken a long time to be sorted out, and making it work is far from a trivial task. Nevertheless, the appeal of TVWS (as well as CBRS) has driven extensive development of white-space databases as well as cognitive radio technology.
The use of databases to facilitate spectrum-sharing underpins WISP use of TVWS and other unlicensed use of CBRS. In the last few years the sophistication in database use has increased dramatically. For example, to identify vacant channels, end-user devices must be equipped with geolocation capability and must be able to access a database through the internet to identify incumbent users.
In 2013, Google was selected to be the database administrator for TVWS in the United States, thanks to its well-known and massive amount of geolocation and other data from Google Maps, Google Earth and Street View, and a database that reveals what spectrum is free in what areas. The company has since become an administrator for CBRS, which uses an even more elaborate approach called a Spectrum Access System (SAS). Other administrators include Federated Wireless, Key Bridge, Comsearch and Sony.
In contrast to microwave bands typically used to deliver rural broadband wireless service, TVWS frequencies are much lower in frequency where signals travel farther and aren’t rendered useless by obstructions. When WISPs use high-gain antennas (see Figure 1), the range using low-power transmitters can reach up to 30 miles depending on terrain and other factors. This enormous improvement makes the use of TVWS a breakthrough for WISPs.
VHF and UHF Challenges
Operating broadband wireless service at VHF and UHF frequencies brings its own challenges, some of which are not encountered at microwave frequencies. An example is radio wave propagation that changes with atmospheric conditions, possibly causing unintended interference. Furthermore, operators of TVWS have to contend with horizontally polarized noise created by TV broadcast stations that still use a limited portion of the spectrum in different geographical areas. Nevertheless, the benefits of using TVWS outweigh these issues, because VHF and UHF frequencies offer cost-effective coverage never before possible to deliver rural broadband wireless service.
Several Institute of Electrical and Electronics Engineers (IEEE) network standards have been created specifically for use in TVWS, the most widely used being IEEE 802.22, the core component of which is cognitive radio. This network technology, which falls within the broad umbrella of dynamic spectrum management, autonomously monitors specific frequencies and when a signal is present reports its findings to the base station and potentially other radios with cognitive capability.
The response dictates whether a radio should autonomously perform one or more functions to ensure the frequency’s occupant receives protection from interference. Mitigating steps the radio may take include changing frequency, signal bandwidth or output power. A half-dozen variants of cognitive radio may be used either singly or in combination, depending on the application, whether the spectrum is licensed or unlicensed, or if a database of known incumbents is used.
Regardless of the chosen technique, cognitive radio represents a form of spectrum-sharing that allows multiple services to use the same band of frequencies. Cognitive radio relies on technologies within the domain of software-defined radio. In an ideal world where spectrum is infinite, no one would voluntarily choose to use spectrum sharing. In the real world where spectrum is limited, spectrum sharing probably will be more widely used in the future.
The other cognitive radio standard, IEEE 802.11af, also known as Super Wi-Fi, has also been developed for white space applications and others. It is a member of the 802.11 Wi-Fi standards. It is being used by some system manufacturers rather than 802.22.
Many companies have been taking advantage of the new rules and have been delivering solutions. One such company is Microsoft, which created the Rural Airband Alliance that takes the approach that the best solution for solving the broadband digital divide is through a combination of whatever technologies are available in a given area, including TVWS. The suggestion is that to deliver maximum throughput to the greatest number of people in a coverage area, a wireless service provider should use microwave frequencies as well as TVWS to achieve the best results. The broadband digital divide is the name given to highlight the difference in service levels often available in cities compared with rural areas.
In addition to helping to resolve technological issues, the alliance will provide training and will help with technology licensing, in addition to other philanthropic activities. It has also established partnerships with a variety of companies throughout the world. Specific locations include Declaration Networks in Virginia and Maryland and Allband Communications and Packerland in Michigan.
One of the alliance’s industry partners is Redline Communications, which offers a product called virtual fiber. It uses TVWS frequencies, 256 quadrature amplitude modulation (QAM), 2×2 multiple-input multiple-output (MIMO) radios with 10 bits per hertz spectral efficiency, latency less than 10 milliseconds and a claimed downstream speed of 186 Mbps over distances up to 30 miles.
Instead of using the 802.22 standard, another member of the alliance, Carlson Wireless Technologies, uses 802.11af to cover 10 miles with line-of-sight radio wave propagation and 15 miles in NLOS conditions. This method gives Carlson the ability to serve as many as 90 customers with a single base station. Downstream data rates are as fast as 18 Mbps using 64 QAM per sector and round-trip latency is less than 35 milliseconds.
Adaptrum, a company with its headquarters in the United Kingdom, last year introduced the first use of carrier aggregation to increase capacity, demonstrating a wireless speed of 30 Mbps using two 8-megahertz-wide carriers, and latency of less than 15 milliseconds.
Deploying spectrum-sharing networks has proven to be extremely challenging, and the spectrum management technologies required to keep interference in check have been some of the most difficult to perfect. As the first application to put spectrum-sharing into practice, TVWS has suffered its share of hiccups, notwithstanding the concerns of the broadcast industry and other users of the TVWS spectrum that have been continuous since the beginning. As the deployment of TVWS increases, the remaining interference concerns will hopefully be resolved to everyone’s — or almost everyone’s — satisfaction.
In addition, other TVWS technology development had been in the works even before the FCC released its final rules, and the pace of advancement in spectral efficiency and capacity has dramatically accelerated in the last few years (see Figure 2). The benefits provided by TVWS lower-frequency operations may begin to close the rural broadband wireless service gap for the first time. And, it has been a long time in coming.
Justin G. Pollock, Ph.D., is a senior antenna engineer at KP Performance Antennas and RadioWaves, which are subsidiaries of Infinite Electronics. He is the technical lead on the product development of industry-leading antenna technologies. He has co-authored refereed journal, conference and white papers for leading publications in the field of RF and microwave engineering, antennas, physics and optics.