For the last quarter century, wireless internet service providers (WISPs) have operated in relative obscurity, delivering broadband and voice service to homes and businesses in rural areas. These are the nearly forgotten territories on the wrong side of the digital divide, in places too sparsely populated for cable multiple system operators (MSOs) and telephone companies to bother with. Unfortunately, existing under the radar isn’t the best route to self-preservation today in an environment where spectrum is an increasingly precious commodity, and everyone wants his share. Fortunately, the Wireless Internet Service Providers Association (WISPA), the industry’s advocacy group, recently stepped up its marketing game to match its lobbying efforts in Washington. This didn’t come a moment too soon.
If you are unfamiliar with WISPs, perhaps a little history is in order. But first, a primer on how they work. A WISP is a fixed wireless access (FWA) provider that uses point-to-point microwave or millimeter-wave links between its towers for coverage extension and backhaul, and point-to-multipoint links from the towers to the customer premises. In addition to residences, they also serve businesses, municipal governments and other entities, sometimes in urban environments in addition to the more usual rural environments.
The data stream to the WISP can originate from dedicated circuits such as T1, T3 or DS3 carrier system lines, fiber-optic cable or, when none of these is obtainable, multiple microwave hops from somewhere else. The data stream provided by the source is then sent to a high point in the coverage area, such as a building, existing tower or whatever else is available. The subscriber receives the signal using an outdoor directional antenna bore-sighted at the tower (see Photo 1).
The signal captured by the antenna is typically sent to a radio (essentially a weatherized access point) from one of the major manufacturers that is mounted with the antenna. The radio delivers the signal via power over Ethernet (PoE) to a computer and then a Wi-Fi access point or both, for distribution throughout the home or business. At this point, the scenario is no different than broadband delivered any other way. Nearly every WISP also offers voice over Internet Protocol (VoIP) telephony as part of its service packages. Downlink data rates range from kilobits to tens of megabits per second, based on what the customer needs or can afford. Even higher data rates are also sometimes offered on a custom basis.
WISPs have traditionally used unlicensed radio-frequency spectrum in the industrial, scientific and medical (ISM) radio-frequency bands, first because the government charges nothing to use the frequencies, and second because unlicensed spectrum is readily available without the entanglements and high costs associated with licensed spectrum. The frequencies used by WISPs have followed developments in Wi-Fi standards, quickly moving from 915 MHz to 2450 MHz with the standardization of IEEE 802.11b, then to 5 GHz, and today to a combination of both. Operation is also conducted at 3650 MHz, and soon through shared usage of the 3.5-GHz Citizens Broadband Radio Service (CBRS) band, the rules for which are currently being disputed by WISPA, for reasons discussed later. Other frequencies include the TV white spaces (channels not used by TV stations in various geographic areas) between 500 MHz and 700 MHz (also shared), and 2.5 and 3.6 GHz for Long Term Evolution (LTE) high-speed wireless communications, in some cases.
In the Beginning
In many ways, the origin of the WISP industry resembles that of community antenna television (CATV), now commonly referred to as cable TV. As television became a must-have in the 1950s and 1960s, reception in some places was difficult if not impossible for communities either too far from the broadcast tower or obscured from a line-of-sight path by mountains or some other inhibitor of RF energy. To solve the problem, some enterprising minds mounted high-gain directional antennas at the highest point in the area and fed the captured TV signal to nearby homes and later to entire communities with coaxial cable as it still is done, with the help of fiber, today.
WISPs emerged as a means of delivering internet access and VoIP voice service rather than over-the-air TV in remote places where it has always been too expensive for cable and telephone companies to build the infrastructure required to serve a few hundred or even a thousand customers. In contrast, a WISP’s cost to serve a customer is a small fraction of this, which effectively makes their service possible, if not immensely profitable.
The industry got its start with Lawrence “Brett” Glass, an engineer with an MSEE degree from Stanford, known for designing integrated circuits (ICs) for Texas Instruments in Palo Alto, California, and known as a prolific contributor to computer trade publications. Glass decided to trade the frenetic density of Silicon Valley for the wide-open vistas of Wyoming, specifically Laramie. It didn’t take long for him to become frustrated by the lack of decent internet access — even at times no internet access at all in the area. Only the University of Wyoming had T1 circuits. At home, he was using CompuServe and dial-up at 2400 baud.
So, in 1992 he and a small group of like-minded people in the area created a users’ group that became a nonprofit co-op to provide wireless internet access throughout the area. They pooled their money to lease a T1 circuit for $6,000 a month, a ridiculously high price, but their only option. For wireless distribution, they used a new 915-MHz unlicensed wireless technology called WaveLAN (the precursor to IEEE 802.11) created by NCR Corp., with which Glass was familiar. They cobbled together a WaveLAN system, found a high location for the transmitter, and distributed the signals to the group’s members and local businesses.
In 2003, the membership of the co-op wanted to be customers rather than having the management responsibilities of co-op members, so Glass and his wife took Laramie Internet Access and Telecommunications (Lariat) public, thus forming the world’s first commercial WISP. Lariat remains in business and, as Glass noted several years ago, it is “growing (in coverage area) about the size of Manhattan” every year. He has become a vocal advocate for the WISP industry and FCC fairness, or the lack thereof, testifying before Congress about net neutrality, about which and other topics he speaks at various venues.
Today, the United States has between 1,000 and 4,000 WISPs, depending on the source of the data, and there are thousands more throughout the world. Their sheer number illustrates that many are still modest operations that serve small geographic areas with several hundred to a few thousand customers. The largest WISP in the United States, Rise Broadband, has more than 200,000 customers. With most estimates saying that about 23 million U.S. households lack adequate internet access, there’s plenty of room for the WISP industry to grow.
And it is indeed growing. The Carmel Group, a market research company that has produced an informative report on the industry, last year predicted that subscribers of fixed wireless services in the United States would nearly double to 8.1 million by 2021, with industry revenues expected to increase to more than $5.2 billion from $2.3 billion today. At last year’s Wispapalooza conference held by WISPA, the industry advocacy group, there were more than 1,800 attendees representing 51 states and territories and 31 countries, 36 sponsors, 131 speakers and 91 sessions.
WISPs are also the most cost-effective (and often the only) way to provide data and voice service in developing countries, where wired infrastructure is limited or nonexistent and would be prohibitively expensive to deploy. Finally, as the internet of things (IoT) gains momentum, WISPs are well suited to provide connectivity from the wireless network edge to the Internet, and this is becoming a new revenue stream.
From technological, economic and management perspectives, WISPs have come a long way, especially in recent years. For example, technologies developed for cellular and Wi-Fi ultimately find their way to the WISP industry, including multiple-input multiple-output (MIMO) communications, channel bonding, cloud-based network management and various other enhancements. The industry is also increasingly making use of LTE, bringing it in line with the cellular industry.
Even antennas, which KP Performance Antennas has been manufacturing for WISPs since 2008, have evolved to meet advances as they have occurred. For example, the sectored approach to serving a specific geographical area is almost universally used by the industry today, as the cellular industry has done for many years.
A typical example is KP’s KP-5HVX8-65 dual-horizontal/vertical-polarization MIMO sector antenna with a 65-degree azimuth beamwidth and eight ports facing the same direction within a 34-inch radome (see Photo 2). It delivers gain up to 17.5 dBi and covers 4.9 GHz to 6.4 GHz. Connecting two four-port radios to the antenna provides redundancy to increase capacity by using separate channels on each radio. Four of these antennas mounted to a tower provide 360-degree coverage, which can be increased to six antennas for more dense applications.
Fighting for Space
Only a few years after the WISP industry began to expand, it began to experience its first growing pains, for several reasons. First, traditional carriers have the resources and influence to fashion the regulatory environment to suit their purposes. This is evident in auctions in which the FCC sells even small swaths of spectrum for many millions of dollars, well beyond what any WISP could afford, and too large in coverage area to make sense for a provider serving a comparatively small geographical area.
Next is the problem of using unlicensed spectrum, which it must share at 900 MHz and 2450 MHz, for example, with other technologies ranging from non-WISP Wi-Fi hotspots to garage door openers, cordless phones, baby monitors and a variety of other consumer devices. This sharing causes interference, the bane of any wireless service because it degrades performance and becomes a major impediment for WISPs that must operate at the same low RF power levels as other unlicensed devices, even though they traverse long distances versus tens of feet for other applications.
In comparison, licensed spectrum would be ideal because WISPs would be free of the massive congestion of the ISM bands such as 2450 MHz, which today is densely populated even in rural areas. This situation has expanded to the 5-GHz bands used by IEEE 802.11n and IEEE 802.11ac, and soon to IEEE 802.11ax, coming in the next year or so. However, as noted, licensed spectrum is far too expensive and tailored primarily for mobile services covering large areas.
Opportunities and Trouble Brewing
In addition to these longstanding issues are two current potentially lucrative opportunities that have the potential to significantly improve WISPs’ ability to increase performance and capacity, the Citizen Broadband Radio Service (CBRS) and new allocations in the band from 3.7 GHz to 4.2 GHz.
In April 2015, the FCC formally established the CBRS, which uses spectrum-sharing in a three-tiered hierarchy to make 150 megahertz of spectrum available between 3550 MHz and 3700 MHz. This is a considerable number, considering that a typical wireless carrier owns about 130 megahertz of spectrum at various frequencies for all its networks.
This so-called innovation band is designed make it economically feasible for entities other than major carriers to create and operate new services such as private LTE networks. At first glance, this should be of major benefit to WISPs, which need all the spectrum they can get to serve customers’ needs for streaming services such as Amazon, Hulu and Netflix, and other bandwidth-intensive applications. And it may be feasible for WISPs, but as usual, the devil is in the details.
In the CBRS three-tiered hierarchy, incumbents occupy the highest rung and must be protected from interference by the two lower-tier operators. One tier below are Priority Access Licensees (PALs) that obtain by auction a 10-megahertz-wide channel in a single, small tract defined by the census bureau. They will operate within 100 megahertz (3550 to 3650 MHz) of the total available bandwidth and are protected from lowest-tier General Authorized Access (GAA) users for which access is free, but without interference protection. In practice, a GAA user should be able to operate in 80 megahertz of the total bandwidth in each tract.
The benefits for WISPs are the ability to obtain licensed spectrum through auction at far less cost than is the case for traditional auctions because the coverage areas are small. They could also avail themselves of the GAA tier, providing additional lightly regulated (although tightly controlled) spectrum.
The issue, and the reason for WISPA’s recent descent on the FCC and Capitol Hill is an FCC notice of proposed rulemaking (NPRM) that would alter the original CBRS rules. WISPs believe that the changes, which were requested by the carriers, would greatly reduce the benefit of CBRS to WISPs. In particular, they are concerned about proposed revisions that would 1) extend PAL licenses from 3 to 10 years and make them renewable without question and 2) expand PAL geographical sizes from census tracts to Partial Economic Areas designated by the FCC a few years ago that are much larger (only 416 nationwide) and best suit wireless carriers rather than WISPs. The result of these changes, according to
WISPA, would be a return to the traditional scenario existing in licensed bands, effectively excluding the nontraditional, innovative entities that CBRS was created in part to serve.
It would also potentially strand investments in LTE many WISPs have already made in the in the section of the band from 3650 MHz to 3700 MHz, which WISPA claims would reduce any incentive to spend more money on advanced technologies required to better serve more customers. WISPs also assumed they would be able to win some PAL licenses and upgrade this equipment to use the entire band.
That said, even if the FCC accepted the changes that the carriers are seeking, WISPs would be able to use the CBRS band as GAAs because half of the band will still not be auctioned, including the portion they’re currently using. They could also still use licensed CBRS spectrum in places where PALs haven’t deployed service, although this adds an element of uncertainty that many WISPs might consider too risky.
WISPA states its concerns concisely in its comments on the NPRM: “The rule changes sought by the incumbent wireless carriers would undercut each and every one of the important statutory requirements of [CBRS] and the public policy objectives clearly articulated throughout the CBRS Order without producing any alternative public interest benefit.
“The spectrum assignment changes they (the wireless carriers) propose … would drive up the costs of initial license procurement and thereby limit the pool of bidders, forcing out smaller and more innovative spectrum users that do not require large geographic areas to implement their business plans. The carriers wish to make the CBRS rules more like every other auctioned spectrum band in which, not coincidentally, the major wireless carriers have obtained the lion’s share of the licenses.”
The FCC is expected to rule on the carrier-requested changes in the next few months.
WISPA, along with long list of WISPs, manufacturers and other organizations petitioned the FCC collectivity as part of the Broadband Access Coalition to revise the current rules to allow the creation of a point-to-multipoint service in the band from 3700 MHz to 4200 MHz that is currently underused for serving rural areas. The FCC has already proposed such a service, and the coalition urges it to do so. However, WISPA’s concern is that this reallocation could also result in the same situation as it currently taking place with CBRS.
The FCC proposal would modernize deployment of high-speed, licensed point-to-multipoint fixed wireless broadband services (i.e., WISPs) while protecting incumbent fixed-satellite service and other incumbents on a shared basis. A significant difference between this allocation and CBRS is that there are no incumbent services to protect, so no complex spectrum management system would be required to eliminate interference.
This band has even greater potential benefit for WISPs than CBRS because it provides 500 megahertz of contiguous bandwidth that would accommodate bonding of the 40-, 80- and 160-megahertz-wide channels needed to achieve gigabit-level LTE service. The band also has propagation characteristics similar to those of the 3.5-GHz band, so, like CBRS, it would potentially make non-line-of-sight transmission paths possible.
As is probably obvious at this point, the WISP industry is growing in coverage, revenue and technical sophistication. However, it also faces daunting challenges, especially in availability and equitable use of spectrum. Challenges are nothing new for the industry, but several factors make the current situation arguably more important than ever.
Foremost, a confluence of circumstances has produced an environment that is, at least potentially, a watershed moment for the WISP industry. There finally appears to be more interest than usual in Washington about ensuring more people have at least one viable (i.e., high-speed) option for internet access. How this ultimately is resolved remains to be seen, but CBRS, making the band from 3700 MHz to 4200 MHz available for fixed wireless access, and potentially other opportunities should provide some benefits.
This spectrum is essential for WISPs to increase minimum downstream data rates to accommodate the inexorable growth of streaming because streaming requires wide channel bandwidths and thus not only more spectrum, but also frequencies unencumbered by the interference that has plagued WISPs for years on the unlicensed bands to which they have been relegated. The FCC presumably will make its decision concerning the revision of the CBRS rules, and the outcome will largely determine how well WISPs fare in the future.
Justin Pollock is an antenna engineer at KP Performance Antennas. Visitwww.kpperformance.com.