Service Rules for Advanced Wireless Services H Block-Implementing Section 6401 of the Middle Class Tax Relief and Job Creation Act of 2012 Related to the 1915-1920 MHz and 1995-2000 MHz Bands

FEDERAL COMMUNICATIONS COMMISSION

47 CFR Parts 1 and 27

[WT Docket No. 12-357; FCC 13-88]

Service Rules for Advanced Wireless Services H Block—Implementing Section 6401 of the Middle Class Tax Relief and Job Creation Act of 2012 Related to the 1915-1920 MHz and 1995-2000 MHz Bands

AGENCY:

Federal Communications Commission.

ACTION:

Final rule.

SUMMARY:

In this document, the Commission adopts rules to auction and license ten megahertz of paired spectrum at 1915-1920 MHz and 1995-2000 MHz—the H Block. This action implements the Congressional directive in the Middle Class Tax Relief and Job Creation Act of 2012 (Spectrum Act) that we grant new initial licenses for these spectrum bands through a system of competitive bidding. In so doing, we extend the widely deployed broadband Personal Communications Services (PCS) band, which is used by the four national wireless providers, as well as regional and rural providers, to offer mobile service across the United States. This additional spectrum for mobile use will help ensure that the speed, capacity, and ubiquity of the Nation's wireless networks keep pace with the skyrocketing demand for mobile services.

DATES:

Effective September 16, 2013 except for 47 CFR 1.2105(a)(2)(xii), 27.12, and 27.17, which contain information collection requirements that have not been approved by the Office of Management and Budget (OMB), Control Number 3060-1184. The Commission will publish a document in the

Federal Register

announcing the effective date of those sections.

We also note that several rules that are not being amended herein are subject to OMB review because they are imposing a new information collection upon a new group of respondents, i.e., the H Block licensees. The rules in question are 47 CFR 1.946 and 27.10. The Commission will publish a document in the

Federal Register

announcing the approval of information collection for those sections.

ADDRESSES:

Federal Communications Commission, 445 12th Street SW., Washington, DC 20554. A copy of any comments on the Paperwork Reduction Act information collection requirements contained herein should be submitted to the Federal Communications Commission via email to

PRA@fcc.gov

and to Judith B. Herman, Federal Communications Commission, Room 1-B441, 445 12th Street SW., Washington, DC 20554 or via the Internet at

Judith B. Herman@fcc.gov.

FOR FURTHER INFORMATION CONTACT:

Matthew Pearl of the Broadband Division, Wireless Telecommunications Bureau, at (202) 418-BITS or

Matthew.Pearl@fcc.gov.

For additional information concerning the Paperwork Reduction Act information collection requirements contained in this document, contact Judith B. Herman at (202) 418-0214, or via email at

PRA@fcc.gov.

SUPPLEMENTARY INFORMATION:

This is a summary of the Commission's

H Block Report and Order,

FCC 13-88, adopted on June 27, 2013 and released on June 27, 2013. The full text of this document is available for inspection and copying during normal business hours in the FCC Reference Information Center, Room CY-A257, 445 12th Street SW., Washington, DC 20554. The complete text may be purchased from the Commission's duplicating contractor, Best Copy and Printing, Inc. (BCPI), Portals II, 445 12th Street SW., Room CY-B402, Washington, DC 20554, (202) 488-5300, facsimile (202) 488-5563, or via email at

fcc@bcpiweb.com.

The complete text is also available on the Commission's Web site at

http://transition.fcc.gov/Daily_Releases/Daily_Business/2013/db0627/FCC-13-88A1.pdf.

Alternative formats (computer diskette, large print, audio cassette, and Braille) are available by contacting Brian Millin at (202) 418-7426, TTY (202) 418-7365, or via email to

bmillin@fcc.gov.

Synopsis

I. Introduction

1. Today we increase the Nation's supply of spectrum for flexible-use services, including mobile broadband, by adopting rules to auction and license ten megahertz of paired spectrum at 1915-1920 MHz and 1995-2000 MHz—the H Block. This action implements the Congressional directive in the Middle Class Tax Relief and Job Creation Act of 2012 (Spectrum Act) that we grant new initial licenses for these spectrum bands through a system of competitive bidding. In so doing, we extend the widely deployed broadband Personal Communications Services (PCS) band, which is used by the four national wireless providers, as well as regional and rural providers, to offer mobile service across the United States. This additional spectrum for mobile use will help ensure that the speed, capacity, and ubiquity of the Nation's wireless networks keep pace with the skyrocketing demand for mobile services.

II. Background

2. In February 2012, Congress enacted Title VI of the Middle Class Tax Relief and Job Creation Act of 2012 (Spectrum Act). The Spectrum Act includes several provisions to make more spectrum available for commercial use, including through a system of competitive bidding, and to improve public safety communications. Among other things, the Spectrum Act states that the Commission, by February 23, 2015, shall allocate the H Block bands—1915-1920 MHz and 1995-2000 MHz—for commercial use, and through a system of competitive bidding grant new initial licenses for the use of each band, subject to flexible use service rules. Congress provided, however, that if the Commission determines that either of the bands cannot be used without causing harmful interference to commercial licensees in 1930-1995 MHz (PCS downlink), then the Commission shall not allocate such band for commercial use or grant new licenses for the use of such band. Additionally, Sections 6401(c)(4) and 6413 of the Spectrum Act specify that the proceeds from an auction of licenses in the 1995-2000 MHz band and in the 1915-1920 MHz band shall be deposited in the Public Safety Trust Fund. Section 6413 of the Spectrum Act specifies how the funds deposited into the Public Safety Trust Fund shall be used, and these purposes include the funding of (or reimbursement to the U.S. Treasury for the funding of) the nationwide, interoperable public safety broadband network by the First Responder Network Authority (FirstNet). The rules we adopt today will enable the H Block spectrum to be the first spectrum specified by the Spectrum Act to be licensed by auction, and thus likely will represent the first steps toward this statutory goal.

3. In response to the Spectrum Act and to help meet the growing demand for wireless spectrum, in December 2012, the Commission adopted the

H Block NPRM.

In the

H Block NPRM,

the Commission proposed to increase the Nation's supply of spectrum for mobile broadband by applying Advanced Wireless Services (AWS) flexible use wireless service rules in 10 megahertz of spectrum adjoining the widely deployed Broadband PCS (PCS) band, at 1915-1920 MHz and 1995-2000 MHz.

4. The

H Block NPRM

also represents a renewed Commission effort to bring this spectrum to market. The Commission first proposed licensing,

operating, and technical rules for this spectrum band in 2004. The 2004

AWS-2 NPRM

sought comment on strict power and out-of-band emission (OOBE) limits for mobile transmissions in the 1915-1920 MHz band, because of concerns about potential harmful interference to PCS mobile reception. Service Rules for Advanced Wireless Services in the 1915-1920 MHz, 1995-2000 MHz, 2020-2025 MHz and 2175-2180 MHz Bands, WT Docket No. 04-356, Notice of Proposed Rulemaking, 19 FCC Rcd 19263 (2004). In response to those proposals, most commenters agreed with such concerns. In 2008, the Commission issued a

Further Notice of Proposed Rulemaking

in which it sought to supplement the record. Service Rules for Advanced Wireless Services in the 1915-1920 MHz, 1995-2000 MHz, 2020-2025 MHz and 2175-2180 MHz Bands, WT Docket Nos. 07-195, 04-356, Further Notice of Proposed Rulemaking, 23 FCC Rcd 9859 (2008). Those 2008 proposals included strict OOBE limits for the Lower H Block of 90 + 10 log

10

(P) dB, where (P) is the transmitter power in watts, within the PCS band, and a power limit of 23 dBm/MHz Equivalent Isotropically Radiated Power (EIRP). The record again reflected the commenters' calls for strict interference limits, particularly the OOBE limit, in 1915-1920 MHz to avoid harmful interference to PCS mobile receivers. In the

2008 NPRM,

the Commission also proposed prohibiting mobile transmissions in the 1995-2000 MHz band, and proposed the typical interference rule of an OOBE limit of 43 + 10 log

10

(P) dB, where (P) is the transmitter power in watts, for base and fixed stations for emissions outside of the 1995-2000 MHz band, and a power limit of 1640 watts EIRP for emissions less than 1 MHz and 1640 watts/MHz for emissions greater than 1 MHz in non-rural areas and double these power limits in rural areas.

5. The spectral proximity of the AWS-4 Band (2000-2020 MHz and 2180-2200 MHz) to the Upper H Block is relevant to the present discussion. The Commission's December 2012

AWS-4 Report and Order

established licensing, operating, and technical rules for terrestrial AWS-4 operations in the 2000-2020 MHz and 2180-2200 MHz bands. The technical rules for the AWS-4 uplink at 2000-2020 MHz balanced the public interest benefits associated with potential uses of the relevant bands. Specifically, the Commission placed limited restrictions on AWS-4 uplink operations that allow for flexible use of the AWS-4 band while also potentially enabling full flexible downlink use of the 1995-2000 MHz band. The Commission explained that it based its determination on, among other things, the asymmetrical nature of broadband traffic (with more downlink than uplink being used), the fact that any limitations on AWS-4 were more than offset by the considerable increase in flexibility that the Commission was providing AWS-4 licensees by granting them terrestrial use rights under the Commission's part 27 rules. In sum, the Commission stated that the AWS-4 technical rules would enable both the AWS-4 band and the 1995-2000 MHz band to be used for providing flexible use services in the most efficient manner possible.

6. In December 2012, the Commission adopted the

H Block NPRM.

Comments on the

H Block NPRM

were due on February 6, 2013 and replies were due March 6, 2013. Fifteen comments and seven replies were filed in response to the

H Block NPRM.

In addition, as permitted under our rules, numerous

ex parte

presentations have been submitted into the record.

7. As observed in the

H Block NPRM,

circumstances have changed in the years since the Commission previously sought comment on the H Block spectrum bands. Wireless broadband technologies and the wireless broadband industry have evolved considerably. Additionally, Congress enacted the Spectrum Act. Accordingly, we provided notice that our determinations here would be based solely on the record developed in response to the

H Block NPRM,

and we invited parties to re-file in this docket earlier comments with any necessary updates.

III. Discussion

8. In this

H Block Report and Order,

we implement the Spectrum Act provisions pertaining to the H Block and build upon recent Commission actions to increase the availability of spectrum for wireless use by adopting rules to grant licenses for the H Block for terrestrial fixed and mobile use via a system of competitive bidding. As explained below, we adopt H Block terrestrial service, technical, and licensing rules that generally follow the Commission's part 27 flexible use rules, modified as necessary to account for issues unique to the H Block bands. Specifically, we take the following actions:

• We find that the Commission's prior action to allocate the H Block for Fixed and Mobile use satisfies the requirement of the Spectrum Act that we allocate this spectrum for commercial use.

• We find that we are required to adopt flexible use service rules for the H Block and that we are required to license this spectrum using a system of competitive bidding, unless we determine that either the 1915-1920 MHz band or the 1995-2000 MHz band cannot be used without causing harmful interference to the broadband PCS downlink band at 1930-1995 MHz.

• We find that, consistent with the technical rules we adopt, the use of both the 1915-1920 MHz band and the 1995-2000 MHz band can occur without causing harmful interference to broadband PCS downlink operations at 1930-1995 MHz.

• We adopt as the H Block band plan the 1915-1920 MHz band paired with the 1995-2000 MHz band, configured as 5 + 5 megahertz blocks, and licensed on an Economic Area (EA) basis.

• We adopt technical rules for the H Block, including rules governing the relationship of the H Block bands to adjacent and nearby bands, with a particular focus on adopting rules for the 1915-1920 MHz band that protect operations in the broadband PCS band at 1930-1995, as required by the Spectrum Act.

• We adopt technical rules that authorize the 1915-1920 MHz band for mobile and low power fixed operations (uplink) and the 1995-2000 MHz band for base and fixed operations (downlink).

• We adopt cost sharing rules that require H Block licensees to pay a

pro rata

share of expenses previously incurred by UTAM, Inc. and by Sprint in clearing incumbents from the 1915-1920 MHz band and the 1995-2000 MHz band, respectively.

• We adopt a variety of flexible use regulatory, licensing, and operating rules for H Block licensees.

• We adopt procedures to assign H Block licenses through a system of competitive bidding.

A. Spectrum Act Provisions for 1915-1920 MHz and 1995-2000 MHz

9. The Spectrum Act, among other requirements, provides that the Commission shall allocate for commercial use and license using a system of competitive bidding the H Block no later than February 23, 2015.

10. Section 6401(b) of the Spectrum Act provides that for certain spectrum bands, including H Block, the Commission must allocate the spectrum for commercial use and grant new initial licenses for that spectrum through a system of competitive bidding by February 23, 2015. Middle Class Tax Relief and Job Creation Act of 2012, Section 6401(b), 47 U.S.C. 1451(b).

However, section 6401(b) also provides that the Commission may not allocate the H Block for commercial use nor grant H Block licenses should it determine that such spectrum cannot be used without causing harmful interference to commercial mobile service licensees operating in the frequencies between 1930 megahertz and 1995 megahertz.

11. To implement these requirements, in the

H Block NPRM,

the Commission sought comment on the Spectrum Act's four main statutory elements relating to the H Block: (1) Allocation for commercial use; (2) flexible use; (3) assignment of licenses; and (4) a determination regarding interference. Below, we address the relevant comments and discuss our conclusions.

1. Allocation for Commercial Use

12. The Spectrum Act requires the Commission to allocate the H Block spectrum bands, 1915-1920 MHz and 1995-2000 MHz, for commercial use. As the Commission observed in the

H Block NPRM,

the Spectrum Act does not define the phrase, “allocate . . . for commercial use.” The Commission posited that the Spectrum Act requires us to make any necessary changes to the Non-Federal Table of Allocations to reflect that the H Block bands could be used commercially by, and licensed to, non-Federal entities under flexible use service rules unless the band cannot be used without causing harmful interference to commercial mobile service licensees in the PCS downlink band. The Commission observed that the H Block spectrum's pre-existing allocation was for non-Federal, Fixed and Mobile use on a primary basis and designated for use in the commercial PCS/AWS bands, and that this prior allocation appeared to be fully consistent with section 6401 of the Spectrum Act. The Commission sought comment on this tentative conclusion. In response, commenters agreed with the Commission's tentative conclusion that the H Block's existing allocation met the requirements of the Spectrum Act.

13. We find that the existing allocation of the H Block for non-Federal Fixed and Mobile use on a primary basis meets the “commercial use” allocation requirement of section 6401(b)(1)(A) of the Spectrum Act. As the record indicates, the Commission has already allocated both blocks of the H Block spectrum for non-Federal Fixed and Mobile use on a primary basis. Specifically, in 2004, the Commission adopted the present spectrum pairing. Thus, no further action to allocate the H Block spectrum bands for commercial use pursuant to the Spectrum Act is necessary.

2. Flexible Use

14. The Spectrum Act also requires the Commission to license the H Block under flexible use service rules. In the

H Block NPRM,

the Commission proposed that any service rules adopted for the H Block permit a licensee to employ the spectrum for any non-Federal use permitted by the United States Table of Frequency Allocations, subject to our part 27 flexible use and other applicable rules, including service rules to avoid harmful interference. Part 27 licensees must also comply with other Commission rules of general applicability.

See

47 CFR 27.3;

see

also infra section III.E.6. (Regulatory Issues, Other Operating Requirements). In addition, flexible use in international border areas is subject to any existing or future international agreements.

See infra

section III.C.3. (Canadian and Mexican Coordination). Thus, the Commission proposed the H Block may be used for any fixed or mobile service that is consistent with the allocations for the band. Commenters uniformly supported this proposal.

15. We adopt the Commission's proposal to license the H Block under flexible use service rules. We find the Spectrum Act's direction on this matter clear and direct—we are required to grant licenses “subject to flexible-use service rules.” Accordingly, adopting the flexible use service rules for the H Block, which we do in the sections below, will give effect to the legislative mandate. Adoption of flexible use service rules, moreover, is consistent with prior congressional and Commission actions that promote flexible spectrum allocations and the record before us. As CCA comments, flexible use allows licensees to innovate and “rapidly respond to changing consumer demands for wireless services . . . [and] encourage[s] the similarly timely deployment of innovative commercial wireless services to the public.”

3. Assignment of Licenses

16. The Spectrum Act mandates that the Commission grant new initial licenses for the 1915-1920 MHz and 1995-2000 MHz bands through a system of competitive bidding pursuant section 309(j) of the Communications Act. In the

H Block NPRM,

the Commission proposed applying competitive bidding rules to resolve any mutually exclusive applications accepted for H Block licenses. Parties uniformly supported the Commission's proposal to assign the H Block spectrum through a system of competitive bidding. For example, MetroPCS voiced its support that the Commission was correctly interpreting the Spectrum Act and that the H Block should be licensed through competitive bidding. We agree and find that the Spectrum Act's requirement that we grant H Block licenses “through a system of competitive bidding” clear and unambiguous. Thus, as detailed below, we adopt rules to govern the use of a competitive bidding process for licensing the 1915-1920 MHz and 1995-2000 MHz bands.

4. Determination of No Harmful Interference to the 1930-1995 MHz Band

17. The Spectrum Act states that the Commission may not allocate for commercial use or license the H Block if the Commission “determines that” the H Block “cannot be used without causing harmful interference to commercial mobile licensees” in the 1930-1995 MHz band (PCS downlink band). Neither the Spectrum Act nor the Communications Act defines the term “harmful interference.” In performing its statutory role to maximize the public interest in the spectrum, the Commission has adopted a definition for this term, as well as for the unmodified term “interference.” Commission rule 2.1(c) defines “interference” to mean “[t]he effect of unwanted energy due to one or a combination of emissions, radiations, or inductions upon reception in a radiocommunication system, manifested by any performance degradation, misinterpretation, or loss of information which could be extracted in the absence of such unwanted energy.” That same rule defines “harmful interference” to mean “[i]nterference which endangers the functioning of a radionavigation service or of other safety services or seriously degrades, obstructs, or repeatedly interrupts a radiocommunication service operating in accordance with [the International Telecommunications Union] Radio Regulations.” In the

H Block NPRM,

the Commission proposed to use this definition of harmful interference in performing the analysis required by the Spectrum Act. No party opposed the use of this definition.

18. We find it appropriate to use the Commission's existing definition of harmful interference. We presume that Congress was aware of this rule, defining both interference and harmful interference, when it crafted the Spectrum Act and used the term harmful interference. Because the Spectrum Act offers no alternative to the Commission's pre-existing definition of

harmful interference, we believe it reasonable to conclude that Congress intended for it to apply to the situation here.

See Hall

v.

U.S.,

132 S.Ct. 1882, 1889 (2012) (“We assume that Congress is aware of existing law when it passes legislation[.]” (internal quotation marks omitted)). Applying the existing definition of harmful interference to the Spectrum Act provision at issue, we find that we may not allocate for commercial use or license the H Block if we determine that the H Block cannot be used without causing serious degradation, obstruction, or repeated interruption to commercial mobile licensees in the PCS downlink band. We further find that we need not set technical rules so restrictive as to prevent all instances of interference, as opposed to harmful interference. Determining

ex ante

when operations in one band will seriously degrade, obstruct, or repeatedly interrupt operations in another band necessarily involves the Commission examining the particular interference scenario that is likely to arise and exercising its predictive judgment, which is entitled to deference.

See Northpoint Technology, Ltd.

v.

FCC,

414 F.3d 61, 69 (D.C. Cir. 2005) (deferring to the Commission's interpretation of “harmful interference” as the phrase was applied under the Rural Local Broadcast Signal Act of 1999);

see also American Radio Relay League, Inc.

v.

FCC,

524 F.3d 227, 233 (D.C. Cir. 2008) (“considerable deference” on “highly technical question” involving harmful interference). For example, in 1999, Congress adopted a statute that directed the Commission to “ensure that no facility [to be newly] licensed or authorized under the [newly enacted Rural Local Broadcast Signal Act of 1999] . . . cause [] harmful interference to the primary users of that spectrum.” In determining technical rules to ensure that the incumbent primary operators were not subject to harmful interference, the Commission established interference parameters designed such that the presence of the new operators' signals “would not be perceptible to the [incumbent operator's] customer in most cases.” The DC Circuit found this “qualitative requirement” to represent a reasonable application of the Commission's harmful interference definition.

Northpoint,

414 F.3d at 69-71. In this similar statutory circumstance, we now establish technical rules (below) for the H Block that will permit use of this block without causing harmful interference (although not necessarily eliminating all interference) to PCS downlink operations.

a. Upper H Block: 1995-2000 MHz

19. The Commission allocated the 1995-2000 MHz band for fixed and mobile use in 2003. In 2004, this spectrum was designated for PCS/AWS base station operations and the Commission proposed service rules. Before the

H Block NPRM

in December 2012, no party had filed technical data or analysis indicating that base station operations in the Upper H Block would cause harmful interference to licensees in the PCS downlink band. Accordingly, in the

H Block NPRM,

the Commission tentatively concluded that base station operations in the Upper H Block posed no likelihood of harmful interference to PCS operations in the 1930-1995 MHz band and that licensing of the Upper H Block could proceed.

20. In light of the technical rules we impose on operations in the Upper H Block, described below, we conclude that operations in the 1995-2000 MHz band will not cause harmful interference to PCS operations in the 1930-1995 MHz band. The rules we adopt herein determine the Upper H Block will be used for base station (

i.e.,

downlink) transmissions. As the 1930-1995 MHz PCS band is used for downlink transmissions, the 1995-2000 MHz band, in many respects, will operate as an extension of the PCS band. As explained below, in contrast to an uplink band adjacent to a downlink band, similarly used bands (

i.e.,

downlink next to downlink) generally do not raise difficult interference scenarios. More specifically, the technical rules we adopt include power limits and OOBE limits for operations in the Upper H Block that are comparable to limits already imposed upon PCS licensees governing the transmission of electromagnetic signals into adjacent PCS bands to prevent harmful interference. As the technical rules we impose for the Upper H Block reflect similar technical constraints as the existing PCS rules—and these rules have allowed robust service to develop in these bands—we find no basis to conclude that the 1995-2000 MHz band “cannot be used without causing harmful interference” to PCS downlink operators at 1930-1995 MHz. Additionally, in response to the

H Block NPRM,

no commenters raised concerns about the potential for harmful interference from the Upper H Block into the 1930-1995 MHz band. In sum, because the 1995-2000 MHz band is adjacent to another downlink band, the technical rules we adopt are comparable to the existing PCS rules for preventing harmful interference and the record demonstrates no concern for harmful interference from the 1995-2000 MHz band into PCS operations in 1930-1995 MHz, we determine the Upper H Block can be allocated for commercial use, assigned via a system of competitive bidding, and licensed subject to flexible use service rules without causing harmful interference to PCS pursuant to the Spectrum Act.

b. Lower H Block: 1915-1920 MHz

21. In designating the 1915-1920 MHz band for PCS/AWS mobile operations in 2004, the Commission concluded that any harmful interference from the Lower H Block to the PCS downlink band (

i.e.,

1930-1995 MHz) could be addressed through service and technical rules. Subsequently, in the

H Block NPRM,

the Commission tentatively concluded that it would be possible to license the Lower H Block under flexible service rules without causing harmful interference to commercial mobile licensees in the 1930-1995 MHz band. Acknowledging the prior concerns with mobile operations in 1915-1920 MHz, the Commission sought comment on the proposed band plan and service rules, and it specifically sought technical analysis on the potential for harmful interference into the PCS downlink band. In response, parties submitted four technical studies and offered numerous comments discussing the potential for harmful interference from Lower H Block operations into operations in the 1930-1995 MHz band. As detailed below, commenters suggest that, with appropriate technical rules, deployment in the Lower H Block can occur without causing harmful interference to the 1930-1995 MHz PCS band.

22. We adopt the Commission's tentative conclusion set forth in the

H Block NPRM,

and determine that operations in the 1915-1920 MHz band, subject to appropriate technical rules, will not cause harmful interference to PCS operations in the 1930-1995 MHz band. As we explain below, in designating the Lower H Block for uplink use, we must address the issue of uplink spectrum in close frequency proximity to the downlink spectrum in the 1930-1995 MHz PCS band. Our analysis is based on our prior findings with respect to similar services, our experience evaluating the probabilistic nature of mobile-to-mobile interference, and our evaluation of the technical studies submitted into the record that examine this specific scenario. Notably, the proponents of these studies acknowledge that the interference scenario at issue—namely, mobile-to-

mobile interference between mobile transmitters operating in the Lower H Block and mobile receivers operating in the PCS downlink band—is inherently a probabilistic one. That is, a number of low probability events all need to occur before an actual Lower H Block transmission would seriously degrade, obstruct, or repeatedly interrupt the ability of the PCS mobile device to receive the PCS signal. As such, the rules we establish below are designed to prevent harmful interference. These rules are not, nor could they reasonably be, designed to prevent all possible instances of interference generally. In sum, we find the technical rules we adopt below will enable commercial use of the Lower H Block without causing harmful interference to PCS operations in the 1930-1995 MHz band. Accordingly, we find no basis to conclude that the 1915-1920 MHz band “cannot be used without causing harmful interference” to PCS downlink operators at 1930-1995 MHz. We therefore determine, consistent with our findings above, that the Lower H Block can be allocated for commercial use, assigned via a system of competitive bidding, and licensed subject to flexible use service rules pursuant to the Spectrum Act. Consequently, we reject Savari's proposal that we make the 1915-1920 MHz band a combination unlicensed PCS (UPCS) and licensed low power band.

See

Savari Comments at 14;

infra

Section III.B.1. (Band Plan, Block Configuration).

B. Band Plan

23. Band plans establish parameters and provide licensees with certainty as to the spectrum they are authorized to use. Here, Congress has identified the H Block bands—1915-1920 MHz and 1995-2000 MHz—as the frequencies for the band plan. To establish the specific band plan for these frequencies, the Commission must determine the block configuration, whether to license the blocks on a geographic area basis and, if so, the appropriate service area. In the

H Block NPRM,

the Commission proposed licensing the H Block as paired 5 megahertz blocks, with the Upper H Block used for high power base stations and the Lower H Block used for mobile and low power fixed operations. The Commission also proposed licensing the H Block on a geographic licensing scheme based on Economic Areas (EAs). Finally, the Commission also sought comment on how best to license spectrum in the Gulf of Mexico. The Commission sought comment on these proposals, including on their associated costs and benefits.

24. In the band plan, based on the record before us, we adopt the H Block band plan of 1915-1920 MHz paired with 1995-2000 MHz, configured as 5 + 5 megahertz blocks, and will license the H Block on an EA basis, including for the Gulf of Mexico. In so doing, we find that 1915-1920 MHz shall be used for mobile and low power fixed (

i.e.,

uplink) operations and 1995-2000 MHz shall be used for base station and fixed (

i.e.,

downlink) operations.

1. Block Configuration

25. In 2004, the Commission designated the H Block for licensed fixed and mobile services, including advanced wireless services. The Commission further decided to pair 1915-1920 MHz with 1995-2000 MHz because it found that doing so would promote efficient use of the spectrum and allow for the introduction of commercial wireless mobile and fixed services. The Commission also observed that it would be advantageous to use the Lower H Block for low power or mobile operations as the adjacent 1910-1915 MHz band is used by PCS mobile operations, and that high power base stations in the band could result in harmful interference to operations in the PCS band.

26. In the

H Block NPRM,

the Commission observed there was no apparent reason to alter the proposed pairing or use of the 1915-1920 MHz and 1995-2000 MHz bands. To ensure the PCS bands were adequately protected from harmful interference due to operations in the Lower H Block, the Commission also proposed to prohibit high power base station operations in 1915-1920 MHz. In response to these proposals, commenters generally supported the Commission's goal of maintaining the pairing of the H Block spectrum and the designated uplink/downlink bands. Additionally, some commenters addressed the Commission's inquiry for alternative configurations of the H Block, which we discuss below. No party presented cost or benefit data in support of its position.

27. We adopt the proposal to maintain the pairing of 1915-1920 MHz with 1995-2000 MHz. In doing so, we observe that Congress, in enacting the Spectrum Act and directing us to license these bands, did not express disagreement with the Commission's earlier determination to pair these bands. We find this approach in the public interest and find that the benefits of this approach likely outweigh any potential costs. As several commenters discuss, pairing the lower and upper portions of the H Block will promote the efficient use of this spectrum and allow for the proliferation of wireless services. In addition, consistent with the record, we expect that adopting the paired spectrum band plan will facilitate the deployment of wireless fixed and mobile services in rural areas. Further, by licensing the H Block as a paired band, we allay the concerns some commenters expressed about the risk of a stranded, standalone block of spectrum that may be unsuitable for mobile broadband use.

28. Additionally, configuring the H Block as a 5 + 5 megahertz band will allow for flexibility and efficiency in the deployment of wireless services and technologies. Five megahertz blocks can support a variety of wireless broadband technologies. While we do not prescribe a specific technology for use in the H Block, we expect that most users of the band will deploy 4G or 3G Frequency Division Duplex (FDD) technologies. Various globally-standardized technologies, including Wideband-Code Division Multiple Access (W-CDMA), High Speed Packet Access (HSPA), and their variants, use 5 + 5 megahertz paired blocks when deployed as FDD. Long Term Evolution (LTE), which commenters indicate is the most likely technology to be deployed in the H Block in the near term, supports a variety of block sizes, including multiples of 5 megahertz. Thus, as C Spire comments, adopting a 5 + 5 megahertz band plan allows an operator using today's LTE technology to deploy in the band.

29. In adopting this band plan, we also adopt the proposal to prohibit high powered fixed and base station operations in the Lower H band,

i.e.,

1915-1920 MHz. Limiting base station operations to the 1995-2000 MHz band will reduce the potential for harmful interference to PCS operations. Because the PCS spectrum immediately proximate to the Lower H Block is used for mobile operations, a high powered signal emanating from 1915-1920 MHz, such as from a base station, may cause harmful interference due to receiver overload. As we discuss below and have concluded previously, the power limits necessary to avoid this potential problem preclude the use of base stations in this band. Therefore, based on the record before us, we determine the 1915-1920 MHz band will be used for mobile operations (uplink) and the 1995-2000 MHz band will be used for base station operations (downlink).

30.

Alternatives.

Our decision today to pair 1915-1920 MHz band with the 1995-2000 MHz band moots concerns that some commenters have raised regarding the possibility of either band standing alone. Specifically, by pairing

these two spectrum bands together, neither the Lower H Block nor Upper H Block will become a standalone “stranded” five megahertz block. In addition, we find it unnecessary to address Savari's suggestion that, as part of its plan to have the Commission license the H Block as a low power guard manager band, the Commission permit the H Block licensee to partner the 1915-1920 MHz band with unlicensed PCS channels in the adjacent 1920-1930 MHz band. Because we decline to adopt Savari's predicate proposal that the H Block be licensed under a low power guard band manager approach, we need not reach the issue of “partnership” with adjacent UPCS channels.

31.

Interoperability.

As discussed below, the H Block spectrum is adjacent to the PCS spectrum and the technical rules we adopt for the H Block would permit the H Block effectively to be operated as an extension of the PCS band. The Commission historically has been interested in promoting interoperability, beginning with the licensing of cellular spectrum. Although the Commission did not adopt a rule to require band-wide interoperability for PCS, it stressed the importance of interoperability by acknowledging industry efforts to establish voluntary interoperability standards. We continue to believe that interoperability is an important aspect of future deployment of mobile broadband services and generally serves the public interest. We note that no party has requested that we impose an interoperability requirement here to further the public interest. We strongly encourage all stakeholders in this ecosystem to develop new equipment in a manner that promotes, rather than hinders, interoperability. We intend to closely monitor the development of the equipment market in the H block and neighboring PCS band as well as other future developments in this band in order to assess whether additional action will need to be taken to promote interoperability.

2. Service Area

a. Geographic Area Licensing

32. In the

H Block NPRM,

the Commission proposed to adopt a geographic area licensing approach for the H Block, reasoning that such an approach is well-suited for the types of fixed and mobile services that would likely be deployed in these bands.

33. We adopt a geographic area licensing scheme for the H Block for the reasons that the Commission articulated in the

H Block NPRM,

namely that it is well-suited for the types of fixed and mobile services that we expect to be deployed in the H Block and will maintain consistency with numerous other bands. Given the record before us, we conclude that this approach is in the public interest and that the benefits of geographic area licensing likely outweigh any potential costs. We find it particularly significant that geographic area licensing in the H Block is consistent with the Commission's licensing approach for other similar commercial bands, including AWS-1, Broadband PCS, Commercial 700 MHz, and AWS-4. As the Commission has observed in the past, geographic licensing also carries many additional benefits, including: (1) Providing licensees with substantial flexibility to respond to market demand, which results in significant improvements in spectrum utilization and allows new and innovative technologies to rapidly develop; (2) permitting economies of scale because licensees can coordinate usage across an entire geographic area to maximize spectrum use; and (3) reducing regulatory burdens and transaction costs because wide-area licensing does not require site-by-site approval, thus allowing a licensee to aggregate its service territories without incurring the administrative costs and delays associated with site-by-site licensing. Further, geographic area licensing in the H Block will allow the Commission to assign initial licenses through a system of competitive bidding in accordance with the Spectrum Act. Finally, we observe that the record supports geographic area licensing for the H Block, which no commenter has opposed.

b. Service Area Size

34. In the

H Block NPRM,

the Commission proposed to license the H Block on an Economic Area (EA) basis. The Commission sought comment on this approach and asked commenters to discuss and quantify the economic, technical, and other public interest considerations of any particular geographic licensing scheme for this band, as well as the impact that any such scheme would have on rural service and competition. Alternatively, the Commission sought comment on nationwide licensing for the H Block, including whether it would maximize or limit the opportunity for licensees to provide the widest array of services and would provide the necessary incentives to expand existing technologies and create new ones. The Commission requested that commenters compare the advantages and disadvantages of nationwide licensing to those of EA licensing. Further, the Commission sought comment on licensing areas smaller than EAs for the H Block, including whether it would facilitate use by smaller and rural operators and whether the benefits of such an approach would outweigh the potential diseconomies of scale. Finally, the Commission requested comment on whether there are any other geographic licensing methods for the H Block that would better meet the Commission's goals.

35. Comments on the proposal were mixed. Some commenters, including both small and large carriers, supported EA-based licensing, while other commenters opposed EAs and advocated license areas smaller than EAs. While one commenter supported either nationwide or large regional (

i.e.,

Major Economic Areas) licenses, several other commenters opposed such a licensing scheme. One party also supported “roadway or highway license[s].” No party, however, provided cost or benefit data to support its position.

36. We will license the H Block on an EA basis. As explained below, licensing based on EAs has been used for similar bands and is a useful and appropriate geographic approach. We believe that licensing the H Block on an EA basis will help us to meet several statutory goals, including providing for the efficient use of spectrum; encouraging deployment of wireless broadband services to consumers; and promoting investment in and rapid deployment of new technologies and services. Given the record before us, we conclude that licensing the H Block on an EA basis is in the public interest and that the benefits of this approach likely outweigh any potential costs.

37. We believe that licensing on an EA-basis strikes the appropriate balance in license size for this band. We find it particularly significant that the two bands adjacent to the H Block, PCS G Block and AWS-4, are licensed on an EA basis. As the record indicates, adopting the same size geographic area as is used in adjacent bands may encourage rapid deployment in and use of the spectrum. Thus, to the extent that licensees for either of those bands ultimately obtain licenses for the H Block, EAs may present opportunities for efficiencies that other geographic license sizes would not offer. For example, AT&T states that EA-based licensing here would be consistent with the Commission's adoption of EA-based licensing in other spectrum bands that will likely be used for mobile broadband. Sprint, moreover, states that the consistent use of EA-based licensing

in PCS, AWS-4, and now H Block will encourage quick deployment in the H Block spectrum.

38. We also believe that licensing this band using EAs will facilitate access to spectrum for both small and large carriers. We believe that it will facilitate access by smaller carriers because EAs are small enough to provide spectrum access opportunities to such carriers. At the same time, EAs are large enough that large carriers can aggregate them up to larger license areas, including into Major Economic Areas (MEAs) and Regional Economic Area Groupings (REAGs), thus achieving economies of scale.

39. Several commenters supported EA-based licensing. For example, as stated above, AT&T and Sprint support EA-based licensing because this band is adjacent to other bands that have been licensed on an EA-basis. MetroPCS explains that EA-based licensing helps to ensure that the bidder that most highly values the spectrum in a particular area acquires that license. C Spire argues that EA-based licensing would “allow for efficient geographic aggregation of licenses. And CCA asserts there are numerous advantages to EA-based licensing, including that it provides “rural and regional carriers [with] reasonable opportunities to bid.”

40. Other commenters opposed EAs as either too large or too small. Commenters proposing smaller geographic license areas advocated for Cellular Market Areas (CMAs), including both Metropolitan Statistical Areas (MSAs) and Rural Service Areas (RSAs). They argued that small and rural carriers cannot afford EAs and that, because EAs include both urban and rural areas, large carriers that purchase EAs can focus their buildout efforts on urban centers to the detriment of rural customers. Another commenter argued that that the H Block should be licensed on a larger-than-EA basis either on a nationwide basis or on a Major Economic Area (MEA).

41. On balance, we are not persuaded that we should adopt geographic license areas smaller or larger than EAs. Rather, we find that—for the H Block—licensing the spectrum on an EA basis best balances the Commission's public interest goals of encouraging widespread geographic buildout (including in rural areas) and providing licensees with sufficient flexibility to scale their networks. We find this particularly so because, as explained above, EA-based licensing will make H Block consistent with two adjacent bands. Moreover, we note that CMAs do not “nest” easily into EAs, which could make it more difficult for licensees to aggregate license areas to match the neighboring bands. Finally, to the extent that an entity desires to obtain access to H Block spectrum for less than an EA geographic area, secondary market transactions (e.g. partitioning) offer a possible way to obtain such access.

42. Finally, we observe that Savari argues that, if the FCC adopts EA-based licensing, it should issue “roadway licenses” that cover highways and areas near highways; areas that, it implies, may lie between EAs. We disagree. To the extent that this commenter suggests that the FCC should issue roadway licenses between EAs, we are not aware of geographic areas that exist between EAs. More generally, we believe that EA, rather than roadway, licenses will lead to more widespread service to consumers in this band. Further, we believe the public interest lies in covering as much area as possible given the economics of the band. In many cases, even in very rural areas, this may extend beyond roadways.

3. Licensing the Gulf of Mexico

43. In the

H Block NPRM,

the Commission sought comment on whether and, if so, how to license the Gulf of Mexico. The Commission sought comment on whether the Gulf should be included as part of larger service areas, or whether the Gulf should be licensed separately.

44. We will license the H Block for the Gulf of Mexico. We find it appropriate to follow Commission precedent from the AWS-1 and AWS-4 bands, both of which licensed the Gulf as a separate EA license. Moreover, the only party who commented on this issue supports the proposal to make available an EA license for the Gulf. Finally, we determine to apply the existing definition of the Gulf of Mexico EA contained in section 27.6 of the Commission rules when licensing the Gulf. Specifically, the Gulf of Mexico service area is comprised of the water area of the Gulf of Mexico starting 12 nautical miles from the U.S. Gulf coast and extending outward.

C. Technical Issues

45. Pursuant to the statutory direction of the Communications Act of 1934, as amended, the Commission adopts rules for commercial spectrum in a manner that furthers and maximizes the public interest. Notably, when developing policies for a particular band, the Commission looks at other bands that might be affected, particularly the adjacent bands. Consequently, the Commission must often balance competing interests of adjacent bands, and potentially competing public interest considerations, when crafting rules. Because the rules for one band, particularly the interference protection rules, affect the use and value of other bands and thus the public interest benefits that can be realized through the use of those adjacent bands, we take a holistic view when establishing the technical rules for each spectrum band.

46. In this section, we adopt the technical operating rules (

e.g.,

interference rules) that will govern H Block operations and licensees. In general, our aim in establishing technical rules is to maximize the flexible use of spectrum while appropriately protecting operations in neighboring bands. Here, we also specifically consider our statutory obligations set forth in the Spectrum Act with respect to the 1930-1995 MHz broadband PCS band, which specifically requires us to determine whether either of the H Block bands “cannot be used without causing harmful interference to commercial mobile service licensees in the [1930-1995 MHz PCS band].”

47. We base the technical rules we adopt below on the rules for the AWS and PCS spectrum bands, which have similar characteristics to the H Block and that we therefore expect would permit optimal use of the H Block by its licensees. In applying these rules to the H Block, we specifically adopt rules to adequately protect operations in adjacent bands, including the existing 1930-1995 MHz broadband PCS downlink band and the 2000-2020 MHz AWS-4 uplink band. Finally, given the record before us and the analyses provided below, we conclude that the benefits of the technical rules we adopt herein likely outweigh any potential costs.

1. Upper H Block: 1995-2000 MHz

48. The Upper H Block is immediately above the 1930-1995 MHz PCS band, which is subject to the Spectrum Act's harmful interference provision. The PCS band currently is used for base station transmit/mobile receive (

i.e.,

downlink) purposes. In the

H Block NPRM,

the Commission tentatively concluded that operating base stations in 1995-2000 MHz would be compatible with similar use of the spectrum in the 1930-1995 MHz band, and that more restrictive technical standards than those established for other AWS stations in similar bands would be unnecessary to protect the PCS band from harmful interference. No technical concerns were raised in the record about interference between the Upper H Block and PCS base stations operating below 1995 MHz. As stated above, the 1995-2000 MHz Upper H band will serve as

downlink spectrum and is thus compatible with adjacent downlink operations below the band.

49. The Upper H Block is also situated immediately below the 2000-2020 MHz band, which is allocated on a co-primary basis for Fixed, Mobile, and Mobile Satellite (Earth-to-space,

i.e.,

for mobile transmit/satellite or base station receive), and is licensed for both Mobile Satellite Service (MSS) and AWS-4 terrestrial wireless services. The Commission recently adopted service rules that permit use of the 2000-2020 MHz band for terrestrial mobile-to-base (uplink) transmissions. In so doing, the Commission concluded that certain protections were needed to avoid harmful interference between the Upper H Block and 2000-2020 MHz band. Having weighed various public interest considerations, the Commission imposed certain limited power restrictions and out-of-band emission (OOBE) limits on AWS-4 uplinks to preserve the capability for full flexible use of the Upper H Block. Additionally, the Commission concluded that 2 GHz MSS operators and AWS-4 licensees must accept harmful interference from future, lawful operations in the Upper H Block due to either Upper H Block OOBEs into the 2000-2005 MHz portion of the AWS-4 uplink band or to Upper H Block in-band power (receiver overload) into the AWS-4 uplink band. DISH Network Corp.'s (DISH) AWS-4 and 2 GHz MSS subsidiaries accepted the

Order of Proposed Modification,

which accompanied the

AWS-4 Report and Order

and which, thus, included these requirements. Commission staff subsequently issued an

Order of Modification

and issued modified licenses. Nothing in our discussion below is intended to revisit these determinations.

a. Upper H Block Power Limits

50. We adopt transmitter power limits for the Upper H Block that will maximize the full flexible use of the spectrum while ensuring against harmful interference to adjacent PCS operations and, in the case of the AWS-4 band, adequately protecting adjacent operations due to receiver overload. Receiver overload may result when signals outside of the receiver's nominal bandwidth cause the receiver to experience an increased noise level or produce non-linear responses. In setting power limits, we balance the power necessary to ensure successful communication in the band against the level of interference that adjacent services can tolerate based on their operational needs and the public interests served. In doing so here, we ensure against harmful interference to the adjacent PCS band and, in the case of the adjacent AWS-4 band, set a power limit necessary to ensure successful communication by H Block licensees based on the public interest balancing the Commission established in the

AWS-4 Report and Order.

51. In the

H Block NPRM,

the Commission proposed and sought comment on adopting the standard base station power limits applicable to AWS and PCS stations. These power limits are 1640 watts equivalent isotropically radiated power (EIRP) for emissions with less than a 1 MHz channel bandwidth and 1640 watts/MHz for emissions greater than 1 MHz in non-rural areas. In rural areas,

i.e.,

counties with population densities of 100 persons or fewer per square mile, the power limits are 3280 watts EIRP for emissions with less than a 1 MHz channel bandwidth and 3280 watts/MHz EIRP for emissions greater than 1 MHz. The AWS and PCS rules also require providers operating in excess of the 1640 watts/1640 watts/MHz EIRP to coordinate with adjacent block licensees within 120 km. Except as detailed below, commenters generally supported these proposed power limits.

52. For H Block operations in the 1995-2000 MHz band, we adopt a power limit for operations in non-rural areas of 1640 watts EIRP for emissions less than 1 MHz and 1640 watts/MHz for emissions greater than 1 MHz. We adopt a power limit for operations in rural areas of 3280 watts EIRP for emissions less than 1 MHz and 3280 watts/MHz for emissions greater than 1 MHz. For purposes of this rule, a rural area refers to a county with a population density of 100 persons or fewer per square mile. Further, we allow operations in excess of the EIRP of 1640 watts and 1640 watts/MHz limits after coordination with adjacent PCS G Block licensees within 120 km, as is allowed for similar operations in the AWS and PCS services. We adopt these power limits because they are the same as those for base stations in other AWS services, including AWS-1 services and the recently adopted limits for AWS-4 base stations and substantially the same as for PCS base stations. Most parties that commented on this issue supported adopting these power limits. As both Sprint and U.S. Cellular observed, the Commission has consistently proposed and adopted these power limits for other services. Additionally, Sprint commented that such power levels will provide adequate protection for PCS licensees in neighboring spectrum bands. No party claimed otherwise. Based on the record and our prior experience with similar services, we conclude that these power limits are consistent with the Spectrum Act's requirement for avoiding harmful interference to the adjacent PCS band. Further, because these limits reflect established measures of efficient use of spectrum for similar services in other bands, we believe they are consistent with the goals of ensuring full, robust, commercial service for mobile broadband, as set forth in the

AWS-4 Report and Order.

53. In adopting these power limits for H Block base stations, we acknowledge that wording in the

H Block NPRM

may have led to confusion on the part of one commenter (DISH). In the

H Block NPRM,

the Commission specifically “propose[d] to adopt the standard base station power limits that apply to AWS and PCS stations,” but did not include the power density limit for emissions greater than 1 megahertz in summarizing the existing rules, despite the fact that the Commission's AWS and PCS rules explicitly include such limits. In supporting the proposed power limits, Sprint correctly referenced “standard power limits of 1640 watts/MHz for non-rural areas and 3280 watts/MHz for non-rural areas.” In its Reply, DISH claimed that the Commission intended for the Upper H Block power to be measured across the entire 5 megahertz of the band, and that Sprint was improperly seeking to measure the power across one megahertz, thereby increasing the radiated power by 7 dB within the Upper H Block. We disagree. The Commission's intent was to propose Upper H Block power limits that would be measured across one megahertz (for emissions greater than one megahertz). In any event, we now determine to measure power limits in a manner consistent with the PCS and AWS bands. Accordingly, we now adopt the standard AWS base station power limits, as described above, based on the record presented in response to the

H Block NPRM.

54. Further, to the extent DISH may be arguing for lower power limits than those in other AWS bands and the PCS band, its argument is unsupported and misplaced. DISH's statement that some existing PCS equipment (we are not aware of equipment presently existing for the H Block band) may operate at lower maximum power levels is not in and of itself dispositive of the appropriate maximum permissible power levels. Rather, this argument appears simply to present an example of PCS equipment operating well within the applicable PCS rules.

55. We also reject DISH's argument that symmetrical power reductions for

the H Block are necessary. DISH suggested that, should the Commission determine that (1) full-power operations of the Lower H Block would cause harmful interference into the PCS band and, (2) it is necessary to mediate this effect by reducing the power limits of the mobiles transmitting in the Lower H Block, then the Commission should adopt similarly reduced power limits for the Upper H Block (1995-2000 MHz). DISH explained that, “[f]or instance, if the Commission decides to limit the H Block uplink transmit power across 1915-1920 MHz to 13 dBm, as opposed to the typical [3rd Generation Partnership Protect] 3GPP power level of 23 dBm, then the base station radiated power should accordingly be reduced by 10 dB to 164 Watts, as opposed to the Commission's proposal of 1640 watts.” Because, as explained below, we do not reduce the permissible power levels for mobile devices in the Lower H Block below the 23 dBm level discussed by DISH, we dismiss as moot DISH's argument to apply symmetrical power restrictions both to the lower and upper bands.

56. In sum, we adopt a power limit of 1640 watts EIRP for emissions with less than 1 MHz channel bandwidth and 1640 watts/MHz for emissions greater than 1 MHz in non-rural areas and of 3280 watts EIRP for emissions with less than a 1 MHz channel bandwidth and 3280 watts/MHz EIRP for emissions greater than 1 MHz in rural areas as sufficient to protect PCS licensees in the 1930-1995 MHz band from harmful interference and to adequately protect AWS uplink operations, while enabling H Block licensees to operate full power base stations. Further, we allow operations in excess of the EIRP of 1640 watts and 1640 watts/MHz limits after coordination with adjacent PCS G Block licensees within 120 km, as is allowed for similar operations in the AWS and PCS services.

b. Upper H Block Out-of-Band Emissions Limits

57. To minimize or eliminate harmful interference between adjacent spectrum blocks, the Commission's rules generally limit the amount of radio frequency (“RF”) power that may be emitted outside of, or in a range of frequencies outside of, the assigned block of an RF transmission. In both the PCS and AWS-1 bands, for example, the Commission established an OOBE limit that requires emissions outside a licensee's assigned spectrum block be attenuated by a level of at least 43 + 10 log

10

(P) dB, where P is the transmit power in watts.

58. To protect operations in adjacent and nearby bands above and below the Upper H Block, the Commission proposed, and sought comment on (including on the associated costs and benefits), a general OOBE limit for H Block base stations of 43 + 10 log

10

(P) dB, where P is the transmit power in watts, outside of the 1995-2000 MHz band. This is consistent with the OOBE limits of the adjacent PCS operations within the 1930-1995 MHz band. In addition to this general limit, the Commission proposed that H Block operations meet a more stringent OOBE limit of 70 + 10 log

10

(P) dB, where (P) is the transmitter power in watts, between 2005 MHz and 2020 MHz to provide interference mitigation to AWS-4 terrestrial uplink operations. As the Commission observed, this additional proposed interference protection is meant to ensure that all of the Upper H Block spectrum can be used for downlink operations, while affording additional protections to most of the AWS-4 uplink band. Commenters generally supported the proposed OOBE limits into the 1930-1995 MHz PCS band, but several commenters proposed alternative OOBE limits for emissions above 2000 MHz. Although a few commenters made general assertions regarding the costs of adopting certain OOBE limits, no party submitted any cost or benefit data.

59. For the reasons discussed below, except as otherwise specified, we adopt the proposed OOBE limit of 43 + 10 log

10

(P) dB, where (P) is the transmitter power in watts, for Upper H Block base station transmissions outside of 1995-2000 MHz, including into the 1930-1995 MHz and 2000-2005 MHz bands. We also establish an OOBE limit of 70 + 10 log

10

(P) dB, where (P) is the transmitter power in watts, for transmissions from the Upper H Block into the 2005-2020 MHz AWS-4 band. We find that this approach both protects the 1930-1995 MHz band and the 2005-2020 MHz portion of the AWS-4 band from harmful interference, and provides adequate protection to the adjacent, lowest five megahertz of the AWS-4 band at 2000-2005 MHz. Thus, these OOBE limits allow us to meet the requirements set forth in the Spectrum Act with regard to the PCS downlink band, and to best manage the use of these spectrum bands in the public interest, consistent with the balancing we established in the AWS-4 proceeding. Further, as detailed below, our evaluation of the record and our consideration of how best to serve the public interest demonstrate that the various alternative proposals for OOBE limits put forth by commenters do not sufficiently balance the use of the H Block and use of the neighboring spectrum bands.

60.

General OOBE Limit.

We adopt an OOBE limit of 43 + 10 log

10

(P) dB, where (P) is the transmitter power in watts, for Upper H Block transmissions outside of the 1995-2000 MHz band, except as described below. We anticipate that H Block systems will be similar in design to PCS and AWS-1, which have effectively relied on the 43 + 10 log

10

(P) dB OOBE limit in the Commission's rules to prevent harmful interference to operations in adjacent and nearby bands. The record also contains support for this OOBE limit. We therefore adopt an OOBE limit of 43 + 10 log

10

(P) dB, where (P) is the transmitter power in watts, for transmitters operating in the Upper H Block, except as detailed below.

61.

Emissions into PCS.

We adopt and apply the general OOBE limit of 43 + 10 log

10

(P) dB, where (P) is the transmitter power in watts, for Upper H Block transmissions into 1930-1995 MHz. The record demonstrates support for our decision as commenters support the proposed 43 + 10 log

10

(P) dB for base station transmissions from the 1995-2000 MHz band into the PCS bands located in 1930-1995 MHz. For example, U.S. Cellular and Sprint support an OOBE limit of 43 + 10 log

10

(P) dB as the emissions restriction imposed on operations in the 1995-2000 MHz band. With respect to emissions into PCS, no party has opposed this limit. Moreover, inasmuch as the Upper H Block can be viewed from a technical perspective as an extension of the 1930-1995 MHz PCS band because they are both adjacent downlink bands, the 43 + 10 log

10

(P) dB OOBE limit that applies between adjacent PCS downlink blocks logically should also apply to Upper H Block emissions into the 1930-1995 MHz PCS bands. Thus, to protect PCS operations in the 1930-1995 MHz band from harmful interference, we adopt an OOBE limit of 43 + 10 log

10

(P) dB for Upper H Block base transmissions.

62.

Emissions into AWS-4.

We adopt an OOBE limit of 43 + 10 log

10

(P) dB, where (P) is the transmitter power in watts, for Upper H Block transmissions into 2000-2005 MHz and an OOBE limit of 70 + 10 log

10

(P) dB, where (P) is the transmitter power in watts, for Upper H Block transmissions into 2005-2020 MHz. We find these limits appropriately balance the difficult technical challenges associated with the Upper H Block (

i.e.,

downlink) being adjacent to the 2000-2020 MHz AWS-4 band (

i.e.,

uplink), which the Commission

addressed in the

AWS-4 Report and Order.

As the Commission previously observed, uplink spectrum bands that are adjacent to downlink spectrum bands raise difficult interference issues that require balancing the needs of both bands. In striking this balance, the Commission must determine what technical limits are appropriate, because the rules for one band affect the use and value of other bands, and the Commission seeks to maximize the efficient use of all bands. In the AWS-4 proceeding, for example, the Commission weighed the potential interference issues between the 2000-2020 MHz AWS-4 band and the 1995-2000 MHz H Block band. The Commission's assessment concluded that, to protect the utility of the Upper H Block, (1) AWS-4 uplink operations must meet a relatively strict OOBE limit of 70 + 10 log

10

(P) dB into the 1995-2000 MHz band and into the 1930-1995 MHz PCS band, and (2) AWS-4 and 2 GHz MSS licensees would be required to accept harmful interference from lawful operations in the 1995-2000 MHz band if such interference is due to OOBE into the 2000-2005 MHz band or due to receiver overload into the 2000-2020 MHz band. In now establishing the technical rules for the Upper H Block, it is appropriate to likewise recognize the impact operations in this band may have on licensees above 2000 MHz.

63. In assessing the needs of both Upper H Block and AWS-4 uplink band, we start from an understanding of the current interference environment. Under the Commission's rules, emissions from the PCS downlink band at 1930-1995 MHz, including the G Block (1990-1995 MHz), into the AWS-4 uplink band at 2000-2020 MHz are limited to 43 + 10 log

10

(P) dB, where (P) is the transmitter power in watts. Our rules, however, are not the only factors affecting the operation and performance of AWS-4 systems. Both Sprint and DISH cite the 3GPP standards to support their differing cases for the OOBE limit into the AWS-4 band. These standards allow for an OOBE limit of −30 dBm/MHz (equivalent to attenuation of 60 + 10 log

10

(P) dB) into the 2000-2010 MHz band, dropping to −49 dBm/MHz (equivalent to 79 + 10 log

10

(P) dB) in the 2010-2020 MHz band. Additionally, the 3GPP standard noted that OOBE limits would only apply 5 MHz or farther from the edge of the PCS base station's operating band. This allows 5 megahertz within which the transmitter's output can roll off to meet the tighter limits.

64. Sprint (which holds all of the licenses for the PCS G Block, as well as some licenses for other PCS blocks) advocated for a limit of 60 + 10 log

10

(P) dB across the 2005-2020 MHz band and DISH (which holds all of the AWS-4 licenses) advocated for a more stringent 79 + 10 log

10

(P) dB limit across the 2005-2020 MHz band. In other words, relatively speaking, DISH would prefer that we impose greater restrictions on the transmissions from the Upper H Block into the AWS-4 band, while Sprint would prefer lesser restrictions on those Upper H Block transmissions. Both Sprint and DISH cite 3GPP standards in arguing for their preferred OOBE limits. Historically, while the Commission may take into consideration the determinations of third party technical standards organizations, such as 3GPP, the Commission also considers other factors not relevant to standards organizations. For instance, the Commission necessarily takes into account its enabling, and any other relevant, statute, which would not be binding on a third party standards organization. We are required, for example, to manage spectrum in the public interest, and to “generally encourage the larger and more effective use of radio in the public interest.” Private standards bodies may have other bases for their determinations, which may reflect compromises among the participants that are not subject to the statutory mandates that must inform our actions. Accordingly, while the Commission may independently incorporate industry standards based on the particular record before it, it does not typically adopt such interference standards as Commission rules. We again decline to do so here. Further, inasmuch as the OOBE limit we establish herein represents a ceiling, not a floor, industry remains free to set a more restrictive value through technical standards bodies, such as 3GPP.

65. In maximizing the usefulness of both bands, we seek to set appropriate limits on OOBE such that the overall interference imposed on AWS-4 uplink operations is no more than currently exists, to the greatest extent possible, without imposing a harsh and undue burden on Upper H Block downlink operations. We therefore adopt an OOBE limit of 43 + 10 log

10

(P) dB, where (P) is the transmitter power in watts, for all Upper H Block emissions above 2000 MHz, including the 2000-2005 MHz portion of the AWS-4 band, except for transmissions into 2005-2020 MHz. As discussed above, this emission limit (10 log

10

) is the same level of protection that the Commission's rules currently provide AWS-4 operations from transmissions from existing PCS downlink operations in the 1930-1995 MHz band. For Upper H Block transmissions into 2005-2020 MHz, we adopt a more stringent OOBE limit of 70 + 10 log

10

(P) dB, where (P) is the transmitter power in watts. This layered approach, encompassing one set of interference standards for emissions into the first five megahertz and a more stringent limit on emissions into the remaining fifteen megahertz, provides some flexibility for the H Block operator to design the emission characteristics of its system to meet the tougher OOBE limits into the 2005-2020 MHz band. This approach, moreover, was contemplated by the Commission in the

AWS-4 Report and Order

where the Commission, in requiring AWS-4 licensees to accept certain interference in the AWS-4 uplink band, stated that “base station transmit filters need 1 to 5 megahertz to roll off to a low level of emissions.” In addition, under the 3GPP standards, out-of-band emissions from PCS LTE operations must satisfy an OOBE limit of 60 + 10 log

10

(P) dB at 2000-2010 MHz and then transition sharply to satisfy a much stricter limit of 79 + 10 log

10

(P) dB at 2010-2020 MHz. As a practical matter, however, out-of-band emissions tend to roll off smoothly and do not mimic the step functions of the limits set by standards bodies, such as 3GPP. As a result, the emissions from LTE operations in the PCS band will naturally decrease smoothly from the 60 + 10 log

10

(P) dB level at 2000 MHz to the 79 + 10 log

10

(P) dB from 2010-2020 MHz. The limit we set at 2005 MHz—70 + 10 log

10

(P) dB—approximates the emissions level that we expect would arise at 2005 MHz as emissions roll off between 2000 MHz and 2010 MHz. Therefore, we expect that the overall harmful interference risk on the AWS-4 A Block operator from future H Block operators would be no more than exists today from existing PCS operators. That is, just as PCS operations are not expected to cause harmful OOBE interference at 2005-2020 MHz, nor are H Block operations expected to cause OOBE interference at the limit we set here.

66. In response to the Commission's proposed OOBE limits into the AWS-4 uplink band, parties commented that the proposed limits were both too lenient and too strict. DISH argued that 43 + 10 log

10

(P) dB is insufficient to protect AWS-4 and 2 GHz MSS operations in 2000-2005 MHz and that 70 + 10 log

10

(P) dB is insufficient protection for operations in 2005-2010 MHz. Rather, DISH suggested a three-fold approach to protect AWS-4/2 GHz

MSS operations. DISH proposed an OOBE limit of 55 + 10 log

10

(P) dB for emissions in the 2000-2005 MHz band, an OOBE limit of 79 + 10 log

10

(P) dB for emissions above 2005 MHz, and an OOBE limit of 116 + 10 log

10

(P) dB for co-located sites. Conversely, Sprint opposed the

H Block NPRM's

proposal of 70 + 10 log

10

(P) dB above 2005 MHz as imposing too stringent a restriction on Upper H Block transmissions and recommended an OOBE limit of 60 + 10 log

10

(P) dB into and above 2005.

67. We reject both proposals as improperly balanced, with the DISH proposal overly burdensome for a full powered, flexible use H Block and the Sprint proposal too burdensome on AWS-4 operations and unnecessary to allow the Upper H Block licensees full, flexible use of that spectrum.

68. First, we reject DISH's proposal that Upper H Block operations be restricted to an OOBE limit of 55 + 10 log

10

(P) dB between 2000 and 2005 MHz. As discussed above, we establish an OOBE limit of 43 + 10 log

10

(P) dB between 2000 and 2005 MHz and believe this represents an appropriate balance between ensuring the utility of the Upper H Block and the AWS-4 uplink band. A level of 55, rather than 43, plus 10 log

10

(P) dB would be 32 times more stringent and would thus restrain the full use of the H Block. DISH argues that this OOBE level is necessary because aggregate power from all H Block base stations in the direction of the satellite would inadequately protect the satellite. We agree with Sprint and U.S. Cellular that DISH's argument is an inappropriate collateral attack on the

AWS-4 Report and Order

and our related order modifying the licenses of DISH's subsidiaries, which they have accepted. The Commission explicitly addressed the issue of how to balance Upper H Block interference into the 2000-2005 MHz band, for both terrestrial and MSS operations, in the

AWS-4 Report and Order.

There the Commission stated:

to the extent that future operations in the 1995-2000 MHz band, operating within the rules established for use of the 1995-2000 MHz band, cause harmful interference to AWS-4 operations or MSS operations due to . . . OOBE in the 2000-2005 MHz portion of the AWS-4 and 2 GHz MSS uplink band . . . AWS-4 and 2 GHz MSS licenses must accept this interference.

We therefore reject DISH's proposed OOBE limit of 55 + 10 log

10

(P) dB between 2000 and 2005 MHz because it conflicts with the full potential use of the H Block and would be inconsistent with the

AWS-4 Report and Order.

69. Second, we reject DISH's proposal for an OOBE limit of 79 + 10 log

10

(P) dB at and above 2005 MHz. DISH argued this limit is needed to protect AWS-4 terrestrial operations in 2005-2020 MHz. We disagree. We find that some of the assumptions underlying DISH's analysis are overly conservative, such as the use of a one kilometer spacing between base stations in both the interfering system and the victim system in determining the minimum coupling loss (MCL). As a result, we find an OOBE limit of 79 + 10 log

10

(P) dB at 2005 MHz to be too restrictive on Upper H Block operations. While DISH has asserted that meeting an OOBE limit more stringent than 43 + 10 log

10

(P) dB would not be difficult for the H Block operator to meet, the evidence it cites does not support the conclusion that an H Block operator could meet an OOBE limit of 79 + 10 log

10

(P) dB at 2005 MHz. In the three test reports cited by DISH, each LTE base station is shown to exceed the Commission's limit of 43 + 10 log

10

(P) dB by 10 dB or more. For instance, the Samsung test report shows that the base station may be able to meet 60 + 10 log

10

(P) dB within the AWS-4 band. However, none of the test results show whether the base stations would be able to meet DISH's proposed limit of 79 + 10 log

10

(P) dB. In addition, we find that an OOBE limit of 70 + 10 log

10

(P) dB, as opposed to a limit of 79 + 10 log

10

(P) dB, is more consistent with the balancing of interference concerns between the AWS-4 and H Block bands discussed in the

AWS-4 Report and Order,

particularly in light of the Commission's determination in that order to require AWS-4 operations to protect future Upper H block operations using an OOBE limit of 70 + 10 log

10

(P) dB. Thus, to avoid harmful OOBE interference to AWS-4 operations at 2005-2020 MHz, we find an OOBE limit of 70 + 10 log

10

(P) dB into 2005-2020 MHz is necessary.

70. DISH further argued that an OOBE limit of 79 + 10 log

10

(P) dB at 2005 MHz is consistent with 3GPP specifications. As an initial matter, as we stated above, while the Commission may take into consideration the determinations of third party technical standards organizations such as 3GPP, the Commission also considers other factors not relevant to standards organizations. Moreover, we observe that, while the DISH proposed OOBE limit is contained in the 3GPP specification for LTE base stations, the limit is for bands other than Bands 23, 2, and 25. Bands 23, 2, and 25 represent the AWS-4 operations, PCS operations in the 1930-1990 MHz band, and PCS + G Block operations in the 1930-1995 MHz band, respectively. Thus, the 3GPP specification, on its own terms, does not apply to the interference scenario at issue here. There is a separate set of OOBE limits that apply to these nearby bands. Notably, the relevant 3GPP specification for Band 25 only requires 60 + 10 log

10

(P) dB between 2000 and 2010 MHz due to its proximity to the AWS-4 band. 3GPP does not require PCS operations to meet the more stringent 79 + 10 log

10

(P) dB limit until at least 15 MHz above the PCS band (

i.e.,

above 2010 MHz). Thus, DISH's suggestion that 3GPP standards provide an example of more stringent OOBE limits is misplaced. We also observe that, as Sprint asserted, current Commission rules allow for much lower attenuation for existing PCS systems, including the G Block, over the entire AWS-4 band.

71. Third, we reject DISH's proposed OOBE limit for co-located sites. Specifically, DISH sought an OOBE limit of at least 116 + 10 log

10

(P) dB for sites containing both an AWS-4 base station and an H Block base station. DISH argued, “when two base stations are co-located, significantly less path loss is encountered, and a much higher interference level may be present at the victim receiver,” which requires more stringent filters. DISH cited a 3GPP LTE standard recommendation for co-location that stated a limit of −96 dBm/100 kHz may be applied for the protection of other base station receivers. Co-location with other communication systems is a common industry practice to resolve coexistence issues. Yet the Commission typically does not impose separate OOBE requirements on co-located sites in other systems operating under either part 24 or part 27. Instead, these interference concerns are routinely negotiated between the affected parties, taking advantage of the flexibility afforded by our rules for affected parties to resolve interference issues at spectral and geographic boundaries. Because co-location is a network design decision, network operators possess incentives to deploy in an efficient and productive manner that minimizes potential harmful interference. In some cases, interference scenarios can be improved through the use of co-location. Additionally, our rules contain a savings provision. In the case that harmful interference results from OOBE, the Commission may, at its discretion, require greater attenuation than the specified limits. Furthermore, while not dispositive of our regulatory determination, the 3GPP standards DISH references specifically exempt base station transmitters operating within 10 megahertz of the affected

receiver's operating band, which is the case here. Indeed, the standard itself states that “the current state-of-the-art technology does not allow a single generic solution for co-location with other systems” and points to site engineering solutions. In sum, we find that to impose a limit of 116 + 10 log

10

(P) on the Upper H Block would be unduly burdensome on the licensee and that setting any OOBE for the specific case of co-location would be inconsistent with general Commission practice. Therefore, we decline to establish a rule pertaining to co-location interference issues.

72. We also reject Sprint's proposal to adopt a 60 + 10 log

10

(P) dB attenuation requirement from 2005-2020 MHz. Sprint argued an OOBE limit of 70 + 10 log

10

(P) dB would significantly increase the cost of deployment in the Upper H Block, but made no attempt to quantify this cost or provide any cost data. According to Sprint, such increases in costs both could stifle interest in an auction of the H Block and would not provide any substantive improvement in interference. This argument is inconsistent with Sprint's agreement in the 3GPP standards process to protect operations in the 2010-2020 MHz band at a level of 79 + 10 log

10

(P) dB. In addition, DISH asserted that an OOBE limit of 60 + 10 log

10

(P) dB is insufficient to protect AWS-4 operations. We agree with DISH. In this instance, a stricter OOBE limit is warranted because the Upper H Block (downlink) is adjacent to the AWS-4/2 GHz MSS uplink band, which raises real interference concerns. An OOBE limit of 70 + 10 log

10

(P) dB, as opposed to a limit of 60 + 10 log

10

(P) dB, is more consistent with the balancing of interference concerns between the AWS-4 and H Block bands discussed in the

AWS-4 Report and Order,

particularly in light of the Commission's determination in that order to require AWS-4 operations to protect future Upper H block operations using an OOBE limit of 70 + 10 log

10

(P) dB. Thus, to avoid harmful OOBE interference to AWS-4 operations at 2005-2020 MHz, we find an OOBE limit of 70 + 10 log

10

(P) dB into 2005-2020 MHz is necessary.

73.

Measurement Procedure.

Finally, to fully define an emissions limit, the Commission's rules generally specify details of the measurement procedure to determine the power of the emissions, such as the measurement bandwidth. For AWS-1, for example, the measurement bandwidth used to determine compliance with this limit for both mobile stations and base stations is generally 1 megahertz, with some modification within the first 1 MHz. The Commission also applied the same OOBE measurement procedure to AWS-4 and to PCS operations. To treat the Upper H Block in an equivalent manner to these similar bands, we therefore adopt the same requirement that compliance with the emissions limits established herein will be determined by using a 1 MHz measurement bandwidth.

c. Co-Channel Interference Between Licensees Operating in Adjacent Regions

74. As discussed above, we determine to license the H Block on an EA geographic license area basis. The Commission observed in the

H Block NPRM

that should the H Block be licensed on a less than nationwide basis, it would be necessary to ensure that licensees do not cause harmful interference to co-channel systems operating along their common geographic boundaries. To resolve any such interference, the Commission proposed adopting a boundary limit approach, with a specific boundary field strength limit of 47 dBµV/m. The Commission also sought comment on whether licensees operating in adjoining areas should be permitted to employ alternative, agreed-upon signal limits at their common borders. With one exception, commenters did not oppose the Commission's proposals to protect adjacent licensees from co-channel interference. Sprint, however, argued that the field strength limit be adjusted to accommodate for varying channel bandwidths.

75. We adopt the proposed boundary limit approach for co-channel interference. As discussed above, the Commission will license the H Block on a geographic area basis that is less than nationwide,

i.e.,

an EA basis. To prevent licensees that operate systems along common geographic borders from causing harmful interference to one another, the Commission must provide operating limits to ensure such licensees do not cause interference to co-channel systems. Adopting a boundary limit approach establishes a default standard, which will enable licensees to deploy facilities in boundary areas without the need for prior coordination. Licensees may use this operating limit as a starting point for negotiations to exceed the limits with agreement of adjacent area licensees. Moreover, in other bands where spectrum has been allocated for fixed and mobile services, similar to the H Block, the Commission has uniformly adopted the boundary limit method to minimize harmful co-channel interference. For instance, the PCS, AWS-1, and AWS-4 bands all use a boundary limit approach. In response to the Commission's proposal, commenters favored the boundary limit approach over a coordination requirement. For example, Sprint comments that “applying a boundary limit consistent with prior proceedings can enable future H Block licensees to deploy facilities in boundary areas without the delays associated with significant pre-coordination efforts while protecting adjacent licensees from co-channel interference at their borders.” Additionally, no commenter proposed a coordination approach for limiting co-channel systems from interfering with one another. Consequently, we find that a boundary limit approach is the best method to address potential harmful co-channel interference between licensees operating in adjacent geographic regions.

76. We set the field strength limit at the boundary at 47 dBµV/m. As the Commission observed in the

H Block NPRM,

in other bands where spectrum has been allocated for fixed and mobile services and licensed for flexible use, similar to the H Block, the Commission has generally adopted a boundary field strength limit of 47 dBµV/m. For example, in the PCS, AWS-1, and AWS-4 bands, the Commission adopted a field strength limit of 47 dBµV/m at the boundary of licensed geographic areas. Because this limit has worked well in limiting co-channel interference in other bands, we find it appropriate to adopt it here for the similarly situated Upper H Block.

77. In adopting this boundary limit, we decline to adopt the alternative limit proposed by Sprint in its Reply. While supporting the boundary limit approach used in other bands, Sprint asserted that we should modify the boundary limit to set a reference measurement bandwidth. In making this recommendation, Sprint claimed that because today's LTE transmissions operate on wider channels than earlier technologies such as CDMA or Digital AMPS, a 47 dBµV/m limit will effectively result in a comparatively lower field strength limit. Specifically, Sprint proposed to adjust the field strength limit from 47 dBµV/m to 62 dBµV/m per MHz. Sprint argued that the power spectral density for a 30 kHz Digital AMPS carrier at a 47 dBµV/m field strength is equivalent to a 62 dBµV/m LTE carrier with a 1 MHz bandwidth, adjusting the field strength limit by the ratio of the bandwidths (10*log

10

(1 MHz/30 kHz) = 15 dB). Sprint stated that its proposed boundary limit would better enable 4G-LTE buildout of the H Block while also providing the appropriate interference

protections. Sprint further suggested that the boundary limits with Canada and Mexico should similarly be based on power density levels.

78. Although we agree with Sprint on a conceptual level that a boundary limit that adjusts for large differences in channel bandwidths may be appropriate, we are not persuaded that Sprint's proposed limit represents the optimal solution. Sprint derived the value for the field strength based on a comparison against a 30 kHz Digital Amps signal. Other technologies may be a more appropriate reference upon which to base the value for the field strength. Also, there are other metrics that may be used to limit the signal at the boundary, such as power flux density. We observe that the Commission has already adopted a bandwidth-independent approach when setting boundary limits with Canada and Mexico. For example, certain international limits are expressed as a power flux density (

i.e.,

dBW/m

2

/MHz), a measure of power, whereas field strength is a measurement of voltage. As Sprint noted, other parties have proposed to set boundary limits in a bandwidth neutral manner, but there is no established consensus on what the value of the limit should be. With no consensus regarding an alternative boundary limit approach, and not having received record input from any other party on Sprint's proposal, we are not prepared to adopt it at this time. We intend to explore the issue of whether to apply a measurement bandwidth to co-channel boundary limits in future service rules proceedings and we encourage all interested parties to explore this issue in such proceedings to develop a full record of the technical concerns and ramifications of such an approach.

79. Finally, we adopt the Commission's proposal that adjacent affected area licensees may voluntarily agree upon higher field strength boundary levels that the 47 dBµV/m we adopt above. This concept is already codified in the field strength rules for both PCS and AWS services, as Sprint acknowledged. No party opposed extending this approach to the H Block. Accordingly, to maintain consistency with the PCS and AWS bands, we permit adjacent area licensees to agree to a higher field strength limit.

2. Lower H Block: 1915-1920 MHz

80. The Lower H Block is immediately above the 1850-1915 MHz PCS band, which is used for mobile transmit/base receive (

i.e.,

uplink) purposes. As the Commission observed, use of the Lower H Block as proposed in the

H Block NPRM

is compatible with this adjacent PCS band. Accordingly, the Commission stated that technical standards more restrictive than those already established for AWS and PCS stations to protect PCS operations below 1915 MHz likely would not be necessary.

81. The Lower H Block is also situated immediately below the 1920-1930 MHz band, which is allocated for Unlicensed PCS purposes (UPCS) and the 1930-1995 MHz PCS base transmit/mobile receive (

i.e.,

downlink) band. As explained below, UPCS operations are not entitled to interference protection from appropriately licensed operators in the Lower H Block. The 1930-1995 MHz band, however, presents technical challenges for use of the Lower H Block. As detailed below, when certain worst-case conditions are present, the potential exists for mobile transmitters in the 1915-1920 MHz band to cause harmful interference to mobile receivers in the 1930-1995 MHz band.

82. As discussed above, the Spectrum Act requires the Commission to conduct an auction of the H Block spectrum unless we determine that the H Block frequencies cannot be used without causing harmful interference to commercial mobile service licensees operating between 1930-1995 MHz (PCS downlink). Against this backdrop, commenters generally argued that the Commission should carefully examine the issue of mobile power limits for the Lower H Block and that, if possible, these limits should be based on technical studies. Four parties submitted technical reports into the record that address the possibility of Lower H Block operations causing harmful interference to PCS operations in the 1930-1995 MHz band. Sprint filed a test report accompanying its Reply filing. On April 18, 2013, Verizon Wireless submitted a technical study. On May 13, 2013, and May 14, 2013, T-Mobile and AT&T separately filed a joint test report.

83.

Sprint and Verizon Wireless Test Reports.

Both Sprint and Verizon Wireless contracted with V-COMM Telecommunications Engineering (V-COMM) to conduct tests on the effects of mobile operations in the Lower H Block on several of each operator's existing CDMA handsets. The handset's receiver performance was tested against interference due to overload (

i.e.,

blocking), intermodulation, and OOBE.

84.

AT&T and T-Mobile Study.

AT&T and T-Mobile contracted with 7Layers to perform tests on the effects of mobile operations in the Lower H block on several of each operator's existing GSM, UMTS and LTE handsets. The mobile receiver's performance was tested against interference due to overload, intermodulation and OOBE.

85. We discuss these test reports and the interference scenarios they examined more fully below. At the outset, however, we observe that AT&T, Sprint, T-Mobile, and Verizon Wireless all stated that, subject to appropriate power limits and OOBE limits, mobile operations in the Lower H Block can occur without causing harmful interference to PCS operations in the PCS band at 1930-1995 MHz. Based on our analysis of the record, which we explain in detail in the sections immediately below, we agree that appropriate technical rules will ensure that mobile or low power fixed operations in the Lower H Block do not cause harmful interference to PCS downlink operations.

a. Lower H Block Power Limits

86. We adopt transmitter power limits for the Lower H Block that will maximize the full flexible use of the spectrum while protecting adjacent operations from harmful interference due to receiver overload. As explained above, receiver overload may result when signals outside of the receiver's nominal bandwidth cause the receiver to experience an increased noise level or produce non-linear responses. Accordingly, we must examine the power limits necessary to avoid harmful interference to PCS downlink licensees under the Spectrum Act and, within this constraint, maximize full flexible use of the Lower H Block.

87. In the

H Block NPRM,

the Commission observed that parties commenting in earlier dockets had expressed concern regarding power limits for the Lower H Block. These comments argued for the establishment of power limits for operation in the Lower H Block that would adequately protect PCS operations in the 1930-1995 MHz band. As discussed above, since these earlier comments, the mobile broadband industry has undergone rapid evolution and new technologies have been developed and adopted. These advances prompted the Commission to seek comment on how newer filtering techniques and duplex designs have improved to adjust for potential harmful interference. Specifically, the Commission sought comment on an appropriate power limit for 1915-1920 MHz mobile devices in light of these advances.

88. The Commission also observed that the 1915-1920 MHz band is allocated for fixed services, but that the possibility of interference from fixed station antennas to PCS mobiles will likely be less than anticipated

interference from Lower H Block mobiles to PCS mobiles because fixed devices are generally located at a fixed height above the ground and thus are vertically separated from PCS mobile devices. Accordingly, the Commission sought comment on what the power level should be for fixed stations operating in the Lower H Block.

89. The record contains three technical studies that examined the potential for Lower H Block operations to cause harmful interference, including overload, intermodulation and interference from out-of-band emissions, to PCS downlink operations. All of these studies assumed that the Lower H Block device would be an LTE FDD mobile device. The Sprint Test Report and the Verizon Wireless Test Report both used existing CDMA devices for the PCS devices. The AT&T/T-Mobile Study used LTE, UMTS, and GSM PCS devices. The studies included testing of the receiver performance of existing PCS devices against overload interference, as well as intermodulation interference that would be caused, in part, by receiver overload. As stated above, receiver overload occurs when the power from a signal outside of the receiver's operating frequency range causes the receiver's performance to degrade. A strong radio frequency (RF) signal can cause the detector in the receiver to operate in a non-linear manner, thereby reducing its ability to decode the desired signal. Intermodulation interference may occur when two RF frequencies pass through a non-linear element in the receive path of the receiver. Two signals at different frequencies passing through a non-linearity will mix and create new frequencies that are related to the sum and the difference of the original signals. These are termed intermodulation products. Although the non-linearity may be caused by hardware flaws, the most common cause of intermodulation interference—and the historical concern for the bands at issue—is from non-linearity that results from receiver overload. Notably, in earlier tests, third order intermodulation products were found to occur within the PCS mobile receiver's B Block frequency range (1950-1965 MHz) due to the mixing the of the PCS mobile device's transmitter frequency (1870-1885 MHz) with the Lower H Block mobile device's transmitter frequency (1915-1920 MHz). Below, we describe the three tests, first presenting the test set-up for all of the tests, followed by the results for all of the tests.

90.

Sprint and Verizon Wireless Test Reports—Test Setup.

In performing tests for Sprint and for Verizon Wireless, V-COMM tested the performance of a number of each operator's existing CDMA devices against overload and intermodulation interference using the same test procedure. Although both types of interference may be caused by strong power levels, the effects of the interference are seen at different receiver frequencies. The greatest potential for overload occurs where the edge of the receiver's passband is closest to the transmitter's operating frequency range. Therefore, tests for overload were conducted with the receiver tuned to the lowest channel in the PCS A Block, closest to the Lower H Block. The tests for intermodulation were conducted at three different receiver operating frequencies within the PCS B Block downlink band.

91. In the testing, V-COMM subjected each of the PCS CDMA receivers to several different interfering signals, each with different center frequencies, channel bandwidths and types of modulation. The set of interfering signals were 5 MHz, 3 MHz or 1.4 MHz bandwidth LTE carriers, centered at 1917.5 MHz, 1916.5 MHz and 1919 MHz, respectively. The types of modulation used represented several worst case conditions, such as maximizing power at the control channels located near the edges of the band, a fully loaded device with all resource blocks allocated, or all power concentrated in a single resource block located on a frequency where it would be most likely to create intermodulation products.

92. In total, twelve different types of interfering signals were tested for each device. First, the receiver sensitivity of each device was measured to determine the minimum received power level at which the device would perform properly in the absence of noise. Successful operation was defined as a 0.5% Frame Error Rate (FER). The level of the desired signal was set at either 1 dB or 3 dB above the measured sensitivity level. Then an interfering signal was introduced and its power level increased until the same 0.5% Frame Error Rate was achieved, marking the 1 dB or 3 dB receiver desensitization level. The 1 dB or 3 dB desensitization level is the power of the interfering signal at which the receiver's sensitivity is degraded by 1 dB or 3 dB, respectively. For each test case, both the 1 dB receiver desensitization and 3 dB receiver desensitization levels were recorded.

93. V-COMM then related the interference levels measured in each test case to their effect on the user's experience in two scenarios. In so doing, V-COMM determined the power level of the out-of-band emissions at the output of the H Block transmitter necessary to generate the measured interference levels at the PCS receiver's antenna terminals. The difference between these two signal levels is determined primarily by the distance between the transmitting and receiving devices and by the manner in which the user is handling the device, which affects the amount of head and/or body losses in the transmission path. The two user scenarios were: (1) Both the transmitting and receiving mobile devices were assumed to be held in the user's hand, as would be likely for data use; and (2) both the transmitting and receiving mobile devices were assumed to be held to the user's head, as would be likely for a voice call. The analysis then set forth assumptions of 3 dB for body loss, 8 dB for head loss, a 0 dBi receive antenna gain for both mobile devices, a separation of 1 meter, and free space path loss to the two user scenarios. Application of these assumptions determined the effective interfering signal level at the receiver input of −21 dBm and of −31 dBm, respectively, for the data and voice user scenarios. The device was deemed to operate normally if the power level of the interfering signal that caused receiver desensitization exceeded these values.

94.

AT&T/T-Mobile Test Report—Test Setup.

AT&T and T-Mobile developed a joint test plan to test the performance of several of each operator's GSM, UMTS and LTE devices against interference due to receiver overload, intermodulation and out-of-band emissions from an H Block mobile transmitter. The tests were performed by 7Layers, a third party. Because much of the 7Layers testing took place after the filing of the Sprint Test Report, AT&T and T-Mobile included several test cases that subjected their devices to similar conditions to those used by Sprint. The test report, submitted jointly by AT&T and T-Mobile, did not provide details of the test setup used. However it did identify several differences between the 7Layers tests and those performed by V-COMM for Sprint and Verizon Wireless. The most significant difference between the test plans is how the desired signal level was set. The 7Layers tests initially set the level of the desired signal at 3 dB above the reference sensitivity level set by the 3GPP standard for the technology under test. To provide a more direct comparison to the Sprint and Verizon Wireless test reports, however, 7Layers then performed its tests using the sensitivity measured for each device individually, both at 1 dB

above measured sensitivity and again at 3 dB above measured sensitivity. Despite characterizing the set of test conditions using a 1 dB desensitization level as representing worst case scenarios, the AT&T Test Report used this assumption in reaching its conclusions. The AT&T/T-Mobile Test Report did so, while at the same time it raised particular concern about the usefulness of testing to 1 dB of desensitization above each device's measured sensitivity, stating that “it is not typically used during conformance or performance testing, primarily because the measurement uncertainty associated with it is rather high. The measurement metric (throughput or BER/FER) displays highly non-linear behavior.”

95. The AT&T/T-Mobile Test Report is different from the Sprint and Verizon Wireless test reports in other ways, as well. Notably, 7Layers subjected each PCS receiver to two different interfering signals to simulate an H Block mobile device. Both signals represented 5 megahertz LTE carriers operating at a center frequency of 1917.5 MHz, but used different resource block allocations. One signal spread the mobile's power over all 25 resource blocks representing a fully loaded mobile, while the other concentrated the mobile's power in 5 resource blocks, but did not define which five blocks were assigned. By comparison, the Sprint and Verizon Wireless test reports used a total of twelve different LTE signals. Another significant difference in the test plans is that the AT&T/T-Mobile Test Report included for the UMTS PCS devices two desired signal conditions, reflecting both lightly loaded and heavily loaded cell conditions for these devices, whereas the Sprint and Verizon Wireless test reports used one signal condition. The AT&T/T-Mobile used two conditions to simulate “cell breathing” on a CDMA network. In the heavily loaded scenario, the power allocated to each user in the downlink spectrum was reduced and the effective cell coverage was reduced.

96. AT&T and T-Mobile reported results for two GSM devices, up to three UMTS devices (depending on the test scenario), and one LTE device. These results note the power of the interfering signal that would create the specified degradation of the receiver. AT&T and T-Mobile also interpreted the results differently than Sprint and Verizon Wireless, using slightly different assumptions for the user scenario. AT&T and T-Mobile used 25 dBm EIRP as the actual operating power of the H Block mobile, rather than using the nominal 23 dBm EIRP assumed by Verizon Wireless and Sprint. The AT&T/T-Mobile Test Report also did not include any body loss for either the transmitting or receiving mobile. The report therefore used an interfering signal level of −13 dBm as a pass/fail criterion. For point of comparison, Sprint and Verizon Wireless set a −21 dBm criterion for the level of allowed interference for the data user scenario. The AT&T/T-Mobile Test Report also observed that the receive antenna gain used by Sprint and Verizon Wireless was likely optimistic, stating that most mobile receivers have a −1.5 to −3 dBm antenna gain. However, the AT&T/T-Mobile Test Report still adopted the 0 dBi value as it is typically used in link budget calculations.

97.

Sprint Interference Tests—Results.

In the Sprint Test Report, in the tests for receiver overload from Lower H Block in the PCS A Block, all six Sprint devices tested met the 3 dB desensitization level at a separation of 1 meter for all 24 test cases (12 interfering signals, 2 user scenarios). Four of the six devices met the 1 dB desensitization level at a separation of 1 meter, with the exception of one device for three test cases (out of the twenty-four total cases tested for that device). That device in that single case experienced blocking at 2 dB below the target level of −21 dBm for data use, which is equivalent to a separation of 1.3 meters. The other two Sprint devices experienced a 1 dB desensitization of their receivers at distances significantly greater than 1 meter in a majority of cases. V-COMM observed that the average interfering signal level that caused a 1 dB desensitization of the receiver was −22 dBm for a majority of devices, equivalent to a 1.1 meter separation.

98. In the Sprint Test Report, in tests for intermodulation and overload of the PCS B Block receiver, the results showed better performance than were observed for overload alone in the PCS A Block. Again, as with the overload tests, all devices met the 3 dB desensitization level for all test cases. Moreover, all devices experienced less than 1 dB of desensitization for the voice call in all instances. There were fewer failures in the data use scenario as well, with four of the six devices meeting the 1 dB desensitization level at less than 1 meter for data use. The other two devices experienced a 1 dB desensitization of their noise floor at distances of greater than 1 meter in half or more of the cases. These results for intermodulation were significantly better than were the results from testing in 2004.

99. After observing the difference in the results for the 1 dB and 3 dB desensitization levels, V-COMM conducted a test using the worst case interfering signal at a 2 dB desensitization level. At this level, all devices passed under the two user scenarios for both overload in the PCS A Block and overload plus intermodulation in the PCS B Block. In other words, no PCS device experienced a 2 dB or greater rise in the noise floor at a 1 meter separation from an H Block mobile device operating at 23 dBm, which is full power under the 3GPP LTE specification.

100.

Verizon Wireless Test Report—Results.

In the Verizon Wireless Test Report, in the tests for receiver overload from Lower H Block in the PCS A Block, all eight Verizon Wireless devices met the 3 dB desensitization level for all test cases. Four of Verizon Wireless's eight devices met the 1 dB desensitization level at a separation of 1 meter for both user scenarios. Of the other four devices, two experienced overload at the 1 dB desensitization level in approximately half of the test cases. V-COMM observed that the average interference levels for 1 dB desensitization for the six best devices was −21 dBm, which represents an H Block device transmitting at a 1 meter separation and at full power under the 3GPP LTE specification of 23 dBm EIRP.

101. In the tests for intermodulation and overload of the PCS B Block receiver, Verizon Wireless observed better performance than it observed for overload alone in the PCS A Block. As with the overload tests, all devices met the 3 dB desensitization level for all test cases. Six of the eight devices met the 1 dB desensitization level at 1 meter of separation for all of the voice call scenarios. There were ten instances out of a total of 144 (combination of six devices, two user scenarios and 12 interfering signals) in which the device experienced more than 1 dB of desensitization at a 1 meter separation. The two poorest performing devices experienced a 1 dB desensitization of the receiver at a distance of 1 meter in approximately half of the user scenarios. These results for intermodulation were significantly better than were the results from testing in 2004.

102. Just as it did for Sprint, V-COMM also conducted a set of tests using the worst case interfering signal at a 2 dB desensitization level. At this level, all devices passed for the two user scenarios for both overload in the PCS A Block and overload plus intermodulation in the PCS B Block. In other words, no device experienced more than a 2 dB rise of the noise floor at a 1 meter separation from an H Block mobile device operating at 23 dBm,

which is full power under the 3GPP LTE specification.

103.

AT&T and T-Mobile Test Report—Results.

The AT&T/T-Mobile Test Report stated that “all three airlink technologies displayed reasonable immunity to blocking and/or overload from an emulated H Block device.” In the AT&T/T-Mobile Test Report, under typical design conditions for light traffic, seven of the ten test cases met their stated criteria. The two GSM devices did not meet their interference criteria of −13 dBm, and “display[ed] noticeable performance impairment when the H Block device transmits at a power level within 2dB from its nominal maximum output power.” As explained above, AT&T and T-Mobile assessed the test results under different assumptions than did Sprint and Verizon Wireless. Based on examination of the test reports by Commission staff, under the data use scenario defined by Sprint and Verizon Wireless, all of AT&T and T-Mobile's devices would meet the criteria for receiver overload corresponding to 3 dB desensitization, for either worst case or typical design. Under 1 dB desensitization performance conditions, AT&T and T-Mobile's devices met their criteria in only one of six test cases.

104. In the tests for intermodulation, the AT&T/T-Mobile Test Report stated that “[n]o B Block performance impairment was noted . . . until the device was exposed to very high H Block signal levels.” Using AT&T and T-Mobile's assumptions, we observe their devices met their criteria in 15 of 18 test cases, over all desensitization levels, when lightly loaded. Based on Commission staff examination, all of the devices would have passed under Sprint and Verizon Wireless's user scenarios.

105. Looking separately at the results for the UMTS devices under high traffic conditions, the AT&T/T-Mobile Test report recorded more sensitivity to interference than under light traffic for the typical design case. Two of four receiver blocking test cases met their stated criteria, as did two of the four intermodulation test cases. We observe that all eight high traffic test cases would meet the criteria under the Sprint and Verizon Wireless data use scenario. Looking at a total of eight test cases for blocking (two devices, two interfering signal types, and two desensitization levels) and eight test cases for intermodulation, the UMTS devices were unable to meet the target BER under high traffic conditions before any interfering signal was applied in all but two of the sixteen cases. In other words, the devices were unable to perform acceptably in the complete absence of interference when the desire signal was set at only 1 dB or 3 dB above the device's sensitivity in high traffic.

106.

Power Limit Proposals Based on Interference Testing.

As a result of these studies, the four largest wireless providers all proposed the Commission adopt mobile and fixed power limits of 25 dBm EIRP, which is equivalent to a power limit of 300 milliwatts EIRP. First, in submitting its initial test results, Sprint concluded that “intermodulation interference is no longer a significant threat to today's PCS devices.” With regard to receiver overload, Sprint determined that the “potential for receiver blocking in today's PCS devices has decreased significantly to a point where blocking interference is unlikely.” Based on the evidence provided in the test data, Sprint proposed that a mobile power limit of 23 dB EIRP with a +/− 2 dB tolerance would protect adjacent PCS devices in the 1930-1995 MHz band. Second, Verizon Wireless recognized a similar improvement in the performance of its devices over time, stating that the newly tested devices “showed less sensitivity to interference than they did in 2004.” Specifically, the Verizon Wireless Test Report concluded that “based on receiver blocking test results, an H-Block mobile power limit of +23 dBm EIRP will prevent interference to the majority of PCS CDMA devices tested at 1 meter device separation.” Relying on the tests, Verizon Wireless stated that a power limit of 25 dBm EIRP “is the minimum needed to protect existing PCS operations from substantial interference.” Third, T-Mobile generally supported the 25 dBm EIRP proposed by Sprint and Verizon Wireless. T-Mobile was concerned, however, that H Block operations at a power level within 2 dB of the nominal maximum output power of 23 dBm could cause harmful interference for consumers with GSM devices and therefore requested that the Commission “require future H Block licensees . . . provide notification to PCS A Block licensees when they turn on service in the H Block on a market-by-market basis.” Fourth, AT&T stated that it “supports an H Block power limit of +23 dBm (+/− 2 dB) as “sufficient to ensure reasonable coexistence between LTE devices operating in the FCC's proposed H Block and UMTS and LTE devices operating in the PCS A and B Blocks.” AT&T further stated that “by the time LTE is widely deployed in the Lower H Block, GSM usage in the PCS Downlink Band on AT&T's network will be much less common than today, as AT&T deploys advanced technologies.”

107. Based on the record before us, we adopt a power limit for fixed and mobile devices operating in the Lower H Block of 300 milliwatts EIRP, which is equivalent to 25 dBm EIRP. As stated above and in the

H Block NPRM,

earlier testing conducted in 2005 identified the primary concern with full power mobile operations in the Lower H Block as intermodulation interference to PCS B Block receivers, with some additional concern regarding overload interference to PCS A Block receivers. The primary remedy to address receiver overload and intermodulation is through limits on mobile transmit power. At that time, parties argued for a severe reduction in the permissible mobile transmit power limit, such as imposing very strict power limits (

e.g.,

6 dBm EIRP) on the 1917-1920 MHz portion of the band, to address this problem. As detailed above, all of the studies showed that technological improvements over the past several years have resulted in mobile devices in the PCS band that can tolerate or mitigate against greater interference levels before overload or intermodulation interference rises to the level of causing harmful interference. In particular, while the testing performed in earlier years showed intermodulation interference to be a significant concern (and a much greater concern than overload interference), the new testing does not identify intermodulation as causing harmful interference. For example, in describing the results for both the Sprint Test Report and the Verizon Wireless Test Report, V-COMM stated that “CDMA devices tested generally showed less sensitivity (better rejection) to intermodulation interference as compared to [r]eceiver [b]locking—this is different from the 2004 devices tested.”

108. Consistent with the results of their studies, AT&T, Sprint, T-Mobile, and Verizon Wireless all proposed a power limit of 25 dBm EIRP, which is equivalent to 300 milliwatts EIRP, for operations in the entire Lower H Block. For example, Sprint “recommend[ed] that the Commission adopt a uniform H Block mobile device power limit of +23 dBm EIRP, with a +/− 2 dB implementation margin of tolerance . . . to protect adjacent PCS operations above 1930 MHz.” Verizon Wireless similarly stated that a power limit of 25 dBm EIRP is “the minimum needed to protect existing PCS operations from substantial interference.” AT&T and T-Mobile, in their joint test report, stated that a full power H Block mobile will not create significant impairment to UMTS or LTE devices, but that GSM devices “display noticeable performance impairment when the H Block device transmits at a power level

within 2 dB from its nominal maximum output power or 23 dBm.” In proposing a power limit of 25 dBm EIRP based on tests that showed significant instances of observed interference, the parties implicitly stated that the overall probability of interference was sufficiently low that it was deemed acceptable and did not rise to the level of harmful interference. No party opposed 25 dBm EIRP as a power limit across the Lower H Block or suggested that this power limit would lead to harmful interference to operations outside of the Lower H Block.

109. We adopt the proposed limit of 25 dBm EIRP, which is equivalent to 300 milliwatts EIRP, as the power limit for mobile and low power fixed operations in the entire Lower H Block and find, consistent with the Spectrum Act harmful interference condition, that operations subject to this power limit will not cause harmful interference to operations in the PCS downlink band. In adopting a power limit of 300 milliwatts EIRP, we observe that this limit is lower than the limits for other, comparable bands. For example, the power limit for mobile operations in the lower PCS Band (1850-1915 MHz) and in the AWS-4 Band is 2 watts EIRP, and in the AWS-1 Band is 1 watt EIRP. We nevertheless adopt the 300 milliwatts EIRP limit because it will protect against harmful interference to the PCS band, as required by statute, while enabling mobile devices deployed in the Lower H Block to operate at power levels sufficient to provide generally robust service quality, consistent with our goal of enabling efficient use of the band. Notably, in performing the testing and reaching the recommendations, the tests all were conducted assuming an LTE mobile device operating at the maximum power level indicated in the 3GPP LTE specifications—23 dBm. Consequently, adopting a power limit at 300 milliwatts (23 dBm, plus a 2 dBm tolerance) will enable the most likely H Block devices to operate without suffering any actual power restriction. That is, this power limit will permit mobile devices using LTE technology to operate at full power based on their design specifications. Moreover, 300 milliwatts EIRP is the level uniformly supported by the interference tests in the record as protecting against harmful interference into the 1930-1995 MHz PCS band.

110. Although we expect that setting the power limit at 300 milliwatts EIRP will not negatively affect mobile operations in either the Lower H Block or the 1930-1995 MHz PCS band, we observe that the test reports may not have fully captured the probabilistic nature of the interference scenario and that some of the assumptions used in performing the calculations in the interference tests may be overly conservative. It is important to identify these concerns with the test report inputs now so that they can be accounted for in future interference studies submitted to the Commission and because they also affect our analysis of OOBE interference, below. For the purpose of establishing the appropriate power limits, including under the Spectrum Act, the Commission determines what transmitter power level will prevent harmful interference, not simply detectable interference. For mobile-to-mobile interference, this is a probabilistic assessment. As we discuss further below in the discussion of OOBE limits, we find that the studies do not sufficiently account for the low probability of mobile-to-mobile interference actually occurring.

111. We are also concerned with some of the specific assumptions used in the test reports. In its analysis of the test data and stated conclusions for both the Sprint Test Report and the Verizon Wireless Test Report, V-COMM bases its conclusions on a number of assumptions, some of which may not be the most appropriate assumptions for calculating interference limits between nearby mobile systems. V-COMM bases its conclusions on the receiver's performance assuming a 1 meter separation between devices, a 1 dB desensitization level, and a data use case, which assumes 3 dB body loss and no head loss. Similarly, the AT&T/T-Mobile Test Report based its conclusions on a 1 meter device separation and a 1dB desensitization level. Further, unlike Verizon Wireless and Sprint, AT&T and T-Mobile made no provision for head or body loss.

112. First, one of several factors that will determine the likelihood of this probabilistic interference actually occurring is the separation distance between the mobile devices. As discussed below, a 2 meter separation between devices is a more appropriate separation distance than the 1 meter separation distance used in the studies. The Commission has adopted a 2 meter separation in the evaluation of other mobile-to-mobile interference scenarios, most recently in the AWS-4 proceeding. Further, AT&T and T-Mobile's concerns regarding the usefulness of testing under worst case conditions were demonstrated by the results for the high traffic test cases. The tested UMTS devices were unable to perform reliably under high traffic conditions, irrespective of the interference environment. Thus, the AT&T/T-Mobile test report lacks sufficient evidence to support any determination of harmful interference under high traffic conditions.

113. Second, as explained further below in setting OOBE limits, a 3 dB desensitization level is a more appropriate criterion than a 1 dB level upon which to judge harmful interference to mobile devices in cellular networks, which are designed to work in the presence of interference. For example, we observe that industry technical specifications for many types of devices that are currently used in the PCS band allow for a 3 dB degradation of the receiver sensitivity. The 3GPP2 standard for CDMA mobile devices sets the receiver performance requirements for intermodulation spurious response and receiver blocking based on a desired signal level of 3 dB above the reference sensitivity level. Based on the 3GPP2 standard for intermodulation, a CDMA device operating at 1% FER with a desired signal 3 dB above the reference sensitivity level is defined in the standard to be operating normally, and thus may be judged as not experiencing harmful interference. Similarly, the 3GPP standards for UMTS and LTE technologies allow the receiver sensitivity to degrade by 3 dB in response to interference. The LTE standard for receiver blocking is, moreover, is based on a desired signal level 6 dB above the receiver's reference sensitivity, requiring the receiver to perform in the presence of a strong interferer.

114. Third, as explained below, we believe it more appropriate to assume that the devices will be subject to both head and body loss, rather than just body loss. In both the Sprint Test Report and the Verizon Wireless Test Report, V-COMM tested for two different user scenarios. In one scenario, it assumed body loss only (that is, signal loss from proximity to the body, but not the head)—the data scenario. In the other scenario, it assumed signal loss from both the user's body and head—the voice scenario. For the data user scenario, V-COMM used a figure of 3 dB for body loss; for the voice scenario, it used 3 dB for body loss and another 5 dB for head loss. AT&T and T-Mobile did not apply any head or body loss in their analysis of the test results. As we describe further below, we believe it is more reasonable to use the voice user scenario, which includes both head and body loss assumptions, when determining interference rules.

115. We discuss our concerns with the use of these assumptions more fully below in establishing the OOBE limit.

116. Nevertheless, because, as explained above, the power limit that results from these tests will permit the deployment of full power H Block mobile devices in the 1915-1920 MHz band while also protecting commercial mobile service licensees in the 1930-1995 MHz band from harmful interference due to receiver overload, we find it unnecessary to adjust the studies for purposes of establishing power limits for operations in this band. Accordingly, we find it in the public interest, and consistent with the Spectrum Act's condition to protect the PCS downlink band from harmful interference, to set the power limit for mobile and fixed use in the 1915-1920 MHz band at 300 milliwatts EIRP.

b. Lower H Block Out-of-Band Emissions Limits

117. To minimize harmful interference between adjacent spectrum blocks, the Commission's rules generally limit the amount of RF power that may be emitted outside of the assigned block of an RF transmission. As explained below, we establish an OOBE limit for transmissions outside of the 1915-1920 MHz band of 43 + 10 log

10

(P) dB, where (P) is the transmitter power in watts, except that for emissions into the 1930-1995 MHz band we set an OOBE limit of 70 + 10 log

10

(P) dB, where (P) is the transmitter power in watts.

118. To minimize harmful electromagnetic interference between operators, the Commission has previously concluded that, in certain circumstances, attenuating transmitter OOBE by 43 + 10 log

10

(P) dB, where (P) is the transmitter power in watts, is appropriate. This limit is generally applied in cases where adjacent services have similar characteristics, such as base-to-base or mobile-to-mobile and adhere to similar power limits. As such, this limit applies to most of the services authorized under parts 24 and 27, including transmitters operating in adjacent blocks in the 1850-1915 MHz PCS band, which is adjacent to the Lower H Block. The Commission proposed requiring the attenuation level of 43 + 10 log

10

(P) dB, where (P) is the transmitter power in watts, to emissions from transmitters in the 1915-1920 MHz band, generally. As explained above, the Spectrum Act requires additional analysis with regard to Lower H Block transmissions into the 1930-1995 MHz band. As stated in the

H Block NPRM

and above, the proximity of mobile-to-mobile operations may require stricter OOBE limits than the Commission might impose in other interference scenarios. Specifically, the Commission proposed an OOBE limit of 70 + 10 log

10

(P) dB, where (P) is the transmitter power in watts, for emissions into the 1930-1995 MHz PCS Band. Finally, the Commission proposed to apply the measurement procedure used in the PCS band to these OOBE limits.

119. As explained above, the record contains three studies that examined the appropriate technical parameters for H Block operations needed to avoid causing harmful interference, including OOBE interference, to existing PCS downlink operations at 1930-1995 MHz.

120.

Sprint and Verizon Wireless Test Reports—Test Setup.

For the Sprint Test Report and the Verizon Wireless Test Report, V-COMM tested both Sprint and Verizon Wireless devices for their performance against out-of-band emissions. Two interference cases were tested. For both sets of tests, the CDMA device was tuned to the PCS A Block and subjected to a desired signal representing first a 1 dB desensitization level, and, second, a 3 dB desensitization level, from the device's measured sensitivity level. A co-channel additive white Gaussian noise (AWGN) signal representing the interfering H Block device was then injected into the device's RF antenna port. The power level of the interfering signal level was increased until the FER was no more than 0.5%, and the results recorded.

121.

AT&T and T-Mobile Test Reports—Test Setup.

As discussed above, AT&T and T-Mobile did not provide details of their test setup, but noted some differences with Sprint and Verizon Wireless's test plan. In performing that evaluation, a key difference from the V-COMM tests was that 7Layers set the desired signal level according to typical design at the device's reference sensitivity. Additional tests were conducted to determine the levels at which 1 dB and 3 dB degradation of the device's measured sensitivity occurs. The AT&T/T-Mobile Test Report did not include GSM devices in the typical design conditions. We observe that the analysis within the AT&T/T-Mobile Test Report did not calculate the necessary OOBE limit directly from the results, but simply decided whether the limit calculated in the V-COMM tests would be sufficient.

122.

Sprint and Verizon Wireless Test Reports—Test Results.

For the Sprint Test Report and the Verizon Wireless Test Report, V-COMM reported an average interference level of −107 dBm when the desired signal was at the 3 dB desensitization level, and an average interference level of −113 dBm when the desired signal was at the 1 dB desensitization level. Examining the same two user scenarios as for the blocking and intermodulation tests, V-COMM interpreted these results as equivalent to an OOBE limit −53 dBm/MHz for the voice user scenario at the 3 dB desensitization level and 1 meter of separation between devices, and −63 dBm/MHz for the data use scenario under the same conditions. For the 1 dB desensitization level, the results showed an equivalent OOBE level of −59 dBm/MHz for voice use and −69 dBm/MHz for data use. V-COMM stated that an OOBE limit of −69 dBm/MHz would prevent desensitization of more than 1 dB for devices at a 1 meter separation. It further stated that an additional implementation margin of 3 dB would be appropriate, resulting in a recommended OOBE limit of −66 dBm/MHz based on the data use scenario. V-COMM asserted that this limit would be “consistent with OOBE limits proposed in the FCC NPRM[s] in 2004 and 2008” and “also consistent with 3GPP OOBE limits for UMTS and HSPA devices.”

123.

AT&T and T-Mobile Test Reports—Test Results.

The AT&T/T-Mobile Test Report stated that the OOBE tests “showed the greatest difference between airlink technologies.” The report noted that “UMTS and LTE displayed good immunity to wideband noise emissions from a nearby H Block transmitter.” The report also stated that “GSM devices displayed relatively poor rejection of OOBE interference.” In the AT&T/T-Mobile Test Report, the average interference level for typical design conditions that produced 3 dB of desensitization of the receiver was −93.8 dBm. Similarly, the average interference levels for worst case conditions were −109.64 dBm and −104.8 dBm for 1 dB and 3 dB desensitization levels, respectively. This compares to the average levels of −113 dBm and −107 dBm for 1 dB and 3 dB desensitization levels, respectively, reported by both Sprint and Verizon Wireless.

124.

OOBE Proposals Based on Interference Testing.

Based on the testing, the parties generally proposed that the Commission adopt an OOBE limit of −66 dBm/MHz, which is equivalent to 96 + 10 log

10

(P) dB (where (P) is the transmitter power in watts) for Lower H Block emissions into the 1930-1995 MHz band. Sprint, however, recognized that this level may be overly stringent. Sprint suggested that, if the low probability of the occurrence of the factors needed for mobile-to-mobile interference were fully taken into account, the necessary OOBE attenuation could be lower. Sprint then observed that “[t]he 3GPP OOBE

standards for similar mobile-to-mobile coexistence situations are more typically −50 dBm/MHz [

i.e.,

80 + 10 log

10

(P) dB] (or −40 dBm/MHz [

i.e.,

70 + 10 log

10

(P) dB] when the two bands have little separation).” Verizon Wireless disagreed with Sprint, arguing that, “[a]lthough Sprint is correct as to the circumstances in which interference will occur, [Sprint] is wrong to imply that these circumstances occur only rarely.” Instead, Verizon Wireless argues that “mobile devices are most likely to be located very near each other at indoor locations where users are likely to receive a weaker signal . . . [which is] precisely what [OOBE] limits are designed to protect against.” Neither AT&T nor T-Mobile addressed Sprint's suggestion that the OOBE could be set at a less stringent level than 96 + 10 log

10

(P) dB. T-Mobile, while supporting the 96 + 10 log

10

(P) dB OOBE limit, expressed concern that the AT&T/T-Mobile Test Report showed that GSM devices had “a relatively poor rejection of OOBE interference at a separation distance of 1 meter.” To address this concern, T-Mobile requested that the Commission require H Block licensees to notify PCS A Block licensees on a market-by-market basis when the H Block licensees turn on service. T-Mobile explained that this “would enable full use of the H Block for LTE service while also assisting PCS licensees in network planning to reduce the probability of interference.”

125. For the reasons discussed below, except as otherwise specified, we adopt the proposed OOBE limit of 43 + 10 log

10

(P) dB, where (P) is the transmitter power in watts, for Lower H Block transmissions outside of 1915-1920 MHz. We adopt this limit below 1915 MHz and above 1920 MHz, with additional protections required for the 1930-1995 MHz band. For emissions into the 1930-1995 MHz band, we establish an OOBE limit of 70 + 10 log

10

(P) dB, where (P) is the transmitter power in watts.

126.

Emissions below 1915 MHz.

We adopt an OOBE limit of 43 + 10 log

10

(P) dB where (P) is the transmitter power in watts, for Lower H Block transmissions below 1915 MHz. Immediately below the Lower H Block is the 1850-1915 MHz PCS band, which is used for mobile transmit/base receive. As the Commission observed in the

H Block NPRM,

because it is anticipated that the Lower H Block systems will be similar in design to PCS and AWS-1, use of the 1915-1920 MHz band would be compatible with this adjacent PCS spectrum. That is, both bands will serve as mobile uplink bands. Thus, the OOBE level currently in the Commission's rules to protect adjacent PCS uplink blocks from harmful interference from each other should also be sufficient to protect PCS blocks in the 1850-1915 MHz band from Lower H Block emissions. Additionally, the OOBE limit of 43 + 10 log

10

(P) dB where (P) is the transmitter power in watts, has effectively served to prevent harmful interference to operations in bands adjacent and nearby to PCS and AWS-1 operations. The Commission thus tentatively concluded that a more restrictive OOBE limit than those established for PCS and AWS-1 transmissions was not necessary for Lower H Block transmissions below 1915 MHz; a conclusion now supported by the record. As Sprint comments, “[n]o industry commenter disputes the Commission's conclusion that [Lower] H Block uplink operations would not cause harmful interference to PCS operations located immediately below the uplink at 1850-1915 MHz.” We therefore adopt an OOBE limit of 43 + 10 log

10

(P) dB where (P) is the transmitter power in watts, for Lower H Block operations below 1915 MHz.

127.

Emissions above 1920 MHz.

Except as specified below for emissions into the 1930-1995 MHz band, we adopt an OOBE limit of 43 + 10 log

10

(P) dB, where (P) is the transmitter power in watts, for Lower H Block transmissions above 1920 MHz. The OOBE limit of 43 + 10 log

10

(P) dB, where (P) is the transmitter power in watts, applies to most of the services authorized under parts 24 and 27, which have effectively relied on this limit in the Commission's rules to prevent harmful interference to operations in adjacent bands. We authorize H Block under part 27, and thus anticipate that H Block systems will be similar in design to PCS and AWS-1. Additionally, with respect to the immediately adjacent 1920-1930 MHz band, that band is designated for unlicensed use and operations in that band are required to accept interference from licensed operations, including those in the Lower H Block. Furthermore, except as discussed below regarding the 1930-1995 MHz band, no commenter opposed an OOBE limit of 43 + 10 log

10

(P) dB above 1920 MHz. Therefore, we adopt an OOBE limit of 43 + 10 log

10

(P) dB, where (P) is the transmitter power in watts, for Lower H Block transmissions above 1920 MHz, subject to the exceptions below.

128.

Emissions into 1930-1995 MHz.

In order to prevent harmful interference into the PCS downlink band at 1930-1995 MHz, as required by statute, we adopt a requirement that out-of band emissions into the 1930-1995 MHz band be attenuated below the transmitter power level by at least 70 + 10 log

10

(P) dB, where (P) is the transmitter power in watts, (equivalent to −40 dBm/MHz) for fixed and mobile devices operating in the Lower H Block. We conclude that as a result of our adoption of this OOBE limit, licensees in the 1930-1995 MHz band will not experience a level of interference that seriously degrades, obstructs, or repeatedly interrupts their services. We base our finding on Commission precedent, experience with the probabilistic nature of mobile-to-mobile interference, and analysis of the test data submitted into the record.

129.

Commission Precedent.

We find an OOBE limit at 70 + 10 log

10

(P) dB, where (P) is the mobile transmitter power in watts, is consistent with Commission precedent. The interference scenario before us involves setting limits for Lower H Block mobile device out-of-band emissions that prevent harmful interference to PCS devices in the 1930-1995 MHz band. Last year, in the

AWS-4 Report and Order,

the Commission addressed the issue of mobile-to-mobile interference from AWS-4 mobile devices operating in the AWS-4 2000-2020 MHz uplink band to operations in the PCS downlink band and to future Upper H Block operations in 1930-2000 MHz. In the AWS-4 proceeding, the Commission had proposed an OOBE limit of 70 + 10 log

10

(P) dB, where P is the transmitter power in watts, from AWS-4 operations in the 2000-2020 MHz band into frequencies below 2000 MHz. The Commission proposed this attenuation level because it was previously set forth in the part 25 rules for Ancillary Terrestrial Component (ATC) operations in the 2000-2020 MHz band into spectrum below 1995 MHz. Parties in the AWS-4 proceeding generally supported the proposed OOBE level, and no party to that proceeding proposed an alternative limit. After reviewing the record before it—a record compiled after enactment of the Spectrum Act—the Commission adopted a maximum attenuation level of 70 + 10 log

10

(P) dB for AWS-4 transmissions into both the Upper H Block below 2000 MHz and the PCS band below 1995 MHz.

130. The scenario in the AWS-4 proceeding is on point with that facing us here. In both cases the interference scenario is mobile-to-mobile interference. In both cases, the Commission was faced with establishing an OOBE limit for transmissions from nearby operations into the PCS downlink band at 1930-1995 MHz. In the AWS-4 proceeding, the Commission also examined the same interference scenario into the immediately adjacent

Upper H Block. Further, in one important respect, the interference scenario before us now represents a scenario less likely to result in harmful interference than the one we addressed in the AWS-4 proceeding. Specifically, the Lower H Block is 10 megahertz away from the PCS downlink band, whereas the AWS-4 uplink band is 5 megahertz away from the PCS band and directly adjacent to the Upper H Block. Lower H Block operators will thus have 10 megahertz of frequency separation from the PCS band for emissions from their devices to roll off, while AWS-4 operators have no frequency separation for roll off between the AWS-4 uplink band and the Upper H Block. Stated otherwise, the interference scenarios here and in the AWS-4 proceeding effectively bookend the 1930-2000 MHz frequencies, with the emissions entering those from frequencies from below 1930 MHz and from above 2000 MHz needing to meet the same attenuation levels, but with Lower H Block operators having 10 megahertz rather than 5 megahertz or zero megahertz of separation in which to roll off to achieve the limit. Accordingly, we find it consistent with AWS-4 precedent to set the OOBE limit for Lower H Block operations into 1930-1995 MHz at 70 + 10 log

10

(P) dB, where (P) is the mobile transmitter power in watts.

131. In adopting the 70 + 10 log

10

(P) dB OOBE limit also set in the AWS-4 proceeding, we observe that this limit is the most stringent limit in the Commission's rules for operations in a commercial uplink band protecting another band. For example, for the 800 MHz cellular band and the Lower and Upper 700 MHz bands (generally), the Commission adopted an OOBE limit of 43 + 10 log

10

(P) dB (with a measurement bandwidth of 100 kHz, which is equivalent to 33+ 10 log

10

(P) dB with a measurement bandwidth of 1 MHz); and for the broadband PCS band, the AWS-1 band, and the AWS-4 band (except below 2000 MHz), the Commission adopted a mask of 43 + 10 log

10

(P) dB (with a measurement bandwidth of 1 MHz). Moreover, within these bands are examples of mobile-to-mobile interference scenarios at frequency separation distances similar to those that exist between the Lower H Block and the PCS downlink band. For example, Lower 700 MHz C Block mobile devices are required to attenuate transmissions at 43+ 10 log

10

(P) dB (with a measurement bandwidth of at least 100 kHz) above 716 MHz, including into the Lower 700 MHz A Block downlink band at 728 MHz. Similarly, in determining the OOBE limit for Upper 700 MHz C Block mobile devices into the nearby public safety downlink band, the Commission set the limit at the equivalent of 43 + 10 log

10

(P) dB (with a measurement bandwidth of 1 MHz). In addition, when 3GPP decided that public safety mobile devices required greater protection than the Commission limit, it set a higher limit of 65 + 10 log

10

(P) dB (with a measurement bandwidth of 1 MHz or greater). As part of the 3GPP deliberations, Verizon Wireless, a licensee of significant Upper 700 MHz C Block spectrum, agreed that this level provided sufficient protection to our Nation's first responders. Yet, here, in the H Block proceeding, wireless providers are advocating for a limit that is 31 dB (

i.e.,

more than 1,000 times) more stringent than the protection afforded public safety. We would expect, to the contrary, that protection levels sufficient for public safety would normally be sufficient to protect commercial mobile service providers.

132. Not only is the OOBE limit of 96 + 10 log

10

(P) dB much more stringent than the limits the Commission has adopted in any other band, it may be very difficult to realize. Sprint submitted a presentation from Avago Technologies that showed one solution using an FBAR (Film Bulk Acoustic Resonator) filter to meet the OOBE limit. The proposed filter was designed to support a single ten megahertz passband covering only the PCS G Block and the proposed H Block. As Sprint is the sole licensee for the PCS G Block, the filter design is very specialized for Sprint's purposes and is unlikely to be useable by other operators that may need to use larger passbands or other more commonly used filter technologies. It is important that the limits we set for H Block operations maximize the utility of the band for all potential licensees and provide for the public good.

133.

Probabilistic Interference.

In evaluating the interference scenario here, it is important to account for its probabilistic nature. In order for mobile-to-mobile harmful interference actually to occur, a number of worst case factors must all happen in conjunction with each other. These factors include that the two mobile devices (1) must be in operation at the same time, (2) must be located in very close proximity to each other, (3) must remain in close proximity for a significant period of time (

i.e.,

proximity must not be transient), (4) must be operating in a weak signal environment with both (a) the interfering mobile transmitter operating at maximum power and (b) the PCS mobile receiver receiving a weak signal and using frequencies most likely to lead to interference (

e.g.,

the interfering device must be capable of using the Lower H Block, actually transmitting on the Lower H Block, and transmitting on a resource block(s) near the upper edge of that band; the PCS device must similarly be operating on a receiver frequency near or at the lower edge of the PCS band), and (5) must be operating in a line of sight environment with respect to each other. Indeed, the Commission has described this issue for these bands previously, stating that “[t]he worst case occurs when the mobile transmitter is operating at maximum power (near the edge of its service area) at the upper edge of the band (near 1920 MHz) and the mobile receiver is trying to receive a weak signal (near the edge of its service area) at the lower edge of the band (near 1930 MHz) and only free space loss is considered.”

134. In addition, mobile devices do not transmit continuously; rather, they transmit data in bursts. For example, for LTE devices, mobile data is organized in resource blocks, which allocate a set of subcarrier frequencies for a 1 ms (millisecond) time interval. The frequency and duration of these bursts, or number of allocated resource blocks, depends upon traffic loads and signal conditions. For interference to PCS mobile devices to occur the H Block mobile must be transmitting in the same time interval that the PCS device is receiving. Thus, by transmitting in bursts, the likely use of LTE devices in the H Block would further dilute the probability of interference occurring. In addition, wireless networks constantly measure performance and seek to switch devices to alternative resources to improve call quality (

e.g.,

handoff to another channel or another base station).

135. The record supports this description of the factors that generally need to occur to give rise to mobile-to-mobile interference. For example, Sprint stated that “many factors come into play for such mobile-to-mobile interference.” It observed that interference would only occur if “(1) the PCS device is attempting to receive a weak signal at the bottom end of the PCS band; (2) the two mobile devices are located very near to each other; and (3) the H block device is transmitting at the same instant, with high power and in the resource blocks at the upper end of the H block.” Verizon Wireless concurred, expressly stating that “Sprint is correct as to the circumstances in which interference will occur.” Accordingly, we reiterate that mobile-to-mobile interference will occur only in specific

situations, such as those described above.

136. The risk of mobile-to-mobile interference occurring is influenced by the low probability of these worst-case circumstances occurring—they may occur, but do so infrequently—and by network management practices, such as hand off and power management, that are designed to mitigate against harmful interference. For example, Sprint states that LTE “spreads across the bandwidth, dynamically controlling the power and number of subcarriers assigned to a particular device and reducing the need for constraining OOBE limits.” Moreover, as Sprint observes, “[p]robability certainly plays a large factor as to when [the above] conditions would occur in the real world.” We believe that the probability of each of the described mobile-to-mobile interactions actually occurring is small individually, and quite small viewed in combination. Thus, we disagree with Verizon Wireless's assertion that the combination of circumstances resulting in interference does not “occur only rarely . . . [because] mobile devices are most likely to be located very near to each other at indoor locations where users are likely to receive a weaker signal.” Although the confluence of worst case scenarios may occur more often indoors than outdoors, it does not necessarily follow that these situations occur indoors with any frequency; nor has Verizon Wireless provided any evidence showing that these factors occur frequently indoors. Further, in areas where wireless providers anticipate recurring high density use of mobile devices, providers typically engineer their networks to provide robust coverage, including for indoor locations.

137. We apply our discussion of the probabilistic nature of mobile-to-mobile interference to our evaluations of the test reports, immediately below.

138.

Test Reports.

While we believe it appropriate to act consistently with the Commission's recent determination in the AWS-4 proceeding that an attenuation limit of 70 + 10 log

10

(P) dB, where P is the transmitter power in watts, from the AWS-4 uplink band into the PCS downlink band at 1930-1995 MHz to set that same limit here for transmissions from the Lower H Block into the PCS downlink band, we believe it appropriate to test this conclusion against the test reports submitted into the record here. As explained above, parties submitted three test reports into the record. We assess these reports based on our engineering expertise and with the goal of auctioning the Lower H Block in a manner that maximizes its usefulness while protecting the PCS band from harmful interference, as required by the Spectrum Act.

139. We have a number of concerns with the test reports. In particular, as we discuss above, although we do not question the science behind the reports, we find a number of assumptions used by the parties in their interference tests are overly conservative for use in setting reasonable OOBE limits. Specifically, we find the testing (1) failed to fully account for the low probability of mobile-to-mobile interference, (2) assumed an overly conservative required separation distance of 1 meter, (3) relied on limiting interference to an overly conservative 1 dB desensitization level, (4) relied on an overly restrictive user scenario that accounted for body loss only, as opposed to head and body loss, and (5) included an unnecessary manufacturer's tolerance. We address each of our concerns with the test reports, below, in turn.

140. First, the test reports do not fully account for the highly probabilistic nature of OOBE interference from the Lower H Block into the PCS downlink band. As explained above, many low probability factors must occur in conjunction for interference to occur in a mobile-to-mobile scenario. Because our charge is to prevent harmful interference, rather than all interference, accounting for the likelihood that an instance of interference will occur is important in assessing whether the interference scenario rises to the level of harmful interference. For example, as the Commission has said previously, whether the user would actually notice the interference may be an important element of determining if interference is harmful. Except for one factor—separation between devices, which we discuss immediately below—no information provided in the test reports indicates that they accounted (or attempted to account) for the probabilistic nature of the interference. Because the test reports did not fully account for the probabilistic nature of the interference at issue, we believe they overstate the protection from OOBE interference needed by licensees operating in the 1930-1995 MHz band.

141. Second, we examine the one probabilistic factor included in the test reports—separation distance. The selection of the separation distance between devices is a key factor in determining the probability of that interference could occur. As stated above, the Sprint Test Report, the Verizon Wireless Test Report, and the AT&T/T-Mobile Test Report all assumed a separation of 1 meter between devices. A 1 meter separation is often used as a minimum separation distance in industry analyses of mobile-to-mobile interference. Distances of less than 1 meter risk the possibility that near field antenna coupling effects may distort the propagation between the two devices and undermine the assumption of free space path loss. Again, as discussed earlier, the simple presence of interference is not necessarily the same as harmful interference. To determine what interference is sufficient to be considered harmful, one should consider whether there is a reasonable probability that the conditions necessary to create that interference will occur. The Commission has previously supported a separation of 2 meters as an appropriate assumption for the purposes of determining an acceptable level of interference. For example, in the

AWS Sixth Report and Order,

the Commission expressed support for a 2 meter separation distance, stating that “this short distance coupled with the low probability of occurrence of the worst-case scenario (both mobiles at the edge of coverage, both operating at the edge of the band, both simultaneously active, and both in close proximity to each other), make interference of this nature highly unlikely.” More recently, in the

AWS-4 Report and Order,

the Commission found it reasonable to rely on the 2 meter separation distance proposed by Motorola Mobility in calculating interference limits. Accordingly, we believe that a 1 meter separation distance represents an overly conservative value and that it is a more realistic scenario to assume that the devices at issue are likely to be at least 2 meters apart.

142. Third, we turn to inputs used in the test reports that are not associated with the probabilistic nature of the interference scenario, and start with the desensitization level. While the reports use a 1 dB desensitization level, we believe a 3 dB level is more appropriate. The Sprint and Verizon Wireless test reports include results of the testing for both the 1 dB and 3 dB desensitization levels, but focused their analysis of the results on the 1 dB desensitization level. For purposes of the AT&T/T-Mobile Test Report, AT&T and T-Mobile designed their test plan to use a 3 dB desensitization of the receiver's sensitivity. The desensitization was based on the device's reference sensitivity per the standard for the technology, rather than by the individual device's measured sensitivity (the approach used by Sprint and Verizon Wireless). AT&T and T-Mobile described this test strategy as a typical design test, observing that most link

budgets, which drive the design of the network, use the standard's reference sensitivity. Further, they stated that the reference sensitivity, as opposed to the individual device's measured sensitivity allows all devices “to be tested in exactly the same environment” for a better comparison of device performance.

143. A 1 dB desensitization level is defined as the level of interference at which the effective noise floor of the system will rise by 1 dB, that is, the receiver sensitivity will be reduced by 1 dB. This occurs when the interfering signal level is 6 dB below the noise floor of the receiver. Similarly, 3 dB desensitization occurs when the level of interference is equal to the level of the receiver's system noise. 1 dB desensitization is most commonly used as an interference protection criterion for noise-limited receiver systems. However, mobile cellular systems are inherently interference-limited; that is, the prevailing interference is greater than noise sources. These systems are designed to perform in a strong interference environment, much of which is often self-generated, coming from other network elements (

e.g.,

other nearby base stations in the same or adjacent bands).

144. We believe that a noise-limited interference criterion (1 dB desensitization) is too restrictive for modern cellular systems. This is reflected in industry standards for receiver performance, such as the 3GPP2 standard for CDMA devices. As described above, the 3GPP2 standard for cdma2000 mobile devices sets several receiver performance requirements, including response to receiver overload (blocking) and intermodulation. For example, 3GPP2 Requirement 3.5.2 for Single Tone Desensitization, similar to the intermodulation tests performed by V-COMM, sets the level of the desired signal at either 3 dB or 10 dB above the reference sensitivity level. Similarly, under the 3GPP2 standard, receiver blocking also permits sensitivity to degrade by 3 dB above its reference level in the presence of overload interference while maintaining a 10% FER. CDMA is not the only technology to require the receiver to operate properly in the presence of interference. The 3GPP standard for UMTS and LTE devices specifies an in-band blocking requirement that sets the interfering signal level 6 dB or more above the reference sensitivity level. Further, for GSM, the desired signal is set at 3 dB above refere

This text is long and has been trimmed here. Open the source document for the complete record.