# Amendment of the Commission's Rules With Regard to Commercial Operations in the 3550-3650 MHz Band

> Briefs, arguments, decisions, and more.

URL: https://www.frixlaw.com/law-library/documents/fr%3A2016-14505

## Record

- **Collection:** Federal Register
- **Document type:** Rule
- **Published:** July 26, 2016
- **Citation:** 81 FR 49024

## Text

FEDERAL COMMUNICATIONS COMMISSION
47 CFR Parts 1, 2, and 96
[WT Docket No. 12-354; FCC 16-55]
Amendment of the Commission's Rules With Regard to Commercial Operations in the 3550-3650 MHz Band

AGENCY:

Federal Communications Commission.

ACTION:

Final rule; petition for reconsideration.

SUMMARY:

In this document, the Federal Communications Commission addresses eight petitions for reconsideration on certain rules adopted in the
Report and Order
(
Report and Order
) in this proceeding governing the Citizens Broadband Radio Service in the 3.5 GHz band. The Commission also finalizes the regulatory scheme established in the
Report and Order
to make this spectrum available for wireless broadband through dynamic sharing among three tiers of users.

DATES:

Effective August 25, 2016 except for §§ 1.9046, 96.3, 96.17(b), 96.25(c)(1)(i), and 96.32(a) and (b) which contain information collection requirements subject to approval by the Office of Management and Budget. The Federal Communications Commission will publish a document in the
Federal Register
announcing the effective date for those sections.

FOR FURTHER INFORMATION CONTACT:

Paul Powell,
Paul.Powell@fcc.gov
, of the Wireless Telecommunications Bureau, Mobility Division, (202) 418-1618. For additional information concerning the Paperwork Reduction Act information collection requirements contained in this document, contact Cathy Williams at (202) 418-2918 or send an email to
PRA@fcc.gov.

SUPPLEMENTARY INFORMATION:

This is a summary of the Commission's Order on Reconsideration and Second Report and Order in GN Docket No. 12-354, FCC 16-55 released on May 2, 2106. The complete text of the public notice is available for viewing via the Commission's ECFS Web site by entering the docket number, WT Docket No. 12-354. The complete text of the public notice is also available for public inspection and copying from 8:00 a.m. to 4:30 p.m. Eastern Time (ET) Monday through Thursday or from 8:00 a.m. to 11:30 a.m. ET on Fridays in the FCC Reference Information Center, 445 12th Street SW., Room CY-B402, Washington, DC 20554, telephone 202-488-5300, fax 202-488-5563.

The Commission will send a copy of this Order on Reconsideration and Second Report & Order in a report to be sent to Congress and the Government Accountability Office pursuant to the Congressional Review Act, see 5 U.S.C. 801(a)(1)(A).

I. Introduction and Executive Summary

1. In this
Report and Order and Order on Reconsideration
(
Second Order
) we finalize the rules governing the innovative Citizens Broadband Radio Service in the 3550-3700 MHz band (3.5 GHz Band). Facing ever-increasing demands of wireless innovation and constrained availability of clear sources of spectrum, the Citizens Broadband Radio Service is an opportunity to add much-needed capacity through innovative sharing. With this
Second Order,
we finalize the regulatory scheme we created in 2015, putting in place the last rules necessary for this service to become commercially available. (80 FR 36163, June 23, 2015)

2. The Citizens Broadband Radio Service takes advantage of advances in technology and spectrum policy to dissolve age-old regulatory divisions between commercial and federal users, exclusive and non-exclusive authorizations, and private and carrier networks. The regulatory framework takes from recommendations from the President's Council of Advisors on Science and Technology (PCAST) and substantial engagement and input from stakeholders representing a cross section of the communications, technology, and public interest realms.

3. The comprehensive regulatory scheme adopted in the
3.5 GHz R&O
included specific licensing, technical, and service rules to enable dynamic sharing between three tiers of users in the 3.5 GHz Band. The Spectrum Access System (SAS) is the advanced frequency coordinator (or coordinators) necessary to assign rights and maximize efficiency in the band. The SAS(s) will incorporate information from the Environmental Sensing Capability (ESC), which will be used to increase available spectrum in coastal areas while continuing to protect incumbent Department of Defense (DoD) radar systems.

4. In this
Second Order,
we reaffirm the regulatory approach adopted in the
3.5 GHz R&O.
In doing so, we deny several petitions for reconsideration that are inconsistent with our goals and grant others that advocated rule modifications that would facilitate more equitable and efficient use of the 3.5 GHz Band. In the
3.5 GHz R&O,
we developed a comprehensive approach intended to balance consideration of the complex issues and competing considerations involved in creating a sharing regime in this band, and each rule is a vital part of that approach. We reaffirm our commitment to add much needed capacity spectrum to the marketplace through innovative sharing rules and techniques, and believe the rules established in the
3.5 GHz R&O
are the best means to do so.

5. Nonetheless, we do agree with some petitioners who argue for an increase in the power level for non-rural Category B CBSDs and greater flexibility in how to measure and direct the power. This will provide additional flexibility for all CBSD deployments to potentially increase their utility, and create additional flexibility for non-rural deployments. While rejecting arguments both to increase and to decrease our out-of-band emission (OOBE) limits for CBSDs, we revise our measurement of such limits to conform to the well-established root mean square (RMS) measurement technique reflected in our rules for other services. We also adopt a limited exception to the PAL assignment rules that would allow a single PAL to be issued in License Areas located in Rural Areas in the absence of mutually exclusive applications. At SIA's request, we also revise our rules to make clear that SASs must be capable of receiving and responding to interference complaints from Fixed Satellite Service (FSS) earth station licensees.

6. While we created a robust and substantial regulatory framework in the
3.5 GHz R&O,
there were several technical issues that required further refinement and input on the record. To bolster the record on these issues, we released a
Second FNPRM
seeking comment on how to: (1) Define “use” of Priority Access License (PAL) areas to determine the availability of spectrum for General Authorized Access (GAA) use; (2) implement and promote a robust secondary market in the band; and (3) optimize protections for licensed in-band and out-of-band FSS earth stations.

7. These are important issues, and are fundamental to the fabric of the Citizens Broadband Radio Service. They explore how to maximize the efficient use of spectrum by allowing opportunistic GAA use of spectrum when and where it is not utilized by Priority Access Licensees. They look at how we can maximize the amount of spectrum available in the band by optimizing the protection of in-band and out-of-band FSS earth stations, while leveraging the SAS and other tools to maximize operations towards the 3700 MHz band edge. They examine how to create reliable and flexible secondary market rules that can be implemented across hundreds of thousands of licenses.

8. In resolving these final issues, we strive to establish simple rules that are implementable in the near term, while protecting DoD radar systems consistent with the process and procedures established in the
3.5 GHz R&O.
We establish a definition of use that allows Priority Access Licensees to certify the extent of their service area to an SAS, while also establishing a maximum point at which they will receive protection. This is a both a flexible and objective way to allow Priority Access Licensees to design and deploy networks, and SAS Administrators to provide objective protection and effective GAA access. We authorize “light-touch leasing” to allow Priority Access Licensees to leverage the secondary market to provide access to any qualified lessee with minimal administrative requirements or transaction costs. Finally, we establish protection criteria for in-band FSS, and out-of-band FSS sites used for telemetry, command, and control (TT&C) that provides a high level of reliability, while also allowing the SASs to optimize based on the characteristics of the FSS earth station, the terrain, the CBSD deployment characteristics near the site, and other factors.

9. With these decisions, we complete the regulatory framework for the Citizens Broadband Radio Service, and set the stage for the commercial availability of a contiguous 150 megahertz of spectrum for wireless broadband use.

II. Procedural Background

A. 3.5 GHz NPRM, Licensing Public Notice, and FNPRM

10. As part of its ongoing efforts to address the growing demand for fixed and mobile broadband capacity, the Federal Communications Commission (FCC or Commission) released a
Notice of Proposed Rulemaking
(78 FR 1188, January 8, 2013) in December 2012 proposing to make an additional 100 megahertz (or up to 150 megahertz under a supplemental proposal) of spectrum available for shared wireless broadband use. Specifically, the
NPRM
proposed to create a new Citizens Broadband Radio Service. The technical rules focused on the use of low-powered small cells to drive increases in broadband capacity and spectrum reuse and an SAS that would coordinate multiple tiers of users.

11. In November 2013, in response to comments received on the record up to that point, the Commission released the
Licensing PN
(78 FR 73794, December 9, 2013), which described a Revised Framework that elaborated upon some of the licensing concepts and alternatives set forth in the
NPRM.
The Revised Framework retained the three-tier model proposed in the
NPRM
but expanded eligibility for access to the Priority Access tier with competitive bidding for assigning licenses within that tier. Like the
NPRM'
s main proposal, the Revised Framework cited the unique capabilities of small cell and SAS technologies to enable sharing among users in the Priority Access and GAA tiers.

12. In April 2014, the Commission released the
3.5 GHz FNPRM
(79 FR 31247, June 2, 2014), proposing specific rules for a new Citizens Broadband Radio Service in the 3.5 GHz Band to be codified in a new proposed part 96. The
FNPRM
built upon the concepts and proposals set forth in the NPRM and the
Licensing PN
and reflected the extensive record generated in the proceeding. Notably, the
3.5 GHz FNPRM
proposed to: (1) Implement the three-tier authorization model proposed in the NPRM; (2) establish Exclusion Zones based on recommendations set forth in the Fast Track Report to ensure compatibility between incumbent federal operations and Citizens Broadband Radio Service users; (3) create an open eligibility authorization system for Priority Access and GAA operations; (4) establish granular, exclusive spectrum rights for the Priority Access tier, consistent with parameters discussed in the
Licensing PN;
(5) set a defined “floor” for GAA spectrum availability, to ensure that GAA access is available nationwide (subject to Incumbent Access tier use); (6) set guidelines to allow contained access users to request up to 20 megahertz of reserved frequencies from the GAA pool for use within their facilities; (7) establish baseline technical rules for fixed or nomadic base stations operating in the 3.5 GHz Band; and (8) set guidelines for the operation and certification of SASs in the band. The
FNPRM
also sought comment on: (1) Protection criteria for Incumbent Access users; (2) potential protection of FSS earth stations in the 3700-4200 MHz band (C-Band); (3) competitive bidding procedures for resolving mutually exclusive applications for Priority Access Licenses (PALs); and (4) the possible extension of the proposed rules to include the 3650-3700 MHz band.

B. Report and Order and Second Further Notice of Proposed Rulemaking

13. On April 17, 2015, the Commission released the
3.5 GHz R&O,
which established the Citizens Broadband Radio Service under a new part 96 of the Commission's rules. The
3.5 GHz R&O
established a three-tier framework for making the entirety of the 3.5 GHz Band available for shared commercial use utilizing an SAS to coordinate operations between and among users in different tiers. This three-tier sharing framework is largely consistent with the proposals put forth in the
FNPRM.

14. Incumbent Access users represent the highest tier in this framework and receive interference protection from all Citizens Broadband Radio Service users. Protected incumbents include federal shipborne and ground-based radar operations and FSS earth stations in the 3600-3700 MHz band and, for a finite period, grandfathered terrestrial wireless operations in the 3650-3700 MHz portion of the band. Non-federal incumbents must register the parameters of their operations with the Commission and/or an SAS to receive protection from Citizens Broadband Radio Service users (47 CFR 96.15, 96.17, 96.21). In addition, an ESC may be used to detect transmissions from DoD radar systems and transmit that information to an SAS to ensure that federal Incumbent Users are protected from interference (47 CFR 96.15, 96.67).

15. The Citizens Broadband Radio Service itself consists of two tiers—Priority Access and GAA—both assigned in any given location and frequency by an SAS. Priority Access operations receive protection from GAA operations. A PAL is defined as a non-renewable authorization to use a 10 megahertz channel in a single census tract for three years. PALs will be assigned via competitive bidding in up to 70 megahertz of the 3550-3650 MHz portion of the band. One Priority Access Licensee may hold up to forty megahertz of PALs in any given census tract at any given time (47 CFR 96.25, 96.29).

16. GAA use will be licensed by rule throughout the 150 megahertz band. Both Priority Access and GAA use will be assigned and coordinated by an SAS, which will also perform additional coordination functions as set forth in the rules. GAA users will be permitted to operate on any frequencies not assigned to PALs. GAA users will receive no interference protection from other Citizens Broadband Radio Service users, including other GAA users, and must not interfere with higher tier operations.

17. The
Second FNPRM,
which was released along with the
3.5 GHz R&O,
sought comment on how to define “use” by Priority Access Licensees and whether the Commission should rely on an engineering definition, an economic definition, or a hybrid of the two to

determine whether frequencies are in use. The
Second FNPRM
also sought comment on the applicability of existing secondary market rules to PALs and the appropriate administration of secondary market transactions in the band. Finally, the
Second FNPRM
a sought comment on the methodology and parameters for protecting in-band and C-Band FSS earth stations.

18. After the adoption of the
3.5 GHz R&O,
and as directed therein, on October 23, 2015, the Wireless Telecommunications Bureau (WTB) released a Public Notice (80 FR 69662, November 10, 2015) seeking comment on the appropriate methodology for determining the contours for protecting existing 3650-3700 MHz wireless broadband licensees from Citizens Broadband Radio Service users during a fixed transition period. Finally, as directed by the Commission in the
3.5 GHz R&O,
WTB and the Office of Engineering and Technology (OET) released a Public Notice seeking proposals for future SAS Administrator(s) and ESC operator(s) in the 3.5 GHz Band. The Public Notice summarized the requirements for both SAS Administrators and ESC operators, as established in the
3.5 GHz R&O,
and described the process for submitting proposals. It also briefly described the process that WTB/OET will use to evaluate prospective SAS Administrators and ESC operators.

C. Petitions for Reconsideration

19. Petitions for Reconsideration on the
3.5 GHz R&O
were due July 23, 2015. The following eight parties filed petitions for reconsideration: CTIA, Jon Peha, Motorola Solutions, NAB, Nokia Solutions, SIA, Verizon, and WinnForum (80 FR 59705, October 2, 2015). The arguments raised in these petitions are described in greater detail in the relevant sections of the
Second Order.

20. CTIA—The Wireless Association Petition. CTIA seeks revisions to the licensing process for PALs, arguing that the Commission should adopt a five-year license term with a renewal expectancy. CTIA asks the Commission to reconsider its decision not to award a PAL in census tracts unless there are mutually exclusive applications. CTIA also seeks change to the technical rules, including changes to the OOBE limits and the measurement procedure for such limits. Finally, CTIA requests that the Commission increase the maximum effective isotropic radiated power (EIRP) and conducted power limits for Category A and Category B CBSDs.

21. Jon Peha Petition. Jon Peha seeks reconsideration of the Commission's decision that “when there is only one applicant for one or more PALs in a given census tract, we will neither proceed to an auction nor assign any PAL for that license area.” Instead he argues that the Commission should grant PALs in every market where there is demand, even if there is only one bidder.

22. Motorola Solutions Petition. Motorola Solutions supports WinnForum's Petition and also seeks reconsideration of the Commission's decision to only issue PALs where two or more parties file an application.

23. NAB Petition. NAB asks the Commission to eliminate professional installation as a method to report the geographic location of a CBSD to an SAS. NAB contends that location data should be reported automatically by a mandatory geo-location capability built into the device.

24. Nokia Solutions Petition. Nokia Solutions asks the Commission to increase the response time from when an ESC communicates it has detected a signal from a federal system in a given area that the SAS must either confirm suspension of the CBSD's operation or relocation from 60 seconds to 600 seconds. Nokia Solutions also argues that the Commission should specify emission limits for End User Devices that are compliant with 3GPP specifications. Nokia Solutions seeks changes to the power limits, asking that the total transmit power for CBSDs be stated simply as maximum EIRP and increased by 6 dB for Category A and 9 dB for Category B CBSDs. Finally, Nokia Solutions asks that the Commission revise the vertical location accuracy requirements to align with US Government Position Accuracy standard for outdoor installation and remove such requirements for indoor installations.

25. SIA Petition. SIA seeks changes to a variety of technical rules and aspects of the FSS protection rules. Among other things, SIA states that the Commission should adopt a stringent OOBE limit at 3680 MHz to protect C-Band operations immediately above the 3700 MHz band edge. SIA also argues that the Commission should: (1) Decrease the maximum power limits for CBSDs; (2) reduce the 60-second timeframe for a CBSD to confirm deactivation or a change in frequency; (3) eliminate or clarify the annual registration requirements for FSS earth stations; (4) establish procedures for reporting FSS interference to SASs and implementing immediate shutdown procedures in response to such reports; and (5) reconsider the freeze on new co-primary FSS earth stations in the band.

26. Verizon Petition. Verizon seeks reconsideration of the power limits, stating that the Commission should increase the EIRP to levels closer to real-world small cell deployments and to rely solely on EIRP rather than imposing limits on both EIRP and conducted power.

27. WinnForum Petition. The WinnForum asks the Commission to reconsider a number of the technical rules governing the 3.5 GHz Band. WinnForum argues that the Commission should: (1) Increase the reconfiguration response time from when an ESC communicates it has detected a signal from a federal system in a given area that the SAS must either confirm suspension of the CBSD's operation or relocation from 60 seconds to 600 seconds; (2) increase Category A and Category B CBSD EIRP limits and provide additional flexibility between EIRP and conducted power limits; and (3) modify the geo-location rules to allow SASs to estimate CBSD elevation above ground level for purpose of determining vertical location accuracy.

D. Oppositions and Replies to Petitions for Reconsideration

28. Oppositions to the petitions for reconsideration were due October 19, 2015, and replies to oppositions were due October 29, 2015. Eight parties filed responses. The arguments raised in these oppositions are described in greater detail in the relevant sections of the
Second Order.

29. CTIA Opposition. CTIA opposes SIA's petition and supports the petitions filed by Jon Peha and Motorola Solutions. CTIA asks the Commission to reject SIA's request to impose stricter OOBE limits and states the 3.5 GHz FNPRM provided adequate notice that that the Commission would extend these limits for the 3650-3700 MHz band. CTIA claims the power limits for non-rural Category B CBSDs should be increased to provide operators with additional flexibility. Finally, CTIA supports Jon Peha's and Motorola Solutions' request that the Commission issue PALs in all census tracts, even if there is only one applicant.

30. Federated Wireless Opposition. Federated Wireless asks that the Commission take the following actions in response to the petitions for reconsideration: (1) Increase maximum EIRP and conducted power limits for CBSDs; (2) modify the elevation accuracy requirement to allow the SAS to play a role in determining CBSD location; and (3) allow PALs to be issued even when there is a single applicant in a given census tract. Federated Wireless also asks the

Commission to reject the petitions that seek elimination of the option to allow a professional installer to report geo-location and petitions that request adoption of a maximum antenna height limitation for Category B CBSDs.

31. Google Opposition. Google argues that the Commission should reject SIA's request to strengthen OOBE limits and eliminate registration requirements for FSS earth station operators. Google also argues that professional installation can protect incumbents and the Commission should retain this option to report geo-location accuracy and that the SAS should not be required to perform additional validation of location data. Google also supports many of the petitioners for technical amendments to the rules to maximize spectrum availability.

32. SIA Opposition. SIA asks the Commission to reject requests to relax OOBE limits and use an RMS detection methodology for measuring a device's compliance with the Commission's OOBE rules. SIA also opposes: (1) Higher EIRP limits for CBSDs; (2) unlimited antenna height for Category B CBSDs; and (3) any increase in the CBSD or SAS reconfiguration time. Finally, SIA supports elimination of the professional installation option for reporting location accuracy.

33. Qualcomm Opposition. Qualcomm supports CTIA's request to allow the use of an RMS detector to measure OOBE. Qualcomm also supports CTIA's request to relax the requirement limiting OOBE below 3530 MHz and above 3720 MHz to −40 dBm/MHz.

34. T-Mobile Opposition. T-Mobile supports increasing the license term for PALs from three years to ten years with a renewal expectancy. T-Mobile also argues that the Commission should: (1) Make the total number of PALs in a census tract for which applicants have applied available for renewal; (2) increase OOBE and EIRP limits for CBSDs and eliminate conducted power limits; and (3) increase the reconfiguration response time when an incumbent user is detected. Finally, T-Mobile asks the Commission to continue to evaluate whether geo-location capabilities can be built into devices in the future.

35. Verizon Opposition. Verizon states that the Commission should deny SIA's request for stricter OOBE limits and that SIA's concerns about FSS protections are premature. Verizon reiterates its position that allowing CBSDs to operate at higher power limits is crucial to the success of this band.

36. WISPA Opposition. WISPA argues that the Commission should retain the majority of its technical rules, including the maximum power limit, absence of height restrictions for Category B CBSDs, elevation reporting rule and the professional installation requirements. However, WISPA supports requests to relax OOBE limits and to use an RMS detector to measure these levels. WISPA opposes the petitions that request increasing the three-year license term for PALs and opposes permitting a renewal expectancy. However, WISPA supports the requests to award PALs in census tracts even if there is only one application. Finally, WISPA supports retaining the FSS earth station registration requirements.

E. Responses to Second FNPRM

37. The Commission received comment on the three outstanding issues in the
Second FNPRM
described above: (1) Defining use by PALs; (2) creating secondary markets in the 3.5 GHz Band; and (3) FSS protection criteria. These comments, and those received in subsequent rounds, are summarized and referenced in the
Second Order
below.

III. Order on Reconsideration

38. Section 1.429 of the Commission's rules establishes the standards for submission, review, and consideration of petitions for reconsideration (47 CFR 1.429). The eight petitions for reconsideration filed in this proceeding were assessed pursuant to the requirements set forth in section 1.429 (47 CFR 1.429). The arguments made by petitioners are addressed on an issue-by-issue basis below. Except as otherwise set forth below, these petitions do not raise any new issues not considered in the
3.5 GHz R&O,
or where they do, we do not find these arguments persuasive. Through this Order on Reconsideration we reaffirm our commitment to the rules and comprehensive regulatory framework established in the
3.5 GHz R&O.

A. PAL License Terms and Renewability

39. Background. In the
3.5 GHz R&O,
the Commission adopted a three-year non-renewable license term for PALs. This represents an increase from the one-year, non-renewable term that was originally proposed in the FNPRM and on which the Commission sought comment in the
Licensing PN.
After review of the record, the Commission found that three-year, non-renewable license terms strike an appropriate balance between the public interest need for targeted, flexible licensing and the need to provide sufficient certainty for licensees to invest in the 3.5 GHz Band.

40. CTIA asks that the Commission extend PAL license terms to five years and grant an ongoing renewal expectancy, provided that the licensee has deployed services and registered with an SAS. CTIA argues that the existing three-year license term does not provide operators sufficient time or assurance to realize a return on investment. CTIA contends that many challenges associated with network deployment, such as developing and certifying equipment, obtaining appropriate zoning and permitting, and deploying infrastructure, are amplified in the 3.5 GHz Band given the novelty and complexity of higher frequency small cell deployments. Further, CTIA cites IEEE's reluctance to develop a standard to support IEEE 802.11 Wireless Local Area Networks (WLAN) for the 3.5 GHz Band as a signal that the
3.5 GHz R&O
is already affecting investment and innovation. Three parties, AT&T, PCIA, and T-Mobile, support CTIA's position.

41. WISPA filed an opposition to the CTIA Petition stating that the Commission should not revisit the carefully balanced compromise that resulted in the Commission's adoption of a three-year license term. WISPA contends that the approach adopted in the
3.5 GHz R&O
reflects a balance between the views of parties that prefer short-term licenses—including WISPA members—and those that prefer longer license terms. Further, WISPA doubts that large wireless carriers will choose not to deploy in this band. Rather, WISPA notes that, in recent years, the mobile wireless industry has embraced unlicensed deployment models and argues that the Citizens Broadband Radio Service will provide similar investment incentives for the industry.

42. CTIA filed a reply to WISPA's opposition reiterating its arguments. CTIA argues that, while WISPA's members may not need the same level of certainty that mobile operators will require, the Commission should not ignore the novelty and complexity that mobile operators will face when deploying in the 3.5 GHz Band.

43. Discussion. We deny CTIA's request and reaffirm our decision to issue PALs with three-year non-renewable license terms. We agree with WISPA that the
3.5 GHz R&O
already reflects a balance among parties that advocated for short license terms and those that prefer longer terms. We originally proposed a one-year non-renewable license term for PALs but, based on the record, we instead adopted a longer, three-year license term and allowed applicants to apply for two consecutive terms, during the first

applications window, for a total of six years. We continue to believe that “three-year non-renewable license terms—with the ability to aggregate up to six years up-front—strike a balance between some commenters' desire for flexibility with other commenters' need for certainty.” We set forth several arguments in favor of these findings in the
3.5 GHz R&O
and CTIA has not provided any new information that would cause us to alter our analysis. Indeed, the arguments raised by CTIA and supporting parties are similar to those raised by commenters in response to the FNPRM. These arguments were already thoroughly considered by the Commission in the
3.5 GHz R&O.
As such, we continue to believe that three-year, non-renewable license terms strike the proper balance of interests for the 3.5 GHz Band.

44. We also continue to believe that the current rules will effectively incentivize network investment. As we found in the
3.5 GHz R&O,
the rules governing the 3.5 GHz Band work in concert to promote shared access to the band, foster innovation, and ensure that Citizens Broadband Radio Service users are able to efficiently target their use of the 3.5 GHz Band to their specific needs. Non-renewable, short-term licenses are an essential component of this overall framework. They allow operators to obtain PALs when and where Priority Access to the band is needed while permitting periodic, market-based reassignment of these rights in response to changes in local conditions and operator needs. The technical rules and band-wide operability requirement ensure that operators can easily utilize both Priority Access and GAA spectrum in their networks and seamlessly switch between tiers without purchasing additional equipment. In addition, our decision not to impose specific construction requirements for PALs further increases the flexibility and fungibility of these licenses and reduces the barriers to fluid movement between service tiers. These unique features of the Citizens Broadband Radio Service effectively negate the risk of stranded investment for operators and incentivize efficient network deployments.

45. CTIA asserts that deploying a network takes “several years,” and that six years is not a sufficient time period to build a network and obtain the financial return an operator would need to justify making such investments. But CTIA offers no support for its assertion that “several years” must be more than six years to do so or that a PAL is necessary to facilitate network construction. Nor does it address our conclusion, as WISPA notes, that, even for larger carriers, the economics and upgrade cycles for small cell use may resemble those for Wi-Fi deployments rather than traditional macro cell deployments. Furthermore, PAL Licenses Areas are significantly smaller, and therefore require less network deployment, than market areas for other wireless services. Given the differences in the nature and scope of service in this shared band, we continue to believe that three-year, non-renewable PAL terms along with the opportunity to acquire two consecutive three-year licenses during the initial PAL auction reasonably balance the stated interests of different users of this shared band. This approach will promote competition, spur innovation, and encourage rapid network deployment in the 3.5 GHz Band.

B. Assignment of PALs

46. Background. The Communications Act, as amended, requires the Commission to use competitive bidding to assign licenses when “mutually exclusive applications are accepted for any initial license,” subject to specified exemptions not applicable in this band (47 U.S.C. 309(j)(1)-(2), (j)(6)(E)). In the
3.5 GHz R&O,
we found that mutual exclusivity exists when multiple applicants elect to bid on more PALs than exist in a given census tract. We also found that, consistent with previous spectrum auctions, mutual exclusivity will be determined based upon the Commission's acceptance of competing applications. Because of the “generic” nature of PAL frequency assignments, when total PAL applications exceed the PAL bandwidth available in a License Area, PAL applications are mutually exclusive because granting one application would create conflict with another application.

47. Once mutual exclusivity has been established by competing accepted applications seeking to acquire more PALs than are available in a particular geographic area, the PALs in that area will be assigned by competitive bidding, without regard to the number of applicants that ultimately decide to bid or the actual number of PALs for which they place bids. Under this approach, when there are two or more applicants for PALs in a given census tract for a specific auction, we will make available one less PAL than the total number of PALs in that tract for which all applicants have applied, up to a maximum of seven.

48. CTIA, Jon Peha, and Motorola Solutions seek reconsideration of the Commission's method for determining mutual exclusivity for PALs. Federated Wireless, UTC, and WISPA support these petitions. Petitioners assert that the Commission should make PALs available even if only one applicant applies for a PAL in any given census tract and that the number of available PALs should not depend on the number requested by applicants. Petitioners claim that prospective licensees may have need for exclusive access to spectrum in the 3.5 GHz Band and those needs are not dependent on other parties. In addition, Motorola Solutions, Federated Wireless, and UTC contend that the Commission's rule would have negative effects on critical infrastructure industries that may have an interest in exclusive spectrum access. Federated Wireless, UTC, and WISPA argue that the Commission's approach to determining mutual exclusivity is likely to have a disproportionate negative effect on applicants in rural areas, where demand is likely to be sparser than in more densely populated urban and suburban areas.

49. John Peha argues that the Commission has the legal authority to auction PALs even when all applications in a given License Area are received from the same source. WISPA and Motorola solutions suggest that the Commission should set a reasonable licensing or administrative fee if a single applicant applies for a PAL in a given census tract. Federated Wireless and CTIA argue that PALs should be assigned on a non-auctioned basis when there is only one applicant in a given License Area.

50. Discussion. After review of the record, we largely affirm our decision in the
3.5 GHz R&O
and deny the petitions for reconsideration of our determination not to assign PALs in the Citizens Broadband Radio Service in geographic areas for which there is only one applicant, with one limited exception. We modify our original decision to address the limited case of applicants in Rural Areas that may exhibit lower demand than other areas. Specifically, in the absence of mutually exclusive applications, if there is a single applicant for one or more PALs in a License Area within a Rural Area, as defined in section 96.3 (47 CFR 96.3), we will allow for the assignment of one PAL in that License Area. We believe that this narrow exception is appropriate to create an opportunity for operators that provide broadband services to Rural Areas to secure assured exclusive access to spectrum, regardless of competitive demand. As described below, other than this very limited exception, we affirm our decision to issue PALs only through competitive bidding.

51. Given the unique features of this band, we concluded in the
3.5 GHz R&O
that our approach is consistent with the Commission's statutory authority and precedent, and best serves the public interest. Specifically, we found that if there is only a single applicant seeking PALs in a geographic area, and therefore no mutual exclusivity (and hence we have no auction authority), the best way to discharge our statutory mandate to “encourage the larger and more effective use of radio in the public interest” (47 U.S.C. 303(g)) is to provide access to such spectrum via shared GAA use.

52. We continue to believe that the approach adopted in the
3.5 GHz R&O
fulfills our statutory mandate because it establishes an auction process that promotes “efficient and intensive use” of this spectrum, it allows for the “development and rapid deployment of new technologies, products, and services for the benefit of the public, including those residing in rural areas,” and it “recover[s] for the public . . . a portion of the value of the public spectrum resource made available for commercial use” (47 U.S.C. 309(j)(3), 309(j)(4)). This is a market-based approach that targets Priority Access rights where and when there is actual market demand. None of the petitioners presented new evidence to cause us to reconsider the conclusion that this approach drives greater productivity and efficiency in spectrum use and promotes innovation and the development of the next generation of shared spectrum technologies by providing ample opportunities for both GAA and PAL operations.

53. Petitioners indicate that there may be certain types of users or applications that will require PALs for their operations, regardless of whether there are competing users filing applications in a given census tract. The fundamental benefit of a PAL is the right to exclusive use of 10 megahertz of spectrum in a given census tract. In the absence of competition for the spectrum, exclusivity is unnecessary. Further, since there is no difference in the technical rules governing GAA and Priority Access devices and users, the permissible use cases for each tier of service are the same. In the absence of multiple competing applications that exceed the supply of PALs in a geographic area, there should be ample GAA spectrum available for interested parties, thereby obviating the need for exclusive rights. To the extent that petitioners advocate for the assignment of PALs in geographic areas for which there is only one applicant because a particular PAL applicant might anticipate operations that it believes will require the interference protection that is associated with those authorizations, we decline to revise the hybrid framework we adopted in the
3.5 GHz R&O.
In balancing competing public interest objectives, as we often must, that framework was designed to select the best approach to spectrum management based on local supply and demand. Accordingly, where competitive rivalry for spectrum access is low, we determined to allow the GAA tier to provide a low-cost entry point to the band. Where rivalry for spectrum access is high, an auction will resolve mutually exclusive applications for PALs in specific geographic areas. We further adopted finite-term licensing to facilitate evolution of the band and an ever-changing mix of GAA and Priority Access bandwidth over time. As we explained in the
3.5 GHz R&O,
this regulatory adaptability should make the 3.5 GHz Band hospitable to a wide variety of users, deployment models, and business cases, including some solutions to market needs not adequately served by our conventional licensed or unlicensed rules. By adopting rules that provide for widespread GAA use of any spectrum for which we have not received mutually exclusive PAL applications, we ensure that the spectrum will be put to a use for which we have identified a clear public interest need.

54. We reject WISPA's assertion that our approach “substitutes the Commission's business judgment about shared spectrum use over an applicant's business decision that may favor exclusive spectrum use.” Whether or not a business desires exclusivity is independent of whether there is a market-based need for exclusivity caused by rising demand for the spectrum. The Commission's approach does indeed promote shared spectrum use—a fundamental feature of the Citizens Broadband Radio Service since its inception—while providing for prioritized access in areas with heightened demand. In fact, the Commission's approach relies purely on market demand to both trigger an auction and allocate PALs according to that demand, consistent with long-standing Commission practices that efficiently assign spectrum licenses via auction. Any method that would allow PALs to be assigned absent competing applications would not, as WISPA suggests, ensure “a marketplace decision,” but rather one likely to encourage speculation, reduce spectrum availability, and discourage innovation in the band.

55. After review of the record, we do however conclude that it would serve the public interest to allow providers in Rural Areas to have limited PAL access, even in the absence of mutually exclusive applications in that area. Petitioners assert that, in the absence of mutually exclusive PAL applications accepted for a geographic area, the approach adopted in the
3.5 GHz R&O
will have a disproportionate negative effect on rural providers, utilities, and critical infrastructure facilities. Petitioners claim that such users may have a need for the “high quality of service and interference protection that can only be afforded through acquisition of a PAL.” We note that many of these entities—including utilities and rural WISPs—currently utilize the 3650-3700 MHz band (and other bands including 2.4 GHz, 5 GHz, and 900 MHz) on a non-exclusive basis without the option of acquiring priority rights. These entities should be able to provide similar services in the 3.5 GHz Band operating on a GAA basis with the added option of purchasing a PAL if and when demand from more than one party exists in a given geographic area. In addition, as described in this section and section III(A), there is no type of service that is permitted with a PAL that would not be technically allowed or viable under a GAA authorization—the only variable is the ability to exclude others from the use of the spectrum to ensure interference protection, a need which has not been fully supported in the scenario of a single PAL applicant in a geographic area.

56. However, given that demand for PALs may well be lower in less populated areas—particularly early in the Citizens Broadband Radio Service deployment cycle—some Rural Areas may not have multiple applicants for PALs. While we believe that rural service providers can and will provide a variety of robust broadband services in these areas on a GAA basis, we believe that the public interest would be served by ensuring that a PAL is available to a provider in these Rural Areas in the unlikely event that there is a single PAL applicant in a given area. Under this limited exception we will allow for one PAL in a License Area located in a Rural Area in which mutually exclusivity does not exist. If the Commission receives only one application that is acceptable for filing for a License Area located in a Rural Area, the Commission will issue a Public Notice cancelling the auction for this license and establishing a date for the filing of a long-form application, the acceptance of which would trigger the relevant procedures permitting petitions to deny. We believe that granting this limited exception to

our decision not to assign PALs in the Citizens Broadband Radio Service in License Areas for which there is only one applicant is an appropriate balance that will serve the public interest by allowing for the opportunity for a rural service provider to acquire exclusive spectrum use in a Rural Area where such access may facilitate its ability to provide innovative services to customers in more remote locations. However, recognizing the unique nature of this exception, the Commission reserves the right to review and reconsider this approach at a later date. We do not believe there is any reason to change any other aspect of the PAL licensing scheme for Rural Areas or any other use case.

57. We also note that the opportunity to purchase PALs is not a one-time event for this band. Because PALs are licensed for three-year, non-renewable terms, we will periodically open application windows for new PALs that take effect upon expiration of previously assigned PALs. Additionally, if sufficient interest is expressed by prospective PAL users, we will open interim filing windows to accept applications for unassigned PALs,
i.e.,
PALs that could be made available for auction, before the expiration of an ongoing three-year PAL term. Therefore, as the band develops, our approach provides mechanisms to make PALs available in response to changing market conditions.

58. While we could issue PALs on a non-auctioned basis—as suggested by Federated Wireless and CTIA—we conclude that doing so in this band would not result in as efficient an assignment of the spectrum as licensing the spectrum for shared GAA use, except for the limited exception described above. As part of its proposal that we assign PALs in a license area with only one applicant, Motorola Solutions asserted that the “interested party would be expected to pay a reasonable licensing/administrative fee for such PAL use, and may be expected to pay a reasonable fee to a SAS database provider for interference protection.” Neither Motorola Solutions nor WISPA put forward any theory as to how we would assess this fee under our statutory authority, or how it could replicate a mechanism reflecting the spectrum's fair market value. We believe the record on this issue is insufficient to support Motorola's proposal. We continue to believe the adopted rules are the best way to “encourage the larger and more effective use of radio in the public interest” and nothing in the record supports reconsideration of this determination.

C. SAS and CBSD Response Time

59. Background. In the
3.5 GHz R&O,
the Commission adopted section 96.15(a)(4) (47 CFR 96.15(a)(4)), which requires that, for CBSDs operating in the 3550-3650 MHz band, “[w]ithin 60 seconds after the ESC communicates that it has detected a signal from a federal system in a given area, the SAS must either confirm suspension of the CBSD's operation or its relocation to another unoccupied frequency, if available.” The Commission adopted identical requirements for CBSDs operating in the 3650-3700 MHz band. The Commission also requires that “A CBSD must receive and comply with any incoming commands from its associated SAS about any changes to power limits and frequency assignments. A CBSD must cease transmission, move to another frequency range, or change its power level within 60 seconds as instructed by an SAS.”

60. Motorola Solutions, Nokia Solutions, and WinnForum petition the Commission to increase the first of these two intervals (SAS reconfiguration response time in section 96.15) from 60 seconds to 600 seconds. WinnForum contends that this increase is necessary to ensure a smooth handover of CBSDs to new frequencies or bands. They emphasize the complexity of optimizing these transitions among a number of different SASs and network operators. WinnForum also argues that some critical infrastructure and emergency use cases may need a longer time to effect a seamless transition from the affected frequencies. However, they acknowledge that most CBSDs could probably be cleared after only 300 seconds. Nokia Solutions also suggests that the reconfiguration time be increased to 600 seconds and indicates that, even in a best case scenario, a complex network cannot be suspended or relocated within 60 seconds. Google and WISPA also support WinnForum's Petition.

61. Google notes that there is a tension between the SAS reconfiguration rule and the second of these two intervals (the reconfiguration requirement in section 96.39 that requires CBSDs to cease operations or move to a non-interfering frequency within 60 seconds of receiving instructions from the SAS) (47 CFR 96.39). According to Google, in practice, the combination of these two rules would be to effectively require CBSDs to take action in less than 60 seconds. Google contends that, to resolve this tension, the Commission should increase the interval for SASs to respond to ESC directions but retain the 60-second timeframe for CBSDs to respond to SAS commands.

62. SIA argues that the 60-second response time in section 96.39 (47 CFR 96.39) for CBSDs to move or discontinue operations is too long and asks that the Commission reduce that timeframe. SIA argues that even a one-minute delay could cause significant damage to incumbent satellite systems. SIA asserts that, since the CBSD response time is in addition to any additional time needed for the SAS to process information from the CBSD and communicate with the device, interference could continue for longer than 60 seconds in practice. SIA asserts that the petitions for increases in SAS response time only reinforce their concerns about how quickly harmful interference into incumbent FSS earth stations can be addressed. Google asserts that SIA misunderstands the different types of commands addressed by the Commission's rules and the arguments made by petitioners. Google contends that nothing in petitioners' requests to increase the SAS reconfiguration timeframe in section 96.15 (47 CFR 96.15) casts doubt on the ability of CBSDs to respond to instructions from an SAS within the 60-second window established by section 96.39 (47 CFR 96.39).

63. Discussion. After review of the record, we believe that the SAS reconfiguration time should be increased. Petitioners contend that 60 seconds is an insufficient window for SASs and licensees to effectively reconfigure their networks in response to reported interference. Indeed, Nokia Solutions argues that it may be impossible to effect such changes even under ideal circumstances. These problems are likely to be more acute with networks consisting of a large number of CBSDs. While we take no position on the veracity of these claims, from the evidence presented, it appears that increasing the SAS reconfiguration timeframe will help to promote robust development and deployment of broadband networks in the 3.5 GHz Band.

64. However, given the importance of the incumbent services present in the band, we do not believe that the 600-second SAS reconfiguration timeframe suggested by commenters is appropriate. Federal Incumbent Users must be assured that their mission critical operations will be protected from harmful interference and that any interference reported will be addressed in a timely manner. Therefore, we amend section 96.15(a)(4) and (b)(4) of the rules (47 CFR 96.15(a)(4) and (b)(4)) and extend the SAS reconfiguration

timeframe to 300 seconds. Both Nokia Solutions and WinnForum indicated that, while not ideal, a 300-second reconfiguration window would be adequate for a majority of CBSDs to effectively cease transmitting or transition to a non-interfering frequency. They do not provide a basis for why as much as 600-seconds is needed, even for a large network. We also amend sections 96.15(a)(4) and (b)(4) (47 CFR 96.15(a)(4) and (b)(4)) to clarify that the 300-second reconfiguration window applies to notifications regarding federal use from the ESC or any other source, including federal Incumbent Users themselves. This modification is necessary to ensure that federal Incumbent Users are protected from harmful interference in all circumstances. However, the 300-second timeframe will not necessarily apply if the President of the United States (or another designated Federal Government entity) issues instructions to discontinue use of CBSDs pursuant to section 706 of the Communications Act of 1934 (47 U.S.C. 157), as amended (War Powers of President) (47 U.S.C. 606). In such cases, SAS Administrators must instruct CBSDs to cease operations as soon as technically possible (but no more than 300-seconds). We also note that at this time there is no indication of how the increase in the SAS reconfiguration time will impact federal radar systems. If it is demonstrated there is an operational impact to the federal radar systems, the Commission will review the SAS reconfiguration timeframe and will take appropriate steps to address the operational impact to federal radar systems.

65. While some commenters claim that even this extended reconfiguration window may cause service interruptions in some cases, we believe that 300 seconds will ordinarily provide operators with sufficient time to smoothly discontinue transmissions or move to non-interfering frequencies. Moreover, given the critical importance of the federal operations in the band, we must ensure that CBSDs are shut down as quickly as possible after the presence of federal operations is reported by an ESC or actual interference is reported by a federal user. This change also resolves the tension between sections 96.15 and 96.39 (47 CFR 96.15(a)(4), 96.39(c)(2)) pointed out by Google. Therefore, we find that a 300-second response timeframe strikes the appropriate balance between protecting incumbent operations and facilitating commercial deployments in the band. In addition, given the technical capabilities of SASs and CBSDs, we believe that it is both reasonable and technically feasible to require Citizens Broadband Radio Service users to comply with this modified response timeframe.

66. We refuse SIA's request to shorten the 60-second CBSD reconfiguration timeframe in section 96.39 of the rules. As Google correctly notes, SIA's arguments on this point were considered by the Commission when the rule was adopted. SIA does not raise any substantive new arguments that would compel us to override our prior decision. To the extent that incumbent FSS earth station licensees may have specific, time-limited requests for protection during certain periods, we encourage FSS licensees to work with SAS Administrators to address these concerns. As detailed in section III(H)(2) and section 96.17(f) (47 CFR 96.17(f)), SAS Administrators must develop procedures to receive and respond to such requests. Accordingly, in light of this requirement, we continue to believe that the 60-second CBSD reconfiguration timeframe in section 96.39 (47 CFR 96.39) is sufficient to ensure that federal and non-federal users are protected.

D. CBSD Power Limits

67. Background. In the
3.5 GHz R&O,
the Commission found that “it is vitally important to establish flexible, yet simple, rules that would allow for a wide variety of innovative services to be deployed in the 3.5 GHz Band.” To advance this goal, the Commission defined two categories of CBSDs—Category A and Category B—with parameters appropriate for different use cases. Category A and Category B CBSDs are differentiated primarily by their maximum permissible power and the rules governing their deployment. In addition, Category B CBSDs may only be authorized in the 3550-3650 MHz portion of the band after an ESC is approved and operational. GAA users and Priority Access Licensees may operate CBSDs in both categories and must operate in accordance with instructions from an SAS which, for interference prevention purposes, may authorize an operational power level below the maximum allowable power level (47 CFR 96.41, 96.43, 96.45).

68. Category A CBSDs are limited to a maximum conducted transmit power of 24 dBm and a maximum EIRP of 30 dBm in 10 megahertz and may be deployed either indoors or outdoors (with antennas for outdoor deployments not exceeding 6 meters height above average terrain) (47 CFR 96.41(b), 96.43(a)). These parameters are consistent with the baseline small cell use case proposed in the FNPRM and the phased federal-commercial sharing plan proposed by NTIA and adopted in the
3.5 GHz R&O.

69. Category B CBSDs, which may only be used outdoors, are permitted to operate at higher power than Category A, providing greater flexibility and ensuring ongoing compatibility with existing 3650-3700 MHz band operations (47 CFR 96.41(b), 96.45). In non-rural areas, the conducted power limit is the same as Category A (24 dBm/10 MHz), but the EIRP limit is 40 dBm/10 MHz. In rural areas, the conducted power limit is increased to 30 dBm/10 MHz and EIRP to 47 dBm/10 MHz (47 CFR 96.41(b)). The EIRP limit was set to encourage the use of higher gain antennas and directional transmission in urban areas to facilitate co-existence of PALs and GAAs in spatially tight spectrum sharing environment. The higher rural power limits reflect challenges for deploying wireless coverage in rural areas as well as decreased contention for spectrum resources due to lower population density in those areas.

70. CTIA, Motorola Solutions, Nokia Solutions, Verizon, and WinnForum petitioned the Commission to increase CBSD power limits. AT&T and Federated Wireless supported these arguments. Petitioners assert that the maximum power levels for Category A devices should be raised to 36 dBm EIRP. Petitioners contend that the Category A power levels adopted by the Commission are insufficient to provide significant indoor coverage. Nokia Solutions and WinnForum also contend that a 36 dBm maximum EIRP would be consistent with levels the Commission has approved for unlicensed devices.

71. Petitioners also argue that the maximum permissible EIRP for Category B CBSDs should be raised to 49 dBm for non-rural deployments and to 56 dBm for rural deployments. WinnForum contends that the proposed increases would bring the Commission's rules in line with the power levels of existing urban pico-cells. Verizon contends that the maximum EIRP that the Commission adopted for Category B CBSDs is well below the power levels of the small cells that are used in current licensed deployments. Verizon also argues that the existing rules would significantly limit the coverage that each cell could achieve, driving up network costs. Federated Wireless agrees and adds that “Even at the increased EIRP limit, CBSDs will still operate at power levels no greater than those employed in typical small cell deployments.”

72. Many petitioners also assert that the Commission should increase the flexibility for operators to deploy lower

gain antennas by relaxing the limitations on conducted power for Category A and B CBSDs. For example, Nokia Solutions and Verizon argue that the limitations on conducted power should be removed entirely to provide additional flexibility network operators in the 3.5 GHz Band. WinnForum proposes that the allowed conducted power be scaled up 1 dB for each 1 dB lost in antenna gain, up to the maximum of 40 dBm conducted power for Category B CBSDs. WinnForum argues that this approach would not preclude the use of omni-directional antennas while still maintaining adequate coverage areas for outdoor deployments.

73. SIA opposes any increase in maximum EIRP for Category A or Category B CBSDs and, in fact, argues that they should be reduced to levels stated in the
FNPRM.
SIA contends that higher EIRP limits will increase the risks of interference with incumbent FSS earth stations and significantly increase the size of required separation distances around these stations. They also see risks associated with not limiting the antenna height for Category B CBSDs due to interference to incumbent in-band and out-of-band FSS receivers.

74. WISPA argues that the Commission should not change the maximum allowable EIRP for Category B CBSDs. In WISPA's view, the Commission's rules strike the proper balance between various interests and encourage operators of outdoor networks to deploy more efficient, high-gain, sectorized antennas. Federated Wireless disagrees with WISPA and contends that increased EIRP and flexibility is essential to promote innovation and enable more efficient spectrum use.

75. Discussion. After review of the record, we agree with commenters that contend that additional flexibility for non-rural outdoor CBSDs would promote deployment in the band and, accordingly, we increase the maximum allowable EIRP for non-rural Category B CBSDs from 40 dBm/10 MHz to 47 dBm/10 MHz, making the power levels allowed for both non-rural and rural deployments the same. Category B CBSDs will continue to be authorized for use in the 3550-3650 MHz band only after an ESC is approved and commercially deployed consistent with sections 96.15 and 96.67 (47 CFR 96.15, 96.67). We also eliminate the conducted power limits for all CBSDs. However, we also conclude that it would not be in the public interest to increase the maximum allowable EIRP for Category A CBSDs and rural Category B CBSDs beyond the levels established in the
3.5 GHz R&O.
Combined, these changes will provide increased flexibility to all network operators without increasing the potential for interference in the 3.5 GHz Band.

76. As we stated in the
3.5 GHz R&O,
we are cognizant that the determination of power limits for all categories of CBSD must balance the consideration of several different public interest objectives. On the one hand, higher limits may provide more technical and operational flexibility for users of the band to increase coverage with fewer CBSDs, potentially reducing deployment costs. On the other hand, lower power limits may lead to greater spatial reuse of the band, reduced coexistence challenges, and increased aggregate network capacity. Our determinations herein strive to balance these considerations to create a flexible regime suitable for a wide variety of use cases.

77. With regard to Category B CBSDs, we agree with commenters that higher maximum EIRP may help promote more flexible use and reduce deployment costs in non-rural areas while not significantly increasing coexistence issues. Specifically, we increase the maximum EIRP for Category B CBSDs in non-rural areas to 47 dBm/10 MHz to match the maximum EIRP permitted in rural areas. Petitioners generally argue that higher power is needed to facilitate network deployment and decrease costs. Although we remain concerned about more substantial power increases in more congested areas, we agree that allowing non-rural CBSDs to match the EIRP of rural CBSDs is consistent with the Commission's goals for the Citizens Broadband Radio Service and is a modest increase that will not adversely affect the interference environment in the 3.5 GHz Band.

78. However, we do not agree that the maximum EIRP for Category B CBSDs should be increased to 49 dBm/10 MHz in non-rural areas and 56 dBm/10 MHz in rural areas as requested by several petitioners. While we see the merit in increasing the maximum power available to network operators using Category B CBSDs in non-rural areas, we believe that an increase to 47 dBm/10MHz to match the level permitted for rural CBSDs will adequately address the concerns raised by Petitioners without negative effects on the interference environment in the band. This change represents a significant increase in power for non-rural applications with a corresponding potential for more coverage area for each CBSD. This change will also simplify the rules by removing the distinction between rural and non-rural power levels, allowing for uniform development and deployment of Category B CBSDs. We also note that Category B CBSDs will continue to be authorized for use in the 3550-3650 MHz band only after an ESC is approved and commercially deployed consistent with sections 96.15 and 96.67 (47 CFR 96.15, 96.67).

79. We continue to believe that the power limit that we adopted for Category A CBSDs in the
3.5 GHz R&O
is appropriate for the baseline—primarily indoor or at street level—small cell use case in the band. Moreover, the Exclusion Zones protecting federal radar systems that were studied by NTIA and adopted in the
3.5 GHz R&O
are based on a maximum EIRP of 30 dBm/10 MHz. Any change to the maximum EIRP for Category A CBSDs would require the Exclusion Zones to be reconsidered and expanded, preventing deployment in large portions of the country prior to the development and approval of an ESC.

80. While we acknowledge that some petitioners would prefer that we increase the Category A power levels to allow higher power levels indoors, we believe that the rules appropriately balance the need for operational flexibility with the need to promote efficient spatial and spectral reuse of the band. Transmitting at higher power levels indoors and low outdoor elevations—especially in high traffic areas with multiple PALs and GAAs operating in the same or nearby locations—would likely present significant coexistence challenges. Higher power levels in dense indoor deployments would also increase the likelihood of interference from operators assigned to adjacent channels due to receiver blocking effects. Thus, given the interference risks associated with higher power levels, the delays in deployment of this new service that would result from revisiting the size of the Exclusion Zones prior to implementing an ESC capability, and the disruption to the balance between PAL and GAA use struck in the
3.5 GHz R&O,
we conclude that the maximum EIRP for Category A CBSDs should remain capped at 30 dBm/10 MHz.

81. We are also cognizant of the concerns raised by SIA regarding the need for greater protections for FSS earth stations in the presence of higher power CBSDs but note that the FSS interference protection criteria described in section IV(C)(1) addresses these concerns. We emphasize that the increase in allowable EIRP for non-rural Category B CBSDs is an increase in the maximum allowable EIRP and should not be construed as a guaranteed power level for CBSD deployments, whether

they are operated on a GAA or Priority Access basis. We note that CBSDs must still comply with the Commission's rules to prevent interference to Incumbent Users, including the requirements to operate only at power levels and in locations authorized by the SAS (47 CFR 96.39(c)). Indeed, given that the potential for co-channel and adjacent channel interference may increase at higher power levels, the SAS's responsibility to authorize lower maximum operational power limits, when and where needed to meet the interference protection requirements as defined in Commission's rules, will be even more important in light of the increased maximum power levels authorized herein.

82. Finally, we find that removing maximum conducted power limits for all CBSDs will provide operators with additional flexibility for network deployments and encourage investment in the band. Several petitioners, including WinnForum, Verizon, and Federated Wireless, contend that the Commission's rules requiring Category B CBSDs to use sectorized, highly directional antennas in urban areas would lead to inefficient deployments. Notably, Federated Wireless contends that, since most CBSDs will be deployed below the clutter in urban areas, sectorized antennas would be unable to provide the coverage needed for urban deployment. In addition, since the Exclusion Zones and other protection contours in the band are based on EIRP, removing the conducted power limits should not increase the required protection areas around incumbent sites. Therefore, we agree with petitioners that, on balance, increased flexibility will serve the public interest and promote investment in the 3.5 GHz Band. We note that this has no impact on our OOBE requirements, which continue to be expressed in terms of conducted power. That is, although the rule changes described in this section will allow higher total conducted power, they do not allow higher OOBE power.

83. In making this change to remove maximum conducted power limits for all CBSDs we also recognize that we must limit the peak to average power ratio (PAPR) of signals in the band so that excessive peak power levels do not cause transient interference into other systems. Many commenters have expressed interest in deploying LTE equipment in the 3.5 GHz Band. We note that such signals use OFDM based modulation, which can have a large PAPR. NTIA recently published emission spectrum measurements for a 3.5 GHz LTE hot spot device shows that the peak to average ratio of such devices may range as high as 12-13 dB. Thus, based on these measurements and consistent with the Commission's rules in other licensed mobile broadband services, we are limiting CBSD PAPR to no more than 13 dB (47 CFR 24.232(d) and 27.50(a)(1)(B) and (d)(5)).

84. Finally, SIA argues that unlimited antenna heights for Category B CBSDs will necessitate larger protection areas for FSS earth stations. SIA does not propose a specific remedy or alternate rule governing antenna heights. We note that Category B CBSDs are required to report antenna height as part of their CBSD registration under section 96.45(d) (47 CFR 96.45(d)) and SASs are required to take such antenna height (along with maximum power, location, antenna configuration, and other registered information) into consideration when calculating potential interference effects and protection distances (47 CFR 96.17(d), 96.45(d), 96.53, 96.55). Indeed, the protection criteria set forth in the rules may require an effective limit on Category B antenna elevation in some cases. We continue to believe that the SAS can utilize information reported by CBSDs to effectively coordinate operations in the 3.5 GHz Band and see no reason to impose restrictions on the height of Category B CBSD antennas at this time.

E. OOBE and Adjacent Channel Emissions Limits

1. OOBE and Adjacent Channel Emissions

85. Background. In the
3.5 GHz R&O,
we adopted emissions and interference limits that will further the Commission's goals and promote effective coexistence of different users in the band. Specifically, we adopted the following conducted OOBE limits for devices in the Citizens Broadband Radio Service:

• −13 dBm/MHz from 0 to 10 megahertz from the SAS assigned channel edge

• −25 dBm/MHz beyond 10 megahertz from the SAS assigned channel edge down to 3530 MHz and up to 3720 MHz

• −40 dBm/MHz below 3530 MHz and above 3720 MHz

86. CTIA, Nokia Solutions, and SIA petition the Commission to change its OOBE limits. CTIA contends that the −40 dBm/MHz OOBE limit simply is too restrictive and is not necessary to protect operations in the adjacent band below 3530 MHz and above 3720 MHz. CTIA also asserts that, if the Commission determines that the −40 dBm/MHz limit is necessary to protect adjacent operations, the Commission should increase the transition gap to 40 megahertz to allow operators using 20 megahertz LTE channels to operate at higher power. Qualcomm supports CTIA's comments and asserts that the FCC should not implement tighter OOBE limits at the 3700 MHz band edge for certain classes of devices to protect C-band FSS earth stations. According to Qualcomm, stringent OOBE limits will challenge equipment designs and likely force mobile devices to use significantly less power and/or operate well inside the 3.5 GHz Band edges to comply. Google, T-Mobile, and WISPA also support relaxation of the OOBE limits.

87. Nokia Solutions recommends that the Commission define OOBE limits that comply with 3GPP specifications and would allow the use of Bands 42 and 43 in the United States. According to Nokia only the requirement of −25 dBm/MHz beyond 10 MHz from the assigned channel edge down to 3530 MHz and up to 3720 MHz complies with the 3GPP specification.

88. CTIA also argues that the Commission should adopt a limit of −13 dBm/MHz from 0-20 megahertz outside the assigned channel edge and a limit of −25 dBm/MHz for frequencies more than 20 megahertz outside each assigned channel edge. Qualcomm agrees and contends that the emissions limits that apply outside of the channel of operation were designed around supporting 10 MHz-wide LTE channels, and thus would force 20 MHz LTE and 40 MHz LTE operations to use substantially lower transmit power than the level 10 MHz LTE operations are permitted to use. According to Qualcomm, such reductions will create coverage challenges and limit the band's ability to support wider bandwidth LTE operations. Similarly, T-Mobile argues that 20 megahertz LTE channels would have to be at least 20 megahertz from the channel-edge to meet the −25 dBm/MHz limit without significantly reducing power levels. The reduced power necessary to meet the −25 dBm/MHz limit would in turn reduce coverage of those 20 megahertz channels and would depress operators' desire to deploy those channels.

89. On the other hand, SIA argues that more restrictive OOBE limits are needed to effectively protect C-Band FSS earth stations from CBSD transmissions. SIA also asserts that the OOBE limits adopted by the Commission were implemented without the required legal notice. According to SIA, under the Commission's current OOBE rules, separation distances between CBSDs and FSS earth stations could be more

than 15 km. GCI also argues that the Commission should implement more stringent OOBE limits at the upper edge of the 3.5 GHz Band. According to GCI, at a minimum, a −40 dBm/MHz limit should be implemented at the band edge to protect C-Band FSS earth station receivers.

90. Some parties support the Commission's current OOBE limits. Notably, Verizon argues that the current OOBE limits are sound and oppose further OOBE restrictions. Federated Wireless also contends that the Commission need not reconsider the OOBE issue now.

91. Discussion. After review of the diverse record on this issue, we deny the petitions for reconsideration that requested changes to the OOBE limits that the Commission adopted in the
3.5 GHz R&O.
We continue to believe that the existing OOBE rules properly balance the need to protect operations in adjacent bands—and in adjacent channels within the 3.5 GHz Band—with the need to create an environment that will promote robust deployment of broadband systems in the band.

92. We also believe that, while the OOBE limits are more restrictive than those in other bands, they are wholly consistent with the capabilities of the equipment and services likely to be deployed in the 3.5 GHz Band. For emissions below 3530 MHz and above 3720 MHz, NTIA measurements show that the OOBE of commercial products that operate within the 3.5 GHz Band can be lower than −40 dBm/MHz at offsets higher than 20 megahertz. Thus, according to NTIA research, the approach adopted by the Commission appears to be practically realizable with existing state-of-the-art products at little or no added cost and will provide additional protection for incumbent systems while allowing for more extensive deployment of CBSDs in the 3.5 GHz Band.

93. We disagree with CTIA and Qualcomm's argument that the Commission's OOBE limits should be changed since they would force operators using 20 megahertz channels to reduce power to comply with the rules. As we noted in the
3.5 GHz R&O,
ten megahertz channels provide a flexible, scalable, and practically deployable bandwidth for high data rate technologies, permitting multiple Priority Access Licensees to operate in the same geographic area. While Citizens Broadband Radio Service users are permitted to aggregate PAL channels or operate across wider bandwidths—consistent with section 96.31 (47 CFR 96.31)—the technical rules required for effective coexistence between and among different users of the band do not change, regardless of the how much bandwidth is in use. We also note that power reduction may not be necessary if Citizens Broadband Radio Service users utilize robust filters or other alternative methods to address our OOBE limits. While the flexibility to aggregate spectrum is a key element of the Commission's licensing regime, reducing OOBE limits solely to accommodate wider bandwidths would not further the principles of shared access that are at the heart of this proceeding.

94. Moreover, petitioners do not provide convincing evidence or technical analysis to support their claims regarding power reduction nor do they address the potential effects such changes could have on adjacent channel operations. We also expect to see more spectrally efficient commercial products enter the marketplace in the near future that will meet or exceed our requirements. The current rules support the development of such new and innovative technologies while ensuring a proper balance between the current and future users of the band.

95. We also reject SIA's arguments that the strictest OOBE limits adopted by the Commission (−40 dBm/MHz) should have been set beginning at 3680 MHz, which is 20 megahertz below the lower edge of the adjacent C-Band, rather than at 3720 MHz. SIA argues that failing to do so will lead to impermissible interference into C-Band FSS earth stations. As we stated in the
3.5 GHz R&O,
the −13 dBm/MHz OOBE limit at the band edge is consistent with Commission precedent both in this band and in other licensed spectrum bands. In addition, the transition gap that requires OOBE to drop to −25 dBm/MHz after a 10 megahertz offset and −40 dBm/MHz above 3720 megahertz is significantly more stringent than limits in other bands or the limits that the Commission previously adopted for the 3650-3700 MHz Wireless Broadband Radio Service. The Commission adopted these more stringent limits in recognition of the need to provide additional protection for important operations in the C-Band. Indeed, as detailed above, several petitioners continue to object to these limits as too stringent for certain wireless broadband uses in the Citizens Broadband Radio Service. After review of the record, we remain convinced that the OOBE limits adopted in the
3.5 GHz R&O
strike the appropriate balance between the need to facilitate innovation and investment in the 3.5 GHz Band and the need to protect licensed C-Band FSS earth stations from interference.

96. However, while we maintain the existing OOBE limits, we do acknowledge SIA's concerns regarding potential interference into C-Band receivers used for critical telemetry, tracking, and control (TT&C) operations at the band edge. Therefore, as detailed in section IV(C)(2), we adopt rules to provide additional protection for these facilities. We also adopt new rules to facilitate coordination between Citizens Broadband Radio Service users and licensed C-Band FSS earth stations to address any interference issues that may arise.

97. Finally, we reject SIA's assertion that the Commission did not provide proper notice prior to adopting the current OOBE rules in the
3.5 GHz R&O.
As SIA itself notes, in the FNPRM, the Commission: (1) Proposed an OOBE limit of −13 dBm/MHz at the band edge and −40 dBm/MHz and 30 megahertz above and below the proposed band edges; (2) sought comment on both OOBE limits and the size of the transition gap; and (3) sought comment on extending the Citizens Broadband Radio Service to 3700 MHz. Even prior to that time, the
Licensing PN
sought comment on “[w]hat provisions would need to be made for incumbent operators” if the band were so extended. And in the
3.5 GHz R&O
itself, the Commission determined to seek further comment on “steps we can take over and above those we've already taken to preempt and mitigate the potential for interference” to incumbent C-Band licensees, referring specifically to “our baseline emission performance rule.”

98. As SIA correctly states, “a final rule need not be an exact replica of the rule proposed in the Notice, the final rule must be a `logical outgrowth' of the rule proposed.” In this case, the Commission had sought comment on the need for interference protections relating to extension of the band edge from 3650 MHz to 3700 MHz. The OOBE limits later proposed in the FNPRM were clearly intended to apply to the upper and lower bounds of the Citizens Broadband Radio Service and the Commission made it clear that those bounds could extend to 3700 MHz. Indeed, the Commission originally sought comment on extending the Citizens Broadband Radio Service to 3700 MHz in the original NPRM released in December of 2012. Thus, the extension of the 3.5 GHz Band—and with it the OOBE rules applicable at and beyond the band edge—was wholly foreseeable and a clear logical outgrowth of the Commission's proposals. In addition, the
3.5 GHz R&O
itself provided parties with yet a further opportunity to comment on the

approaches that the Commission could utilize to protect C-Band FSS earth stations.

2. Emission Power Measurements and Testing Methodology

99. Background. In the
3.5 GHz R&O,
we adopted a rule that requires that emission power measurements be performed with a peak detector in maximum hold. CTIA objects to this testing methodology and asks the Commission to adopt a different measurement technique. Qualcomm, T-Mobile, WinnForum, and WISPA support CTIA's request. CTIA contends that the use of an RMS detector to measure emissions would be wholly consistent with the Commission's rules governing most other commercial licensed and unlicensed services. In addition, CTIA states that the peak to average ratio for emissions from LTE signals can easily exceed 10 dB and compelling Citizens Broadband Radio Service users to operate with that much less power would effectively cripple the band's ability to support mobile broadband operations. WISPA agrees and adds that, not only would measuring at peak power require mobile operations to operate at significantly less power, but this would similarly impinge upon the ability of fixed providers to operate at the maximum authorized power.

100. In addition, WinnForum argues that 10+ dB signal strengths over average captured by the current rule would exist for less than 0.01% of the time for any one signal. WinnForum also contends that requiring devices to be tested using a peak detector at maximum hold effectively requires that devices be certified at the maximum possible signal strength at any given time and is a very poor representation of actual interference impact. According to WinnForum, the part 96 emission limits are already stringent, and become simply unattainable when adding over 10dB penalty through the peak detector/max hold requirement. WinnForum also claims that the effects would likely be similar for other wideband systems (Wi-Fi, WiMAX, etc.).

101. SIA disagrees with WinnForum and argues that the Commission should retain the peak measurement test for OOBE. SIA states that ignoring peak emission levels in favor of reliance on average measurements would undermine the prophylactic objectives of the OOBE limits. SIA contends that, by CTIA's own admission, the change would allow power increases of 10 dB or more. According to SIA, because peak emissions can have significant interference effects, the Commission must continue to require use of a peak detector to determine OOBE limit compliance.

102. Google supports WinnForum's filing and argues that SIA's claims should be rejected. Google asserts that all signals, including LTE, Wi-Fi, WiMAX, and even Gaussian thermal noise will have statistical variations in the instantaneous amplitude of the waveform and argues that, for this reason neither cellular, AWS, PCS, or 700 MHz emission are measured using peak hold. Google also asserts that, since the PAPR and signal statistics of LTE and Gaussian thermal noise are similar, the measurement of their interference potential should be treated in the same way. Accordingly, Google argues that if SIA insists on measuring CBSD emissions using peak values, the system noise of FSS receivers should be characterized in the same manner.

103. Discussion. After careful review of the record, we conclude that emission power measurements may be performed using either RMS-detection or peak-detection. We agree with petitioners that requiring the use of a peak detector operating at maximum hold to test emission limits does not serve the public interest. As WinnForum argues, requiring the use of peak measurements may effectively prevent the development and deployment of equipment in the band. Moreover, the decision to allow the use of RMS measurements is consistent with existing Commission rules for several other licensed services in the past, including the AWS bands 47 CFR 27.50(b)(11), (c)(11), (d)(6), (h)(4)(i), 24.132(d)-(f). In other services, the Commission has adopted the emission power measurement by giving the option of detecting peak value or average value 47 CFR 27.53(a)(7), (h)(3)(iii). This decision will provide the measurement lab with a great deal of flexibility to select the appropriate detection type during the certification process.

104. RF power measurement is a function of the receiver bandwidth and detection method whether the signal is detected using a peak or average technique. LTE signals are using OFDM based modulation in downlink which are known to have large PAPRs which may be beyond the 10 dB margin. Google also points out that the PAPRs and signal statistics of LTE and Gaussian thermal noise are generally similar, and thermal noise is typically evaluated using mean measurements. Recent NTIA lab measurements of emission spectrum for a commercial LTE hot spot device operating in the 3.5 GHz Band has shown PAPRs of up to about 12-13 dB. The PAPR for an LTE signal is a random value that fluctuates over a wide range and depends on modulation type and number of sub-carriers used.

105. We reject SIA's argument that retaining the peak detector at maximum hold measurement requirement is necessary to prevent harmful interference into C-Band FSS earth stations. SIA contends that this measurement approach is necessary because “peak emissions may have significant interference effects.” However, the issue is not what is commonly referred to as “peak power” but rather extremely short duration transient signals that typically have little energy and, therefore, generally do not reflect interference potential. In effect, requiring devices to be tested using a peak detector at max hold requires devices to be certified at their “worst case” configuration which would present an unrealistic view of the actual interference potential of any given device. This approach is inconsistent with our oft stated rejection of worst case approaches to measurements and interference protection analysis. Moreover, as Google notes, SIA's assertion that CBSD emission levels should be measured using a peak detector, while their own system noise levels are exempt from such a requirement, is logically inconsistent and mathematically unsound.

106. In addition, WinnForum argues that, since incumbent protections in the 3.5 GHz Band will be calculated using aggregate interference from multiple CBSDs, certifying CBSDs using a peak detector at max hold will compound the effects of these worst case certifications, yielding an unrealistic picture of the RF environment. On the other hand, calculating aggregate interference effects based on average measurements will present a more realistic picture of the actual RF environment for the purpose of determining protection of incumbent systems, including FSS earth stations. We agree with CTIA, Google, and WinnForum that maintaining the peak detector at maximum hold requirement would be unnecessary, particularly in light of the cap on peak-to-average emissions we adopt below. Maintaining this approach would also be inconsistent with the Commission's goals for the Citizens Broadband Radio Service and would not promote spectral efficiency and co-existence among various users in the 3.5 GHz Band and adjacent bands.

107. It is also typically easier to measure emissions using the peak detected signal as part of standard

measurements. Accordingly, under our revised rules, if the device passes the peak detection requirements, no further RMS-detection is needed to meet the OOBE conditions; otherwise, the RMS-detection method can be applied. However, in order to circumvent any effect of peak power spikes, as indicated in the CBSD power requirement section, we will also require that the PAPR of the transmitter output power not exceed 13 dB consistent with the Commission's previous rules in other licensed mobile broadband services 47 CFR 24.32(d), 27.50(a)(1)(B) and (d)(5). NTIA lab measurements on LTE hot spot devices also support our finding that a 13 dB margin is reasonable for industry to achieve.

108. We believe the combination of changing the requirement to include the use of RMS detection for emission measurement, along with setting the PAPR limitation, will diminish the potential for interference between and among Citizens Broadband Radio Service users and Incumbent Users while promoting efficient use of the band. We disagree with SIA's assertions and note that RMS measurement is commonly used by the Commission and, in fact, is commonly used in other bands. Indeed, allowing such flexible measurement techniques here will help promote the next generation of shared spectrum technologies, and will drive greater productivity and efficiency in spectrum usage.

F. Device Geo-Location

1. Location Accuracy and Alternative Measurement Approaches

109. Background. In the
3.5 GHz R&O
we required that all CBSDs must accurately report the location coordinates (referenced to the North American Datum of 1983, NAD83) of each of their antennas to within ±50 meters (horizontal) and ±3 meters (vertical) (47 CFR 96.39(a)). We found that, for the SAS to accurately predict and evaluate interference and channel availability, it must receive and store accurate location information for all CBSDs.

110. Motorola Solutions, Nokia Solutions, and WinnForum filed petitions for reconsideration requesting that Commission relax the existing accuracy requirements and suggest, alternatively, that the Commission allow the SAS to play a role in estimating CBSD location. Google and Federated Wireless also support alternative approaches to ascertaining the location of CBSDs. Specifically, Federated Wireless explains that there are a variety of methods the SAS could use to verify location, such as coordinating with downstream infrastructure or reference to its power levels and other measurements. Google suggests that even if devices cannot meet the specific requirements established by the
3.5 GHz R&O,
the Commission should permit an SAS to calculate spectrum availability based on the geolocation reported by the device, making appropriate adjustments for differences in specificity. Google argues this would incentivize manufacturers to improve location accuracy.

111. WinnForum proposes that the SAS should estimate CBSD elevation and ground level using detailed terrain databases based on the device's reported operating location. Further, WinnForum states that while the ability to meet the horizontal accuracy requirement is readily achievable, the elevation accuracy requirement significantly exceeds the capability of standard GPS equipment, which will be utilized by both CBSDs and professional installers. WinnForum suggests that, in lieu of the vertical location accuracy requirements, for Category A CBSD's, professional installation reports should include the highest floor from which the device will operate and, for Category B CBSDs, the reports should include the antenna height above ground level.

112. Nokia Solutions also recommends that the Commission establish separate vertical location accuracy requirements for outdoor and indoor installations. Nokia Solutions states that, since the primary method used by many equipment vendors for outdoor location is GPS-based, the vertical location accuracy requirement should be aligned to the US Government Position Accuracy standard for worst site conditions as stated in the Global Positioning System Standard Positioning Service Performance Standard. Nokia Solutions argues that, since GPS does not work well or at all indoors, the Commission should eliminate the elevation reporting requirement for indoor installations, allowing the SAS to estimate the CBSD elevation, and require only the GPS location of the building for the horizontal location.

113. SIA and NAB both stress the importance of reliable location accuracy necessary to protect incumbent operations. SIA recognizes that complying with the current requirements may be challenging, particularly with respect to indoor devices where GPS data may not be readily available and both SIA and NAB would support looser requirements so long as “worst case” assumptions are built into the calculations to account for the reduced accuracy. However, in regard to vertical location, simply relaxing the accuracy requirements and allowing the SAS to “estimate” or “compute” a device's elevation is not an acceptable solution, given the importance of a device's vertical position in calculating the potential for harmful interference. Therefore, NAB and SIA argue, the Commission must implement a larger separation distance to account for this uncertainty, if a device cannot meet the requirements or the SAS cannot independently verify a device's elevation.

114. WISPA opposes the petitions that propose to relax or eliminate the existing vertical location accuracy requirements and argues that there is no current mechanism for CBSDs or an SAS to determine the antenna height above ground within the required accuracy. WISPA states the elevation of the CBSD becomes irrelevant for CBSDs installed using external antenna systems and that only the elevation of the actual antenna is relevant for interference mitigation purposes. According to WISPA, the only way for the SAS to ascertain the CBSD antenna system elevation is by using location information provided by a professional installer.

115. Discussion. We maintain the location accuracy requirements established in the
3.5 GHz R&O
and decline the Nokia Solutions and WinnForum Petitions insofar as they request that we modify these rules. We recognize that there are technological challenges to achieving indoor location accuracy. However, as we stated in the
3.5 GHz R&O,
CBSD location is essential for coordinating interactions between and among users in the band and for protecting Incumbent Access users from harmful interference. Without accurate location data, SASs cannot fulfill their core functions in effectively instructing CBSDs to discontinue their operations or change frequencies to protect Incumbent Users.

116. Further, we believe that the location accuracy requirements in the rules are achievable. First, CBSDs are fixed devices, simplifying the reporting of accurate geo-location information, either automatically or with the input of a professional installer. Second, automated reporting of geo-location to our location accuracy requirements may already be achievable in some conditions (
e.g.,
outdoors with clear line of sight to GPS). In addition, at least one party has stated on the record that it has developed technology that can meet the indoor location accuracy rules set forth in the existing rules. Finally, as discussed in section III(F)(2),

professional installation will play an important role in ensuring the SAS can accurately locate devices while automatic location technologies that meet our requirements are tested and developed.

117. Some commenters also suggest that location accuracy requirements could be met alternatively via SAS calculations. We anticipate that SASs will play a key role in verifying the geographic locations of CBSDs and, as technology continues to develop, we encourage SAS Administrators to offer functions to supplement and reinforce CBSD geo-location functions. However, the CBSD is the best source of its own location information, and such features will not discharge the CBSD from complying with our rules.

118. Finally, regarding Nokia Solutions' suggestion that we allow operators to meet vertical location accuracy requirements at a certain confidence level, we decline to make changes to the existing rules. For the aforementioned reasons, the current rules ensure that the SAS can properly locate CBSDs in order to perform its core functions, and we believe them to be achievable over time.

2. Automated Geo-Location and Professional Installation for CBSDs

119. Background. In the
3.5 GHz R&O,
we concluded that Category A CBSDs may utilize either a technical geo-location capability or be professionally installed while Category B CBSDs must be professionally installed (47 CFR 96.39(a), 96.45(a)). We noted that, since CBSDs will be fixed installations, the professional installation option should allow for network deployment in the near term while automatic geo-location technologies for this band are tested and developed that meet our accuracy requirements. We also strongly encouraged the SAS and user community, through multi-stakeholder fora or industry associations, to develop programs for accrediting professional installers who receive training in the relevant part 96 rules and associated technical best practices.

120. NAB and SIA argue that the Commission should eliminate the option for professional installers to report the locations of CBSDs and, instead, require all CBSDs to include a geo-location capability. NAB contends that the Commission's rule is analogous to a similar professional installation requirement adopted in the White Spaces proceeding. NAB argues that, in that proceeding, it identified several errors in device registrations made by professional installers and that such errors prove that the professional installation option is not acceptable in either the White Spaces or the Citizens Broadband Radio Service. NAB contends that professional installation is not necessary for indoor deployments, citing both technological advances and a compromise approach that it submitted in the White Spaces proceeding. NAB also claims that the professional installation is inherently flawed and cannot be rehabilitated by a certification process. SIA agrees with NAB and contends that, regardless of the safeguards adopted, it will be impossible to remove the risk of human error from installations. In addition, on February 26, 2016, the Commission adopted a Notice of Proposed Rulemaking and Order (81 FR 15210, March 22, 2016) that proposed to require automated geo-location capabilities in White Spaces devices, consistent with an agreement between NAB and several White Spaces device manufacturers.

121. Federated Wireless, Google, T-Mobile, and WISPA disagree with NAB and SIA and argue that the Commission should permit professional installation of CBSDs in the Citizens Broadband Radio Service. Google contends that: (1) Discussions of individual records in the White Spaces proceeding are not relevant to this proceeding and that, in any case, the White Spaces entries may have been good faith test cases; (2) the record demonstrates that professional installers can protect Incumbent Access users; and (3) the industry is working collaboratively to develop an effective framework for certifying professional installers in the band. Federated Wireless agrees and argues that, given the requirements of the band, SAS Administrators and Citizens Broadband Radio Service users will be incentivized to ensure that all geo-location information provided to the SAS is accurate. Federated Wireless also notes that professional installation has been used successfully in a number of other licensed services—including two-way satellite broadband.

122. Discussion. We deny NAB and SIA's petitions for reconsideration of the professional installation rule. We also decline to mandate automated geo-location capabilities for CBSDs. As described in the
3.5 GHz R&O,
accurate CBSD location information is essential for coordinating interactions between and among users in the band and for protecting federal and non-federal Incumbent Users from harmful interference. However, we also noted that, while we expect location accuracy technology to continue to develop, in many circumstances, automated reporting of geo-location information that complies with our accuracy requirements will be challenging in this band given currently available technology. Professional installation is intended to fill that gap and facilitate deployment of CBSDs with accurately reported geo-location information while the next generation of automatic geo-location technology is developed.

123. Based on the record, we are not convinced that the capabilities of today's equipment and technology are sufficiently developed to ensure that CBSDs will be able to perform automated geo-location functions in order to reliably meet the location accuracy requirements for the Citizens Broadband Radio Service. As a result, limiting CBSDs to automated geo-location as the only way to meet these requirements would deter near-term deployment on any reasonable scale in the 3.5 GHz Band. As discussed in detail above, several petitioners highlighted the difficulties associated with attaining an accurate vertical reading within +/− 3 meters. Federated Wireless also argues that, while current technology may be sufficient to provide the SAS with a CBSD's location at the requisite degree of accuracy in some outdoor situations, such readings may not be currently possible for a variety of indoor deployments in this band. Since we expect much of the deployment in the 3.5 GHz Band to be indoors, the inability of a CBSD to provide its location indoors would be fatal to many potential use cases for the Citizens Broadband Radio Service. While we are encouraged by iPosi's claim that its technology can provide indoor accuracy readings that meet or exceed or requirements, it has not yet been used commercially in the 3.5 GHz Band, so it is yet to be determined if this technology is appropriate—or economically viable—for all use cases at this time. Thus, while the accuracy of geo-location technology is improving, integrated geo-location technology may not be a viable option for all potential network deployments in the 3.5 GHz Band at this time.

124. We also find unconvincing NAB and SIA's reliance on NAB's claims regarding inaccurately entered location information in the White Spaces databases. NAB and SIA assert that, since professional installers allegedly entered inaccurate locations of devices in White Spaces databases, the entire notion of a professional installation regime is inherently flawed. Indeed, NAB claims that professional installation has proven to be inherently unreliable and that it cannot be rehabilitated through any kind of certification regime. NAB and SIA reach

these conclusions despite the fact that no SASs have been approved or CBSDs deployed in the Citizens Broadband Radio Service and, as such, there is no evidence of actual harm or impropriety in the band to support their claims. Moreover, these parties have provided no convincing evidence that a professional installation option in this band presents any significant potential for such harm. The alleged failures of a dissimilar, uncertified professional installation regime in another service do not warrant eliminating the professional installation option for the Citizens Broadband Radio Service.

125. The Commission noted that the recent changes proposed in the White Spaces NPRM, which included a proposal to eliminate the professional installer option for fixed White Space devices, were “based upon the circumstances specific to fixed white space devices and white spaces databases.” In the White Spaces service, the Commission determined not to “define the qualifications of a professional installer in the rules.” Here, in contrast, as explained in the
3.5 GHz R&O
and detailed below, the Commission will require professional installers to be trained and certified using an established industry-led process.

126. NAB and SIA unfairly dismiss the importance of a robust industry certification process for professional installers. By relying on such a certification process here, as the Commission has in a variety of other contexts, the rules provide an important protection against the prospect that “any purchaser of a device” could serve as a professional installer. We reiterate that industry-led professional accreditation processes have been used by the Commission and have, in fact, proven successful in other similar situations. In the
3.5 GHz R&O,
we recognized the importance of accurate geo-location information and we strongly encouraged prospective SAS Administrators and Citizens Broadband Radio Service users to develop programs for accrediting professional installers and associated technical best practices. WinnForum announced that, consistent with the Commission's wishes, its members are developing a set of professional installation standards to be implemented by SAS Administrators. Any certification regime developed by WinnForum—or any other entity or organization—must ensure that registered CBSDs comply with the Commission's geo-location rules. WTB and OET will review the SAS's ability to implement and verify the information submitted by professional installers as part of the SAS approval process.

127. Most importantly, the White Spaces service itself is not directly analogous to the Citizens Broadband Radio Service. While both White Spaces devices and CBSDs rely on the White Space databases and SASs, respectively, to protect incumbent services, White Space devices are unlicensed and have no expectation of interference protection. On the other hand, the Citizens Broadband Radio Service is a licensed service in which SASs must be able to effectively coordinate CBSD interactions (both PAL and GAA) to prevent interference between and among the three tiers of users and ensure a stable spectral environment for commercial operations in the 3.5 GHz Band. In other words, in the Citizens Broadband Radio Service the accuracy of the information is important both to protect incumbent services and to protect and enable every other user. This licensed nature of the service coupled with industry certification requirements for professional installers provides a higher degree of accountability for Citizens Broadband Radio Service users and SAS Administrators, ensuring that CBSD locations are accurately reported and verified. In addition, all Citizens Broadband Radio Service users have the rights and obligations incumbent on all Commission licensees, which include serious consequences for violation of Commission rules, including potential revocation and license qualification issues. The Commission has extensive mechanisms available to it to ensure that licensees comply with its rules.

128. In addition, as the Commission has stated on several occasions, approved SASs will have capabilities and responsibilities that exceed those of White Spaces database administrators. Drawing on the lessons learned from the White Spaces proceeding, the Commission will expect SAS Administrators to take appropriate steps to authenticate and verify information that is submitted by professional installers and to immediately correct any inaccurate information in their databases (47 CFR 96.53(d), 96.57(a), 96.63(f)). Our rules require authentication of CBSDs with an SAS and require that SAS Administrators maintain the accuracy of stored data, including CBSD records. The latter requirement places a duty on SAS Administrators to take reasonable steps to validate newly entered data and to purge obsolete data (47 CFR 95.55). Federated Wireless also notes that there are a variety of “quality control methods” that an SAS Administrator may employ—including IP validation, Wi-Fi assistance, and downstream infrastructure coordination—to help verify a CBSD's location. We expect SAS Administrators to develop and implement technological safeguards appropriate to ensure the integrity and accuracy of location data submitted by CBSDs, and we will carefully review proposals from prospective SAS Administrators to determine whether they have demonstrated the capability to do so.

129. While we believe that professional installation is necessary and appropriate for the Citizens Broadband Radio Service at this time, future technological developments may obviate the need to rely on professional installation to ensure the accuracy of CBSDs' location information in some circumstances. Accordingly, we direct WTB and OET to seek input on developments in geo-location technology for CBSDs and the status of the professional installation regime in the Citizens Broadband Radio Service no later than April 28, 2020.

3. End User Device Requirements

130. Background. In its petition, SIA seeks reconsideration of the Commission decision not to mandate that End User Devices include geo-location capabilities. SIA argues that such a mandate is necessary so that an SAS is aware of the location of End User Devices and without such a requirement, the SAS calculations to protect FSS earth stations must be based on worst-case assumptions about location. SIA states these assumptions would include the maximum operational distance between the End User Device and CBSD and the maximum number of End User Devices that could be served by the CBSD. In the alternative, the Commission could define a maximum deployment radius. However, SIA argues, “the use of such worst-case assumptions would result in fewer End User Devices being authorized—and therefore less efficient utilization of the spectrum—than if the SAS had actual location data for each device.”

131. Google and WISPA expressly oppose mandating End User Devices to include geo-location technology. Google argues that a geo-location requirement would unnecessarily limit the types of devices available to consumers, as Wi-Fi dongles and other miniature broadband devices are so small that adding geo-location technology would fundamentally alter the form of the device. Both WISPA and Google claim that such a requirement is not needed to protect users from interference, as the

SAS can take into account the “cloud” of End User Devices associated with a particular CBSD when calculating interference protection and the Commission requires End User Devices to positively receive and decode authorization signals from CBSDs.

132. Rajant states that while it is not opposed to requiring geo-location in End User Devices, it would add additional costs to operation in the band. Further, Rajant states that it plans to deploy in places such as enclosed stadiums and underground mass transit tunnels where it would be difficult to obtain GPS location data and while GPS simulators are available, they would be burdensome and hinder flexibility. Therefore, Rajant argues that the Commission should not require geo-location for consumer devices and limit such a requirement to devices intended for industrial, public safety, or commercial use in confined, managed sites.

133. Discussion. We deny SIA's request to mandate geo-location technology in all End User Devices and find that such a requirement is not necessary to ensure compliance with our location accuracy rules or to effectively mitigate interference into incumbent systems. We recognize that FSS earth station licensees are concerned about interference from End User Devices and, indeed we sought comment on how to address these issues in the
Second FNPRM.
However, we agree with Google and WISPA that it is not necessary to mandate that End User Devices include automatic geo-location capabilities to effectively protect Incumbent Users from interference. In addition, such a requirement would unnecessarily limit the types of consumer devices that may be deployed and utilized in the 3.5 GHz Band.

134. Indeed, the rationale we articulated in section III(F)(2) for not requiring automatic geo-location reporting by CBSDs is even more compelling in the case of End User Devices. End User Devices operate at a much lower power than even Category A CBSDs, lowering their potential interference effects and reducing their range of operation. End User Devices are also inherently limited in their area of operation by the coverage of a given CBSD or network of CBSDs. Moreover, since End User Devices will likely include mobile devices—as opposed to fixed CBSDs—reporting their location to the level of accuracy required by our rules would likely exceed the limits of current technology in many locations.

135. Further, the SAS is responsible for managing CBSDs, not End User Devices. Requiring End User Devices to report their locations to the SAS and requiring the SAS to track and manage these devices would greatly exceed the limits of the SAS's responsibilities. As such, it is not appropriate to include End User Devices in our location accuracy rules. However, as noted by WISPA, the rules do require End User Devices to “positively receive and decode an authorization signal transmitted by a CBSD, including the frequencies and power limits for their operation,” (47 CFR 96.47(a)) and any device to be certified by the Commission must meet these requirements. Both Google and WISPA also state that WinnForum is reviewing how to treat End User Devices in interference calculations, which will further supplement the SAS's ability to account for End User Device locations. WTB and OET will review any such approaches submitted during the SAS approval process.

G. PAL Protection Criteria

136. Background. To ensure that Priority Access operations are protected from harmful interference, we adopted an aggregate received signal level at PAL license boundaries to be at or below an average power level of −80 dBm when integrated over a 10 MHz reference bandwidth with the measurement antenna placed at a height of 1.5 meters above ground level (47 CFR 96.41(f)). We also permitted Priority Access Licensees to agree to an alternative limit other than −80 dBm/10 MHz at their Service Area boundaries and communicate it to an SAS. In addition, we noted that these signal level requirements would not apply to adjacent census tracts held by the same Priority Access Licensee.

137. WinnForum asks that the Commission modify its PAL protection criteria to more effectively reflect real world interference concerns and protect Priority Access Licensees. WinnForum contends that the PAL protection rule creates several problems that the Commission did not consider in developing the
3.5 GHz R&O.
According to WinnForum, these problems include: (1) The requirement would place a significant burden on the SAS by requiring it to calculate point-to-line interference along a lengthy border; (2) border protections may not effectively protect interior portions of a Priority Access Licensee's Service Area; (3) high elevation census tracts will have a disproportionate effect on CBSD deployments; and (4) the requirement will unnecessarily block co-channel devices. WinnForum suggests that the SAS implement an alternate protection scheme whereby the SAS would protect an operator-defined contour around Priority Access CBSDs to a protection level of −80 dBm/10 MHz anywhere within the contour. WinnForum claims that this revised approach addresses all of the concerns raised in its Petition. Federated Wireless, Google, and Motorola Solutions support WinnForum's Petition. WISPA also agrees that the −80 dBm criterion is inadequate for the reasons described by WinnForum.

138. Discussion. We agree with WinnForum's Petition in part and, accordingly, we revise the rule. Under the revised rule, allowable interference will be calculated for the area within the PAL Protection Area (47 CFR 96.3) described in detail in section IV(A) below rather than along the borders of a Priority Access Licensee's Service Area (47 CFR 96.3). To protect CBSDs authorized to provide service on a Priority Access basis, the SAS must not authorize other CBSDs—whether Priority Access or GAA—on the same channel in geographic areas and at maximum power levels that will cause aggregate interference in excess of −80 dBm/10 MHz channel within a PAL Protection Area. Consistent with our approach elsewhere in this Order, the aggregate co-channel interference level will be defined by a common models utilizing common inputs and assumptions. These models, inputs, and assumptions—including the propagation model and any clutter or terrain assumptions—will be determined during the SAS approval process. This approach is also consistent with the methods that will be used to model and measure the aggregate interference to protect incumbent FSS earth stations and incumbent federal radar systems.

139. Several commenters, including Federated Wireless, Google, Motorola Solutions, and WinnForum support a protection methodology based on modeled aggregate interference protections within the area served by a Priority Access Licensee rather than along the border of a given Service Area or census tract. Notably, Google and WinnForum contend that a protection methodology that utilizes point-to-area interference models to calculate aggregate interference into a Priority Access Licensee's service area will be relatively simple and inexpensive for SASs to implement. Motorola Solutions, WinnForum, and Google also highlight several negative unintended consequences of the Commission's rule requiring CBSDs to meet an aggregate interference threshold along the border of a Service Area.

140. We find the evidence presented by Petitioners compelling and modify section 96.41(d) (47 CFR 96.41(d)) to address the concerns raised in their filings. We note that there were no objections to the protection level of −80 dBm/10 MHz and, indeed, several petitioners supported this interference protection level. Therefore, under the revised rule, the SAS must assign CBSDs such that the modeled aggregate power of co-channel CBSDs is no greater than −80 dBm/10 MHz within the PAL Protection Area. Consistent with our approach to geographic guard bands, described in section IV(A), we conclude that the SAS may not consider adjacent channel interference when calculating these protections and assigning CBSDs. We believe that the stringent out-of-channel emission limits set forth in section 96.41 (47 CFR 96.41) are sufficient to make adjacent channel interference unlikely, particularly for synchronized systems and Category A CBSDs.

H. FSS Protection

141. In its petition, SIA asked the Commission to reconsider or clarify several of its rules regarding the protection of in-band and out-of-band FSS earth stations. These issues included: (1) The status of new FSS earth stations in the band; (2) interference notification procedures; (3) protections for international FSS earth stations; (4) FSS registration requirements; and (5) clarification of protections afforded to in-band and out-of-band earth stations. Specific protection methods for in-band and out-of-band FSS earth stations were raised by the Commission in the
Second FNPRM
and, as such, are addressed in section IV(C) below. SIA's other requests are addressed in this section.

1. Status of New In-band FSS Earth Stations

142. Background. In the
3.5 GHz R&O,
the Commission adopted a change to the Table of Allocations limiting co-primary FSS earth stations in the 3600-3650 MHz band to those authorized prior to, or granted as a result of an application filed prior to the effective date of the
3.5 GHz R&O,
and constructed within 12 months of the

[Text truncated at 120,000 characters. The full text is on the page linked above.]

---

Source: Frix Law Library, https://www.frixlaw.com/law-library/documents/fr%3A2016-14505. Public record. Not legal advice.