The Citizens Broadband Radio Service (CBRS), which makes 150 MHz of shared spectrum available between 3,550 and 3,700 MHz, is arguably the FCC’s most interesting offering in years. If it delivers on even half of its promises, the agency will have created a genuinely useful opportunity for innovation. As spectrum sharing is exceedingly difficult to achieve without dissolving into chaos, CBRS rules are very specific in certain areas, although antennas receive surprisingly little treatment. Consequently, it’s the responsibility of those developing CBRS networks to fully understand their significance in determining overall performance and meeting FCC rules.
CBRS in a Nutshell
CBRS was formalized by the FCC In April 2015 with the goal of making additional spectrum available on a shared basis for a wide variety of uses, without creating interference to existing services. The agency hopes that the new allocation will enable wireless carriers to increase network performance, provide an alternative for Internet of Things (IoT) connectivity, allow cable companies to get into the wireless business, and perhaps of greatest interest, make it possible for organizations other than wireless carriers to operate private networks using digital (presumably LTE) technology for the first time. These are just a few cases where CBRS will have an impact—there are potentially many others.
CBRS offers two classes of users in addition to those already operating there (the incumbents), which are primarily coastal Navy radars and some fixed satellite services. The incumbents have priority access to the spectrum and protection from interference by devices in the lower classes. Below the incumbents are Priority Access Licensees (PALs), for which allocations are acquired through FCC auctions. PALs are protected from interference by the lowest-tier General Authorized Access (GAA) users. While GAA users are afforded no interference protection, they also pay nothing for spectrum access. Figure 1 shows the hierarchy of this architecture.
Each PAL license gives the owner a 10-MHz channel between 3,550 and 3,650 MHz in a small geographic area called a census tract (a portion of a county or other clearly-defined area that the Census Bureau uses to divide up the U.S.). Up to seven PALs can operate in each tract, in which a licensee can use only 70 MHz of the full 150-MHz allocation. Although GAA users can operate anywhere within the band not assigned to higher-tier users, only 80 MHz is available when PALs are in operation in a census tract. Digital transmission technologies are mandatory; while LTE will be the choice of most users, other wireless access methods can be used, as well.
As census tracts cover small areas, the auction price for each one should be a fraction of what wireless carriers pay for widespread coverage, allowing spectrum to be acquired by smaller organizations. Not surprisingly, wireless carriers seeking spectrum for backhaul, cable companies wishing to become mobile virtual network operators (MVNOs), large corporations wanting to create their own secure networks, the IoT industry, and components manufacturers looking for new sources of revenue all have been anxiously awaiting the arrival of CBRS. The only question that remains is if (or how well) spectrum sharing will work, as it hasn’t been terribly successful in the past.
The FCC mandate that incumbent users have ironclad interference, the PALs less, and GAA users none, makes for an “interesting” operating environment. Spectrum sharing is an extremely complicated process, as it requires a mechanism to ensure protection from interference among users and the various classes. Some spectrum sharing schemes use a “listen-before-talk” approach in which a device senses activity on a channel before transmitting.
If it detects the presence of another user, it moves to another channel. This is used by Wi-Fi, Bluetooth, and other standards with reasonable success in most environments. However, another technique called a Spectrum Access System (SAS) used in other spectrum-sharing scenarios such as the “TV white spaces” is also used by CBRS, but is more ambitious than its predecessors.
An SAS (there will be many within CBRS) works by managing and enforcing the shared frequencies using a cloud-based database of all CBRS devices to coordinate channel assignments and prevent interference. Among its many tasks are assigning channels for CBRS devices (called CBSDs); determining their maximum power at every location and ensuring it is not exceeded; registering and authenticating them; and receiving and addressing reports of interference from incumbent users. To protect them, sensors for Environmental Sensing Capability (ESC) will be deployed nearby each location to detect activity from other services. If it occurs, the sensor alerts the SAS, which commands the interfering emitter to change channels (Fig. 2).
There is no guarantee that the CBRS SAS will work this time, as managing hundreds of thousands of devices throughout more than 70,000 census tracts will be an enormously complex task. However, the CBRS SAS is arguably the most advanced such system to date. If successful, it will result in the rapid and successful deployment of many applications.
Antennas: Critical CBRS Elements
In every RF or microwave system, antennas are a key determinant of overall performance. In CBRS, they take on the major responsibility of helping to ensure that interference is kept in check. The CBRS rules call for two basic classes of CBSDs. Class A devices are suited for indoor use (effectively making them typical small-cell base stations) or low-power outdoor use. The maximum effective isotropic radiated power (EIRP) is 1 W (+30 dBm), although many will likely deliver less. The Class A CBSDs will generally be used with 2-dBi-gain omnidirectional antennas or directional antennas with up to 6-dBi of gain.
A Class B CBSD is targeted for outdoor use with a maximum EIRP of 50 W (+47 dBm). At this power level, an antenna delivering very high gain is well suited for fixed wireless applications, such as Wireless Internet Service Providers (WISPs).
A typical directional antenna that can be used for outdoor Class B systems that require directivity (such as point-to-point and point-to-multipoint) is the L-com HG3517DP-090 sector panel antenna (Fig. 3). It combines vertical and horizontal polarization, gain of 17 dBi, a 90-deg. beamwidth, and a 25-dB front-to-back ratio. As the antenna accommodates transmits and receive paths with different polarizations, unwanted signals from adjacent channels or co-located equipment can be attenuated.
The FCC doesn’t mandate what type of antenna should be used in a specific application. Therefore, it is up to the user to ensure that CBSDs adhere to the rules designed to mitigate interference. There is a broad array of antennas likely to be used for various CBRS applications, thereby presenting a significant challenge for network designers. A variety of factors must be considered, including the characteristics of the antenna, signal propagation, EIRP, terrain, and other factors, to ensure that signal levels at the mandated coverage boundary are within prescribed values.
All potential PAL and GAA users provide specific information about antennas and their characteristics with their registration, including geographic coordinates within an accuracy of ±50 m horizontal and ±3 m elevation. This and much more information must be reported to the respective SAS when the system is first activated. The managing SAS determines the PAL Protection Area (PPA) boundary by a contour based on maximum allowable radiated power, an RF propagation model, antenna height and gain effects, and radiation patterns.
Each CBSD is assumed to use a single antenna. If it has multiple antennas, such as an array for multiple-input multiple-output (MIMO) operation, the antennas must produce an aggregate signal with radiated power that conforms to all the registration parameters. CBSD transmissions must also be managed so that their received signal strength measured at the boundary with any co-channel PAL cannot exceed an average power level of −80 dBm.
If Category A CBSDs are located outdoors, their antennas must be no higher than 6 m above average terrain. Otherwise, they will be considered Category B CBSDs and subjected to their requirements (including mandatory professional installation). Information provided about CBSD Category-B devices must include antenna gain, beamwidth, azimuth, down-tilt angle, and antenna height. The rules provide a procedure for describing antenna gain to determine aggregate interference in the direction of a receiver. If the antenna beamwidth and the antenna pattern are not available, the SAS will assume the antenna is omnidirectional.
The FCC also does not dictate what technology an ESC device uses for detecting the presence of signals from an incumbent system. This means that ESC developers can choose any type of sensing technique that will produce the desired result. While this provides significant flexibility, it also requires that ESC designers consider the device’s antenna along with the other factors that collectively determine its performance.
CBRS is the first spectrum-sharing effort by the FCC that has made it from concept through various legal and other challenges to the stage where it is likely to be realized in operation, as the first systems should be deployed early in 2018. It will nevertheless take some time before the CBRS SAS approach can be evaluated based on actual performance.
That said, CBRS offers benefits to so many different applications. Unlike other spectrum shared approaches, it has “multi-partisan” support. It also doesn’t hurt that CBRS sweetens the deal for those sitting on the fence by offering new revenue-producing opportunities ranging from cellular backhaul to providing LTE coverage in rural areas. This creates private networks for various purposes including IoT connectivity, boosting the resources of WISPs, and allowing cable companies (among others) to create their own LTE networks.