The benefits of coordinated shared spectrum
Mobile network operators (MNOs) continually strive to quench their customers’ seemingly inexhaustible thirst for data, and their response to this relentless growth in mobile data traffic is to improve network capacity and performance. This can be done in any or all of three ways. Network performance can be increased by improving spectral efficiency using the latest access protocols with advanced modulation and coding schemes. Spectral efficiency can also be increased through cell splitting and/or adding more sectors or overlaying small cells in the network. A third route to improved network capacity and performance is to add more spectrum and use larger channel bandwidths. However spectrum is a scarce and limited resource and efforts are underway to coordinate and share access to new spectrum in the 3.5 GHz band in the US.
One of these efforts is the Federal Communications Commission’s (FCC) Citizens Broadband Radio System (CBRS) for shared commercial use which makes available 150 megahertz of spectrum in the 3.5 GHz (3550 to 3700 MHz) band. Commercial users of this spectrum must share it with the incumbent U.S. Department of Defense/naval radar, wireless internet service providers (WISPs), and satellite earth stations. CBRS operates a three-tier priority access model defining Incumbent Users, Priority Access License (PAL) users, and General Authorized Access (GAA) users and access to the spectrum is coordinated by a cloud-based Spectrum Access System (SAS)/Environmental Sensing System (ESC) that distinguishes between Incumbent, PAL, and GAA users and used for detecting transmissions in the band.
There seems to be broad agreement that the FCC will certify the SAS/ESC operators towards the end of 2017 and commercial operations can start at that point. First will be fixed wireless for WISPs, and then later in 2018 we expect MNO network densification, new MSO entrants for mobile services, and an increase in private LTE and neutral host deployments. The indoor enterprise and home consumer gateway device market will ramp up in 2019, and this equipment may include embedded Citizens Broadband Radio Service Devices (CBSDs) or base stations in Wi-Fi cable modems with interoperability with major MNOs.
The promises of CBRS
CBRS promises to give MNOs an increase in capacity, to transform in-building wireless economics, to create private LTE networks run by non-traditional enterprise and venue operators and to offer an entry to mobile for multiple system operators/mobile virtual network operators (MSO/MVNOs). CBRS promises to attract not only MNOs, enterprise, and private virtual network operators (PVNOs), but also new entrants, thanks to its ability to offer densification, neutral host solutions, and low-cost deployments, and to assign spectrum with near real-time updates, unlike other shared spectrum approaches.
An MNO will view CBRS as a tool to improve the economics of network densification and coverage extension by leveraging early and low-cost access to a greater quantity of spectrum than can be had via refarming. The MNO also has the prospect of generating additional revenue by offering managed cellular services to enterprises and property owners with CBRS, while simultaneously coordinating with the macro network.
CBRS can transform the in-building wireless market by offering a low-cost alternative to distributed antenna systems (DAS). The low power (30dBm/1W) CBSD Class A radios are best suited to the small cell form factor for indoor or venue deployments. With deployment costs which promise to be similar to Wi-Fi, in-building wireless penetration in the largely untapped middle-sized and enterprise verticals will increase dramatically with CBRS since these venues could not justify an expensive DAS system.
CBRS will make it possible for any property owner or enterprise to create their own private LTE network. Private LTE networks will find many applications in many verticals such as industrial, mining, ports, entertainment, transportation, and enterprises. Private LTE networks, unlike public LTE networks, are managed locally, and the network is tailored and optimized for the local services and applications for which it is used. Sensitive or proprietary data stays local, as may be the case in manufacturing or enterprise applications.
MSOs can leverage CBRS to expand their services to quad-play by offering a supplementary mobile service as an MVNO. Using their existing cable infrastructure MSOs can offer 4G cellular in addition to cable TV, telephone and broadband internet. MSOs can also offer to offload MNO traffic transiting a host network in return for reduced fees, or even negotiate a “swap” agreement with the MNO offering roaming onto CBRS in return for reduced fees.
The devil’s details: stranded assets and stranded spectrum
As it stands, there remain a few open questions surrounding CBRS, including GAA user coexistence and mechanisms to reduce mutual interference since GAA users are not guaranteed protection from interference from each other in the same channel. Also, the PAL competitive auction rules, license duration, and geographical area all remain to be defined by the FCC.
In recent submissions to the FCC, the CTIA and T-Mobile US, now supported by AT&T and Ericsson, lobbied to change the current PAL license term from 3 years to a standard 10-year license term and to allow PAL license holders to expect renewal after the 10-year term expires. Under the current rules only the first PAL license holder in a census tract may expect to renew for a further 3 years. With a renewable 10-year duration a PAL license holder has an increased incentive to build out a network for 5G since the return on investment from the network over a 20-year period would be more favorable. This change, according to CTIA/T-Mobile US, would avoid the risk of stranded assets under the current rules with the prospect of losing a PAL license after 3 years.
CTIA/T-Mobile US also lobbied the FCC to enlarge the PAL areas from census tracts, of which there are 74,134 nationwide, to the much larger Partial Economic Areas (PEAs), of which there are 416. CTIA/T-Mobile US also asked the FCC to grant a PAL regardless of how many applications are received. Under the current framework, no PALs would be granted if only one company requests a PAL in a specific area, potentially giving rise to geographical areas without a PAL tenant and no continuity of service, effectively stranding the spectrum. The current PAL rules dictate that if there is insufficient demand for the seven PAL licenses in a geographic area then the FCC should proceed to auction one less license than there are applicants. This so-called “n-1” rule risks phasing out PALs at each renewal period – again stranding spectrum.
In counter arguments public interest groups have said that PAL markets in areas larger than census tracts have the potential to undermine the business case for private networks and in-building wireless. Very large PAL geographic areas are incompatible with what will be predominately small cell deployments in indoor and high capacity zones and would not maximize the number of users in the band. This in turn limits the potential for highly localized mobile data offload, neutral host LTE, and private networks for utility, enterprise, smart city, and other uses. These groups also noted that large area licenses have the potential to strand spectrum by leaving it unused in low-density markets, whereas the current framework would allow users access. Finally, these changes have the potential to exclude many smaller users, such as private network operators, utilities, municipalities, and others since the cost of a PAL license will increase drastically. The current census tract based rule is an attempt to use market forces to match license cost to population density – the higher the population density the higher the license cost.
The way forward
If the FCC finds itself unable to reject the CTIA/T-Mobile US petitions, reaffirm the current rules, and move quickly to certify the SAS/ESC operators, then a way forward would be to adopt 10 year/census tract based rules.
The current three-year license term with no expectation of renewal will limit the uptake of PALs, and ABI Research believes the FCC should relax this term in line with the CTIA/T-Mobile US request. However, we believe the large-area PEA license will hinder innovation in this band and exclude smaller operators who may not have the capital to purchase a costly 10-year PEA-based license, and also the disrupt the business models of some vendors who have already entered the US market under the assumption that the current CBRS license plan will be used in the future.
The FCC should move to improve the CBRS rules by adopting a 10-year census tract-based scheme, simplify the PAL auction rules and resist making significant changes to the CBRS framework so that the last two years work from those in the CBRS ecosystem can be brought to market quickly and the band can be put to commercial use as soon as possible.
image source: air802.com