On the theory that old news is still interesting from time to time, we will archive
FCC news articles so they may age gracefully on this page. Look around and see what was big "back
when!"
*************************************************************************
Before the
FEDERAL COMMUNICATIONS COMMISSION FCC 96-36
Washington, D. C. 20554
In the Matter of)
Amendment of Parts 2 and 15 of the ) ET Docket No. 96-8
Commission's Rules Regarding Spread ) RM-8435, RM-8608, RM-8609
Spectrum Transmitters )
NOTICE OF PROPOSED RULE MAKING
Adopted: January 30, 1996 ; Released: February 5, 1996
Comment Date: [75 days from date of publication in the Federal Register]
Reply Comment Date: [105 days from date of publication in the Federal Register]
By the Commission:
INTRODUCTION
1. By this action, the Commission proposes to amend Parts 2 and 15 of the rules
regarding the operation of spread spectrum transmission systems in the 902-928 MHz, 2400-
2483.5 MHz and 5725-5850 MHz bands. Specifically, we are proposing to eliminate the
limit on directional gain antennas for spread spectrum transmitters operating in the 5800 MHz
band. We are also proposing to reduce, from 50 to 25, the minimum number of channels
required for frequency hopping spread spectrum systems operating in the 915 MHz band.
These proposals are in response to Petitions for Rule Making filed by Western Multiplex
Corporation (WMC) and SpectraLink Corporation (SpectraLink), respectively. We also are
denying a Petition for Rule Making from Symbol Technologies, Inc. (Symbol) that seeks a
reduction in the minimum number of required hopping channels, from 75 to 20, for frequency
hopping spread spectrum systems operating in the 2450 MHz and 5800 MHz bands. On our
own motion, we are proposing a number of amendments to the spread spectrum regulations to
clarify the existing regulations, to codify existing policies into the rules, and to update the
current definitions. The changes will expand the ability of equipment manufacturers to
develop spread spectrum systems for unlicensed use.
BACKGROUND
2. Part 15 of the regulations permits the operation of radio frequency devices without
a license from the Commission or the need for frequency coordination. The technical
standards for Part 15 transmission systems are designed to ensure that there is a low
probability that these devices will cause harmful interference to other users of the spectrum.
Indeed, the primary operating conditions under Part 15 are that the operator must accept
whatever interference is received and must correct whatever interference is caused.
3. Spread spectrum communications systems use special modulation techniques that
spread the energy of the signal being transmitted over a very wide bandwidth. The
information to be conveyed is modulated onto a carrier frequency by some conventional
technique, such as AM, FM or digital, and the bandwidth of the signal is deliberately widened
by means of a spreading function. The spreading technique used in the transmitter is
duplicated in the receiver to enable detection and decoding of the signal. This spreading
reduces the power density of the signal at any frequency within the transmitted bandwidth,
thereby reducing the probability of causing interference to other signals occupying the same
spectrum. In addition, the signal processing tends to suppress undesired signals, enabling
these systems to tolerate strong interference.
4. Two types of spread spectrum systems are permitted under the Part 15 regulations:
direct sequence systems and frequency hopping systems. Direct sequence systems modulate
the carrier with a combined information signal and a much faster binary code signal. The
binary code signal is a fixed-length pseudorandom sequence of bits. It dominates the
modulating function and is the direct cause of the spreading of the transmitted signal.
Frequency hopping systems spread their energy by changing, or "hopping," the center
frequency of the transmission in accordance with a pseudorandomly generated list of channels.
5. The existing regulations limit spread spectrum systems to a maximum peak
transmitter output power of one watt. When operating at that power level, the maximum
directional gain of the associated antenna may not exceed 6 dBi, resulting in a maximum
equivalent isotropically radiated power (EIRP) of four watts. Direct sequence systems must
employ a minimum 6 dB bandwidth of 500 kHz with a processing gain of at least 10 dB.
Frequency hopping systems in the 915 MHz band must use at least 50 hopping channels with
a maximum 20 dB channel bandwidth of 500 kHz, while hopping systems in the 2450 MHz
and 5800 MHz bands must use at least 75 hopping channels with a maximum 20 dB channel
bandwidth of 1 MHz.
PETITIONS FOR RULE MAKING
Western Multiplex Corporation (WMC)
6. The Petition. WMC filed a combined Petition for Rule Making and Request for
Immediate Waiver with the Commission. It requests that the limits on antenna gain be
removed for spread spectrum systems operating in the 2450 MHz and 5800 MHz bands.
WMC also requests a waiver of this regulation pending consideration of the Petition for Rule
Making by the Commission. In its petition and comments, WMC indicates that spread
spectrum systems employing antennas with levels of gain higher than that permitted under the
current rules are being used to provide 1.5 Mbps T1-type links without having to go through a
frequency coordination and licensing process. It states that this enables rapid setup of the
system and reduces costs to the user. WMC states that typical users include manufacturing
and service companies, oil and gas pipeline companies, mobile and SMR operators, common
carriers, public safety services, state and local governments and the U.S. Government. WMC
states that typical applications include communications to oil platforms, emergency restoration
of communications in disaster situations, low density spurs off microwave fiber optic
backbone systems to serve remote field offices and service centers, extensions of local area
networks, law enforcement and fire prevention communications, transmission of supervisory
control and data acquisition circuits, seismic monitoring equipment transmissions, data
transmissions, remote control of stacker reclaimer equipment, power plant control yard
communications, and connections of mobile radio cell sites to the main telephone switching
site. WMC argues that there are no known cases of harmful interference involving these
transmitters operating with higher antenna gains. As stated by WMC, the use of directional
(high gain) antennas, because of the line of sight propagation characteristics above 2 GHz,
permits greater reuse of frequencies in an given area and yields a significantly higher
transmission capacity per unit bandwidth than can be achieved using non-directional (low
gain) antennas.
7. Most of the commenting parties support WMC's request. Two parties, Alcatel
Network Systems, Inc., (Alcatel) and WINDATA, Inc., (WINDATA), oppose this request.
Alcatel states that the WMC proposal could have the effect of reallocating the 2450 MHz and
5800 MHz bands for point-to-point use by unlicensed devices, whereas these bands could be
used more efficiently by licensed microwave systems. WINDATA states that the use of
higher gain antennas could result in unacceptable interference, especially to indoor Part 15
spread spectrum systems, such as wireless local area networks, that could be in the main beam
of a WMC system.
8. Cylink Corp., Metricom, Inc., and Tetherless Access Ltd. filed comments
requesting that the limit on maximum directional antenna gain also be deleted for systems in
the 915 MHz band. This request was opposed by Pinpoint Communications, Inc., which
expresses concern for potential interference by spread spectrum systems to its automatic
vehicle monitoring system/location monitoring service operating in the same band under
Part 90 of our regulations.
9. Discussion. We recognize the advantages of being able readily to establish radio
links capable of transmission distances of 10 km, or greater, without the delays and costs
associated with frequency coordination and licensing. The ability to establish such
transmission links quickly could be critical in emergency situations. However, because the
use of high power radio links without prior frequency coordination could result in significant
interference problems to other operators using these frequency bands, we believe it is
necessary to restrict their use. The frequency bands addressed in the WMC petition are
allocated for operation by industrial, scientific and medical devices, U.S. Government stations,
the Amateur Radio Service, the Private Land Mobile Radio Services, the Private Operational-
Fixed Microwave Services, and Television Broadcast Auxiliary Stations. The other spread
spectrum band, the 915 MHz band, is allocated for operation by industrial, scientific and
medical devices, the Private Land Mobile Radio Services, and the Amateur Radio Service.
Accordingly, we are proposing to eliminate the antenna directional gain limit only for non-
consumer, fixed, point-to-point spread spectrum systems operating in the 5800 MHz band.
10. The spread spectrum rules, as originally adopted, did not specify a limit on
antenna gain. At that time, there were few other operators in these bands and little potential
that interference would be caused to other users. Further, we wished to offer an incentive to
spur the development of spread spectrum systems. These bands, especially the 915 MHz and
the 2450 MHz bands, are now becoming more crowded, particularly with mobile units,
increasing the potential that spread spectrum systems using high gain antennas will cause
harmful interference. In addition to the licensed radio services, wireless computer local area
network systems and various consumer products, such as cordless telephones, are being used
under Part 15 in the 915 MHz and 2450 MHz bands.
11. However, as there are few operators in the 5800 MHz band, the potential that
harmful interference will occur from the use of directional antennas is much lower. There are
also fewer mobile users in the 5800 MHz band. It is easier to engineer a fixed, point-to-point
system to operate without causing harmful interference problems if the other stations in that
band are fixed in location. Further, the 5800 MHz band is ideal for fixed, point-to-point
wideband microwave operations, the type of applications desired by WMC. Accordingly, we
believe the limit on directional antenna gain should only be eliminated for spread spectrum
systems operating in the 5800 MHz band. We request comment on this proposal. While we
are not inclined to provide a similar relaxation for the 2450 MHz band, we also ask for
comment on whether we should eliminate the 6 dB limit on directional antenna gain in this
band.
12. While the use of high gain directional antennas can reduce the potential for
interference to radio operations located outside the directional beam of the antenna pattern, the
potential for harmful interference to radio systems located in the beam of the directional
antenna increases significantly. Obviously, if multiple antennas are employed, providing
point-to-multipoint or omnidirectional operation, the coverage area increases, but so does the
potential for harmful interference. In addition, we recognize that the use of high gain
directional antennas benefits primarily fixed applications. Therefore, we also believe that the
restriction on directional antenna gain should be eliminated only for those 5800 MHz systems
used for fixed, point-to-point operations. This proposal would not eliminate the limit on
antenna gain for point-to-multipoint or omnidirectional systems, transmitters employing
multiple directional antennas, or multiple co-located transmitters transmitting the same
information.
13. We further believe that if spread spectrum transmitters employing high gain
antennas were made available to the general public, it would be difficult to ensure that these
systems are used only for fixed, point-to-point applications. In addition, high gain directional
antenna systems, because of their narrow transmission beamwidth and the problems associated
with aligning the transmitter with the receiver site, are not products that would normally be
employed by the general public. Further, as indicated by WMC and the supporting
comments, these high gain transmission systems are being employed by commercial and
industrial operators for communication back-haul systems in order to avoid the expenses and
delays associated with licensed systems. Accordingly, we believe that the marketing of spread
spectrum systems employing high gain antennas should be limited to commercial or industrial
operators and exclude sales to the general public. We further propose to hold the operator of
a spread spectrum system responsible for ensuring that the system is operated in a compliant
manner. In addition, we propose to require that the manual supplied with the spread
spectrum transmitter contain language in the installation instructions notifying the operator of
this responsibility. Commenting parties should note that the transmitter must be authorized
under our certification procedure along with the specific antenna with which it will be used.
These components are authorized as a system to ensure that the emission limits, especially
those limits designed to protect other sensitive radio services or services used for safety-of-life
applications, are not exceeded.
14. Absent controls regarding the locations and manner in which spread spectrum
transmitters may be used, systems employing high gain directional antennas could expose the
public to potentially harmful signal levels that exceed the radio frequency exposure limits of
our rules and recommended by various standards-setting organizations. In order to meet our
obligation under the National Environmental Policy Act, we propose to hold the holder of the
grant of certification for the transmitter, the grantee, responsible for ensuring that the
equipment is designed to minimize exposure of the public to excessive radio frequency (RF)
signal levels. While we proposed to make the operator responsible for ensuring that the
system is used only for fixed, point-to-point applications, the means to prevent excessive
exposure levels can be incorporated into the equipment design. In addition, absent some
action by the grantee to incorporate a warning into the equipment, the operator would not
necessarily realize that there was a potential for excessive RF exposure levels. A possible
method is a sign, attached to the antenna and of sufficient size and visibility, warning the
public of the potential danger of RF exposure. Another possible method is the incorporation
of proximity sensors that cause the transmitter to automatically decrease output power if
someone wanders too close to the transmitting antenna. Comments are requested concerning
possible biological hazards from the high effective radiated power levels that could be emitted
from these systems, any additional methods that can be employed to prevent unnecessary
exposure of the public, and whether we should prescribe the use of specific means for
preventing such exposure.
15. Commenting parties should note that informal comments raising concerns with the
WMC petition, particularly operation in the 2450 MHz band, have already been received from
the staff at Industry Canada, an agency of the Canadian Government. Similarly, the Mexican
Government has expressed its concern regarding unlicensed spread spectrum operations
between stations in the U.S. and stations in Mexico. Thus, commenting parties may also
wish to address actions that could be taken to limit operation near the Canadian and Mexican
borders to avoid unauthorized crossborder operations and interference to licensed systems in
Canada and Mexico.
16. Comments are also requested in two additional areas regarding the technical
standards for spread spectrum transmission systems operating without a limit on directional
antenna gain. The first of these concerns a reduction in the output power of the transmitter
based on the amount that the increase in directional antenna gain exceeds the current limit of
6 dBi. We propose that the output power of a transmitter would need to be decreased by
1 dB for every 3 dB that the antenna gain exceeds 6 dBi in order to maintain an "equivalent"
area of interference, i.e., the geographic area over which interference could result with a
directional antenna as compared to the area obtained with an omnidirectional antenna.
While this would result in a slight reduction in the effective radiated power level of the
system, the higher gain employed by the antenna would still be available to amplify the
received signal.
17. Comments are also requested on whether the rules should specify limits on the
horizontal and vertical beamwidths of antennas used with these point-to-point systems.
Certain antenna designs, e.g., a horizontally polarized yagi antenna, concentrate the signal
strength in azimuth (horizontal) but not in elevation (vertical). However, a fixed, point-to-
point system employing an antenna with a wide elevation beamwidth that is pointed towards
an office building with multiple floors could result in severe interference problems to any
party in that building who is in line with the system and is operating in the same band.
Several antenna designs concentrate the radiated signals in both azimuth and elevation, e.g.,
circular dish antennas and stacked yagi antennas. We believe that any interference problems
resulting from excessive vertical emissions could be resolved if the 3 dB beamwidths, in both
the vertical and the horizontal planes, of the high gain directional antennas employed with
these fixed, point-to-point systems differ by no more than a factor of two and are proposing
such a limit.
Symbol Technologies, Inc. (Symbol)
18. The Petition. In its Petition for Rule Making, Symbol requests a reduction in the
number of hopping channels, from 75 to 20, required for frequency hopping spread spectrum
systems operating in the 2450 MHz and 5800 MHz bands. Under this plan, the average time
of occupancy for the hopping transmitter would be limited to no longer than 0.4 seconds on
any one channel in any 8 second interval, as opposed to the current limit of 0.4 seconds on
any one channel in any 30 second interval. Further, the output power would be reduced from
the current one watt peak output to (number of hops/75) watts. Symbol indicates that the
proposed changes would raise the effective signal bandwidth of a frequency hopping system
from 1 MHz to about 5 MHz.
19. Symbol states that its proposed changes would align U.S. standards for unlicensed
spread spectrum systems more closely with the European standards for such equipment,
permitting U.S. manufacturers to produce the same equipment for domestic use and for export
to Europe. It adds that this, in turn, would reduce the cost of exported goods and increase
U.S. exports to Europe. Symbol further states that its proposed language is compatible with
the European Telecommunications Standards Institute standard ETS 300-328 for the
2450 MHz band. Symbol also indicates that frequency hopping equipment manufactured
under this proposal would be capable of substantially higher data speeds than can be achieved
under the present frequency hopping rules. It states that such systems could be used to satisfy
customer demands for high speed, low cost wireless computer local area networks (LANs),
making it possible for wireless LANs to compete with wired LANs while offering the wireless
advantage of portability and flexibility. Symbol also argues, in its petition, that these
amendments would not increase the threat of interference to other users in this spectrum.
Further, Symbol adds that the proposed changes to the rules would correct a competitive
disparity between direct sequence systems and frequency hopping systems. Finally, Symbol
adds that its proposal would not significantly detract from the major advantages of spread
spectrum long recognized by the Commission: its ability to enhance spectrum efficiency by
sharing spectrum with other services in ways that minimize cross-interference.
20. Comments. Apple Computer, Inc. (Apple), Norand Corporation (Norand), and
SpectraLink Corporation (SpectraLink) filed comments supporting Symbol's petition.
Norand states that these proposals would foster the joint development of products for both the
U.S. and the European markets. Norand also concurs with Symbol's assessment that the
proposed changes would not increase the potential for interference to other radio operations,
and that frequency hopping systems operating under these relaxed standards would create no
greater interference potential than currently caused by direct sequence systems. Apple
expresses support for Symbol's proposal, stating that these changes would permit wireless
LANs to operate at higher data rates while protecting other users of the band from increased
interference and would harmonize our rules with European standards. Apple, however, also
adds that Symbol's proposal does not appear to address the risk that narrow band systems
employing a few hopping channels could dominate a frequency range and cause that range to
appear occupied and unavailable to devices employing wideband channels. Apple is also
concerned that the adoption of this proposal could adversely affect the efforts of the Institute
of Electrical and Electronic Engineers (IEEE) Standards Working Group P802.11 to develop
industry-wide standards for wireless LANs. Apple submits that other Commission actions,
including the allocation of spectrum at 2390-2400 MHz and the development of the millimeter
wave bands above 40 GHz, may also provide new opportunities for high bandwidth wireless
transmissions. SpectraLink concurs with Symbol that a reduction in the minimum number
of frequency hopping channels, coupled with a corresponding increase in the permissible
channel bandwidth, would allow a higher data throughput, but also urges us to stipulate that
the maximum transmitter dwell time per hop on any channel may not exceed 0.4 seconds.
21. In addition to the concerns raised by Apple, Aironet Wireless Communications,
Inc. (Aironet) and AT&T oppose Symbol's request. Aironet contends that the proposed
changes would lead to potentially serious interference to other spread spectrum users as well
as to primary (licensed) users of these bands. It also states that the use of wider bandwidth
transmissions would significantly reduce the processing gain and interference rejection
capabilities of frequency hopping systems. Aironet agrees that there will be substantial
demand for wireless LANs, but argues that these systems should use direct sequence and other
technologies which, it contends, are more than sufficient to meet the technical requirements
for wireless LANs. In regard to Symbol's statement that its proposal is similar to European
standards, Aironet states that the new European standards came into effect after considerable
controversy in 1994, and very few, if any, of these frequency hopping systems have been
developed in Europe to date because of the continuing controversy within the industry.
AT&T also expresses concern that these rule changes would increase the potential for
interference to other devices in the bands. It states that frequency hopping systems using a
wider bandwidth would have a greater chance of transmitting on a frequency used by a nearby
system and, thus, interfering with that other system. AT&T further states that with fewer
hopping channels, frequency hopping systems would frequently collide with each other and
with direct sequence systems. This interference would increase as spread spectrum systems
proliferate.
22. Some of the comments addressed methods that could provide additional versatility
for frequency hopping spread spectrum systems. Apple indicates that it may be advantageous
to consider possible alternative numbers of hopping channels, as well as changes in output
power for increased bandwidth transmissions. Metricom, Inc. (Metricom) states that the
Commission should specify the maximum bandwidth of a hopping channel as a function of
the number of hopping channels and should specify a channel distribution in order to ensure
that there is an even distribution over the entire frequency band employed. This would
prevent the band from becoming "channelized" and would prevent the "bunching" of
frequency hopping channels in any particular segment of the spectrum.
23. Discussion. We have serious concerns that implementing Symbol's requested
changes could result in severe increases in the potential for harmful interference, both to the
authorized radio services and to other Part 15 devices operating in these bands. Symbol's
request to decrease the number of hopping channels would result in an increase in the average
time during which the channels are occupied by a spread spectrum transmission. In addition,
Symbol's request to increase the bandwidth of the hopping channels would broaden the
spectrum over which interference from the frequency hopping systems could be received.
Thus, we believe that implementing these changes would be detrimental to other narrowband
and wideband systems operating in these bands. While this increased interference potential
could be partially offset by a reduction in the output power of the frequency hopping
transmitters, we are not convinced that a linear power reduction alone is sufficient to offset
this interference potential. We also note that any benefit from the reduction in output power
could be negated if we adopt the proposal from WMC, described above, to eliminate the
current restriction on antenna gain in the 5800 MHz band. In addition, the resulting increase
in hopping channel bandwidth would open frequency hopping systems to several new
consumer applications, such as analog video transmissions, in addition to the wireless local
area network applications described in the petition. Normally, the Commission seeks to
encourage new uses of the spectrum. However, in this case we feel that the large increase in
the proliferation of these transmitters from additional consumer applications, combined with a
smaller number of hopping channels, an increased bandwidth, and increase in average channel
occupancy time, and, in some cases, a higher effective radiated power, would result in a
significant increase in the probability that harmful interference will occur to other radio
operations in these bands.
24. We also observe that there appears to be sufficient spectrum, either currently
available or under proposal, to support high data speeds for wireless local area network
systems. For example, we have opened 5 GHz of spectrum in the 59-64 GHz band, and have
proposed to open additional bands above 40 GHz, for consumer applications. In addition,
licensed wideband digital transmission systems may be operated under the Personal
Communications Services, the Private Land Mobile Radio Services, and the Private
Operational Fixed Microwave Service. Unlicensed wideband digital signals may also be
transmitted in the 2450 MHz and 5800 MHz bands using either direct sequence spread
spectrum modulation or, at a lower power level, conventional modulation techniques and in
the bands 1910-1920 MHz and 2390-2400 MHz. In addition, it is possible under the
existing rules to construct frequency hopping systems in the 2450 MHz and 5800 MHz bands
that can operate at high data rates through the use of more efficient modulation techniques,
such as quadrature amplitude modulation. We are also considering the possibility of
providing additional spectrum for unlicensed operations under a Petition for Rule Making
from Apple to open the frequency bands at 5150-5300 MHz and 5725-5875 MHz as well as a
Petition for Rule Making from the Wireless Information Networks Forum (WINForum) to
open spectrum at 5100-5350 MHz.
25. Finally, as indicated in the comments, the IEEE Standards Working Group
P802.11 is currently developing industry standards for wireless LANs, including operation in
the 2450 MHz band. We are reluctant to propose any changes to the existing spread spectrum
standards regarding the transmission of wideband digital systems prior to the release of that
Committee's recommendations. While we agree with Symbol that harmonization with the
European standards would be advantageous, harmonization is not sufficient, by itself, to
overcome all of the potential problems associated with reducing the minimum number of
hopping channels. Accordingly, in light of the above considerations we are denying the
Petition for Rule Making from Symbol to reduce the minimum number of hopping channels
for frequency hopping spread spectrum systems operating in the 2450 MHz or 5800 MHz
bands.
SpectraLink Corporation (SpectraLink)
26. The Petition. In its Petition for Rule Making, SpectraLink requests a reduction,
from 50 channels to 25 non-contiguous channels, in the number of hopping channels required
for frequency hopping spread spectrum systems operating in the 915 MHz band.
SpectraLink also requests that frequency hopping systems that use fewer than 50 hopping
channels be limited to a maximum transmitter peak output power of 500 mW. It states that a
system operating with 25 hopping channels and a corresponding 3 dB reduction in output
power would have the same spectral power density as a system using 50 hopping channels.
27. Adoption of the SpectraLink proposal would allow a reduction, from 25 MHz to
12.5 MHz, in the minimum spectral occupancy of frequency hopping spread spectrum systems
operating at the maximum channel bandwidth. SpectraLink indicates that this is necessary
because of the recent rule making actions by the Commission allocating the 915 MHz band to
the Location Monitoring Service (LMS) under Part 90 of our rules. By decreasing the
number of frequency hops, Part 15 spread spectrum systems could avoid operating in the
frequency bands used by wideband multilateration LMS systems, preventing mutual
interference problems. SpectraLink indicates that this change to reduce the spectral
occupancy of spread spectrum systems would maximize the number of Part 15 devices and
LMS users that can coexist in the band, would accommodate the future deployment of
frequency hopping systems in the band, and would maximize spectral efficiency.
28. Comments. All of the commenting parties generally support SpectraLink's
petition. Most parties cite SpectraLink's proposals as an effective means of avoiding the
bands used by wideband multilateration LMS systems and state that such an approach would
avoid potential interference problems both to and from Part 15 spread spectrum systems.
Apple notes that wideband systems utilize virtually the entire 915 MHz band and must hop to,
and transmit in, portions of the band that have been allocated to multilateration LMS systems.
It adds that SpectraLink's proposal to reduce the number of hopping channels, with an
attendant power reduction to retain the same spectral power density, has substantial merit and
few, if any, downside results. Apple also points out that since the rule change provides only
further latitude, not further restrictions, it will not force Part 15 manufacturers who are
content with the current rules to redesign their products.
29. Itron, Inc. requests that the Commission provide a graded power curve based on
the number of hopping channels actually used by the spread spectrum transmitter. It adds that
the power reductions should be correlated linearly to the number of frequencies on which a
device hops. Metricom, however, states that the proposed 500 mW output power limit
doesn't fully address the potential for crowding if many transmitters begin to concentrate in
the remaining portion of the 915 MHz band. It submits that the potential for interference is
exponential, not linear, due to the statistical nature of systems' response to interference
(collisions). Metricom also points out that SpectraLink indicates that its products will operate
indoors. According to the new rules regarding sharing between Part 15 devices and the LMS,
LMS systems are not protected from interference from Part 15 devices that operate indoors.
Thus, Metricom contends that SpectraLink's claim of needing to reduce the number of
hopping channels to avoid interference between Part 15 and LMS devices is not sufficient to
justify changing the rules.
30. Discussion. As SpectraLink observes, there could be mutual interference
problems between wideband, multilateration LMS systems and Part 15 frequency hopping
spread spectrum systems, and it would be beneficial if these two operations could avoid
sharing the same spectrum. The modification sought by SpectraLink would appear to
promote frequency sharing within this band. For these reasons, we believe that the
SpectraLink petition, unlike the petition from Symbol, should be adopted. Therefore, we are
proposing to amend the rules to permit frequency hopping spread spectrum systems in the
915 MHz band to use only 25 hopping channels, provided that those systems employ hopping
channel bandwidths of at least 250 kHz and the transmitters operate at a reduced power level.
Hopping systems using channel bandwidths less than 250 kHz already can avoid operating in
the bands used by broadband multilateration LMS systems and require no decrease in the
minimum number of hopping channels. For frequency hopping systems employing channel
bandwidths of 250 kHz or greater, we propose to reduce the minimum number of hopping
channels to 25. Consistent with this plan, we are also proposing to modify the maximum
average time of occupancy on any hopping frequency to 0.4 seconds in any 10 second period
to correspond to the reduction in the number of hopping channels.
31. While SpectraLink's petition raises some of the same interference concerns as the
Symbol petition, we see several important differences. We note that, unlike Symbol's request,
there would be no increase in the bandwidth of the hopping channel. In addition, the
reduction in the number of hopping channels and the corresponding increase in the average
channel occupancy time would not be as great as under Symbol's proposal. Further, unlike
operation in the 2450 MHz and 5800 MHz bands, we note that multilateration LMS services
are expected to grow at a rapid rate, causing frequency congestion problems for Part 15, LMS
and other users of the 915 MHz band.
32. We also request comments as to whether the rules should specify a formula for
the minimum number of hopping channels based on the amount by which the bandwidth of
the hopping channel exceeds 250 kHz. Under this approach, the minimum number of
hopping frequencies would be equal to 25 x (500/20 dB bandwidth of a single hopping
channel in kHz) or 50 hopping frequencies, whichever results in the lowest number of
hopping frequencies. The use of this formula would result in an even distribution of the
hopping channels over that portion of the 915 MHz band that is not employed by wideband
multilateration LMS systems. This would prevent frequency hopping systems employing
between 250 kHz and 500 kHz bandwidth hopping channels from being concentrated in any
single portion of the 915 MHz band. Adopting this formula would also require that the
average time of occupancy on any hopping frequency not exceed 0.4 seconds within a 20 x
(number of hopping channels/50) second period.
33. In order to reduce the potential for interference due to the smaller number of
hopping channels, we propose to require that frequency hopping spread spectrum systems in
the 915 MHz band that use fewer than 50 hopping channels operate with a maximum peak
transmitter output power of 500 mW. However, as suggested by Metricom, while the
potential for harmful interference can be offset by a reduction in operating power, a linear
reduction may not be sufficient to provide this offset. We recognize that the chance of
collisions with other transmissions, and resulting interference, will be increased since there are
a fewer number of hopping channels resulting in a change to the average time of occupancy
on any frequency and the crowding of transmissions into less spectrum. Accordingly,
comments are requested as to whether or not a greater reduction in output power should be
applied. Comments are also requested on whether a limit on spectral power density, similar
to that currently applied to direct sequence systems, should be applied to frequency hopping
systems operating with less than 50 hopping channels. We also request comments regarding
Itron's suggestion to specify a linear power reduction based on the actual number of hopping
channels employed. This would result in a maximum peak transmitter output power of no
greater than (number of hopping channels/50) watts or 1 watt, whichever is the lesser power.
34. Commenting parties should note that a number of petitions for reconsideration
have been received in response to the recent Report and Order implementing the LMS system.
Any changes to the LMS rules in response to those petitions may result in modifications to
changes for the spread spectrum regulations under Part 15 proposed for the 915 MHz band.
ADDITIONAL PROPOSALS
35. There are several additional regulations concerning Part 15 spread spectrum
transmission systems that need to be clarified, codified or amended. These are discussed
below.
36. Most U-NII proponents support allowing higher power and higher antenna gain in
the U-NII spectrum. They claim that the propagation characteristics at 5 GHz are such that
operation at power levels higher than the proposed limit is required to provide reliable
communications for most local-area networks and for longer-range networks. They state that
the signal attenuation caused by walls is one of the primary reasons why higher power is needed
for LANs. Higher power, they state, is also needed for community networks to achieve reliable
communications over the necessary distances. For example, WINForum argues that, in order to
meet on-premises communication requirements, the maximum transmitter output power limit in
the 5.15-5.35 GHz band should be 100-250 mW (-10 to -6 dBW), and directional antennas
should be permitted with up to 6 dBi gain. It also suggests allowing use of even higher gain
antennas, as long as the transmitter power is decreased on a dB-for-dB adjustment basis
(i.e.,transmitter output power would be decreased by one dB for every dB increase in antenna gain).
WINForum states that these maximum power and gain parameters would provide a desirable
balance between permitting sufficient in-building signal penetration by U-NII devices and
ensuring adequate interference protection to incumbent 5 GHz and other U-NII operations.
37. Apple supports adoption of a maximum transmitter output power of 100 mW (-10
dBW) in the 5.15-5.25 GHz band. Apple states that this power limit would protect incumbent
operations and would allow the lower power band to be used for personal/portable type
operations that would generally operate indoors. Apple also argues for a higher limit on
transmitter output power in the 5.25-5.35 GHz and 5.725-5.825 GHz bands. Specifically, Apple
states that we should set the power limit at 316 mW (-5 dBW), with unrestricted antenna gain in
these bands. It claims higher power is needed in these bands to provide for fixed point-to-point
operations that would meet the requirements of community networks.
38. WINForum also urges the Commission to adopt higher power and antenna gain
limits for the upper band, 5.725-5.825 GHz. It notes that Part 15 spread spectrum devices in this
band are currently authorized to operate with up to 1 W transmitter output power and with up to
6 dBi of antenna gain. Further, it notes that even higher power limits for spread spectrum
devices are currently under consideration by the Commission in ET Docket 96-8.
39. Motorola recommends adoption of a maximum transmitter output power limit of
250 mW in the 5.15-5.35 GHz band, and 1 W in the 5.725-5.825 GHz band, for bandwidths
equal to or greater than a certain threshold, e.g., 25 megahertz. Motorola also supports allowing
transmitter antenna gains of up to 23 dB in both bands, without any associated reduction in
transmitter output power. Motorola argues that directional transmitter antennas will provide
reliable communications with lower risk of interference. It further states that U-NII device
power limitations should be based on the output power spectral density to reduce interference
concerns irrespective of the emission bandwidth. That is, transmitter output power should be
reduced in direct proportion to any reduction in emission bandwidth below some threshold. With
regard to community network links, Motorola recommends that, consistent with the proposal in
ET Docket No. 96-8, the Commission allow the use of even higher gain transmitter antennas
whenever the transmitter output power is reduced by 1 dB for each 3 dB of antenna gain above
23 dB.
40. Mulcay points out that the proposed 100 mW (-10 dBW) EIRP limit is
substantially lower than the European HIPERLAN standard of 1 W (0 dBW) EIRP. Mulcay
states that the maximum transmitter output power limit for U-NII devices should therefore be
raised to 1 W (0 dBW) EIRP to be consistent with the HIPERLAN limit. It claims this would
facilitate U.S. firms' ability to compete in global markets.
41. Parties currently utilizing the 5 GHz spectrum generally support the 100 mW
EIRP limit proposed in the NPRM. NTIA recommends adoption of the 100 mW EIRP limit in
the 5.15-5.25 GHz band to provide adequate interference protection to primary operations in that
band. L/Q adds that U-NII operations at any higher power would degrade the sharing capacity in
that band and would greatly increase their potential to cause harmful interference to FSS
operations in the band. L/Q also opposes allowing directional antenna use by U-NII devices
operating in the band. It contends that though interfering signals from a directional antenna may
not be received by all satellites overhead, they could certainly be received by satellites close to
the horizon and, thus, FSS capacity to operate in the band could be impaired. The ARRL argues
that permitting the power of U-NII operations to exceed 100 mW EIRP or permitting the use of
high-gain antennas by non-spread spectrum U-NII devices would represent a significant
departure from the underlying precepts of Part 15, which require unlicensed operations not to
cause interference to other services. Although the ARRL opposes the operation of U-NII devices
in the 5.725-5.825 GHz band, it states that if a 100 mW EIRP limit and a power spectral density
limitation of 0.03 mW in any 3 kHz bandwidth were adopted, then U-NII devices should be able
to share this band with incumbent operations. Finally, entities with spread spectrum interests
oppose the operation of higher power, non-spread spectrum U-NII devices in the upper band on
the basis that such operations could prevent existing spread spectrum devices from sharing that
band.
42. Decision. We find that the 100 mW power limit proposed in the NPRM is not
sufficient to accommodate the range and scope of communications envisioned for U-NII devices.
We believe that increasing the U-NII device power limits will enable these devices to provide for
a variety of operations including local areas networks, campus-type settings, or as part of
community networks. At the same time, we recognize the need to ensure that primary operations
are adequately protected from harmful interference. In this regard, we note that the primary users
and the considerations that relate to interference with their operations, vary in different parts of
the spectrum we are providing for U-NII devices. Specifically, the 5.15-5.25 GHz band will be
shared with MSS feeder links; the 5.25-5.35 GHz band will be shared with Government
radiolocation operations; and the 5.725-5.825 GHz band will be shared with Government
radiolocation, Amateur, ISM, and other Part 15 operations. Therefore, the sharing environment
for U-NII devices will be different for each of these three 100 megahertz segments. We find a
balance between providing sufficient power limits for U-NII devices and protecting primary
operations may be struck by adopting different power levels for U-NII devices in each of the
three 100 megahertz bands. This approach will provide the needed flexibility to allow U-NII
proponents to design and manufacture equipment to meet a variety of communications needs
while ensuring a successful spectrum sharing environment with other spectrum users.
43. Accordingly, we will divide the 300 megahertz available to U-NII devices into
three bands of 100 megahertz each and will establish the following maximum U-NII device
power limits for each band: a) in the 5.15-5.25 GHz band, the maximum peak transmitter output
power limit will be 50 mW with up to 6 dBi antenna gain permitted, which equates to 200 mW
EIRP; b) in the 5.25-5.35 GHz band, the maximum peak transmitter output power limit will be
250 mW with up to 6 dBi antenna gain permitted, which equates to 1 W EIRP; and c) in the
5.725-5.825 GHz band, the maximum peak transmitter output power limit will be 1 W with up to
6 dBi directional antenna gain permitted, which equates to 4 W EIRP. To permit manufacturers
flexibility in designing U-NII equipment, we will permit the use of higher directional antenna
gain provided there is a corresponding reduction in transmitter output power of one dB for every
dB that the directional antenna gain exceeds 6 dBi.
44. In the 5.15-5.25 GHz sub-band, we believe a 50 mW peak output power with up
to 6 dBi gain antenna will provide U-NII devices great flexibility in how this band is used.
Specifically, these power limits will allow U-NII devices to provide a variety of short-range
communications, such as those between computing devices (such as computers, servers, printers,
etc.) within a very local area, such as in a room or in adjoining rooms. We also believe that
restricting U-NII devices to this low power will allow U-NII devices to share this band with co-
channel MSS feeder link operations. In this regard, we note that the initial European Conference
of Postal and Telecommunications Administrations ("CEPT") studies conclude that HIPERLAN
systems, which have technical characteristics similar to those of U-NII devices, can share this
band with the MSS operations without causing harmful interference to the MSS feeder links.
(See Section E "Spectrum Sharing Considerations" below.) While some commenters have
argued that based on the CEPT studies that U-NII devices could operate at higher powers than
we are adopting without causing interference, we recognize that since the CEPT study was made
Globalstar has changed some of the parameters of its system and that its MSS feeder links
potentially could be more susceptible to interference. In any event, we believe the power we are
adopting is appropriate to ensure that U-NII devices do not cause harmful interference to MSS
feeder link operations. We are also restricting U-NII use of this band to indoor operations. This
will provide additional protection to co-channel MSS operations due to the attenuation of U-NII
device signals as they pass through the walls and ceilings of buildings. Accordingly, we believe
this power limit, along with the restriction on outdoor operations, will provide the desired
balance of providing sufficient power for U-NII devices in this band, high frequency reuse, great
flexibility in the types of U-NII operations that are accommodated in this band, and protection of
co-channel MSS operations.
45. In the 5.25-5.35 GHz sub-band, we are adopting a higher maximum peak
transmitter input power limit of 250 mW, along with the associated higher power spectral density
limit noted below. We are not restricting U-NII devices to indoor operations in this band
because it will not be shared with MSS operations. We believe that U-NII operations with a peak
transmitter output power of up to 250 mW and a directional antenna with up to 6 dBi of gain will
be sufficient to accommodate communications within and between buildings, such as are
envisioned for campus-type LANs. The only operations in this band are Government
radiolocation systems (radar), and NTIA has supported allowing higher power for U-NII
operations in this portion of the band. These power and antenna gain limits are comparable to
the 1 W EIRP limit used for HIPERLAN and therefore should provide manufacturers with
economies of scale in developing equipment useable in both the domestic and international
markets.
46. In the 5.725-5.825 GHz band, we note that spread spectrum Part 15 devices are
already authorized to operate with 1 W transmitter peak output power and with up to 6 dBi gain
transmitting antennas. Accordingly, we are authorizing similar peak power and antenna gain
parameters for U-NII devices in this band. We believe that U-NII operations that comply with
this power limit will be able to provide community networks with a typical range of several
kilometers. Further, we believe that longer-range communications could be possible in areas
with a low interference environment (i.e., rural areas) where high gain receive antennas could be
used. (Such antennas do not affect the transmitted emission level or EIRP.) We recognize that
the commenters recommend that we allow the use of even higher gain transmitting antennas in
this band. However, the record in this proceeding does not provide enough technical support for
us to conclude that U-NII devices with 1 W transmitter power and high gain transmit antennas
would not cause interference to the primary service, Government radiolocation. Specifically,
NTIA has expressed concern about higher powers in this band and supports further
experimentation before either higher power or gain is authorized.
47. In ET Docket No. 96-8, we are currently considering whether to authorize the use
of transmitting antennas with higher gain for Part 15 spread spectrum operations in this band. If
we decide in that proceeding to permit the use of higher antenna gain for spread spectrum
operations, we may consider similar action for U-NII devices in this band in a separate rule
making. However, we note that permitting use of high gain antennas with U-NII devices without
requiring an equal reduction in power could have a significant impact on the interference
environment in this band, and this issue would have to be addressed should a further rule making
be initiated.
48. With regard to sharing this band with amateur operations, we believe that U-NII
devices will cause little interference to amateur operations because of the relatively low power
with which U-NII devices will operate. Further, we note that the amateur service has access to
all spectrum within the 5.65-5.925 GHz range. We therefore believe that amateur operations will
be able to avoid using frequencies within the 5.725-5.825 GHz band that are available to U-NII
devices, in those rare cases where such avoidance may be necessary.
49. Additionally, in all three bands we are adopting peak power spectral density limits
to ensure that the power transmitted by U-NII devices is evenly spread over the emission
bandwidth. Specifically, we will require U-NII devices to decrease transmitter output power
proportionally to any decrease in emission bandwidth below 20 MHz. These requirements will
decrease the potential for interference to other services and will encourage the use of the U-NII
bands for the broadband operations for which they are intended. For U-NII devices operating
with less than 20 megahertz of emission bandwidth, we will limit power spectral density as
follows: a) in the 5.15-5.25 GHz band, the transmitter peak power spectral density will be 2.5
mW/MHz for an antenna gain of 6 dBi; b) in the 5.25-5.35 GHz band, the transmitter peak power
spectral density will be 12.5 mW/MHz for an antenna gain of 6 dBi; and c) in the 5.725-5.825
GHz band, the transmitter peak power spectral density will be 50 mW/MHz for an antenna gain
of 6 dBi. Finally, to allow manufacturers flexibility in designing U-NII devices, we will allow
operations with antenna gains exceeding 6 dBi if the peak power spectral density is reduced by
the same amount the directional antenna gain exceeds 6 dBi.
50. In the 5.15-5.25 GHz band, we will require transmitting antennas to be an integral
part of the U-NII device. This will ensure that our authorized power limits are not exceeded in
this band. In the 5.25-5.35 GHz and 5.725-5.825 GHz bands, we shall require that the U-NII
device use a permanently attached antenna or an antenna that uses a unique coupling to the U-NII
devices in accordance with Section 15.203(a) of the rules. These requirements will limit
potential interference to other systems and will provide for greater frequency reuse by U-NII
devices.
3. Emissions Outside the Band of Operation
51. In the NPRM, we proposed to require that all emissions from U-NII devices
occurring outside of the U-NII bands be attenuated by at least 50 dB or to the radiated emission
limits set forth in Section 15.209, whichever is the lesser attenuation. In addition, we proposed
that any emissions occurring in the restricted bands comply with the radiated emission limits
set forth in Section 15.209. We also proposed to amend Section 15.205 to delete 5.15 - 5.25
GHz as a restricted band. Further, we proposed to require that any unwanted emissions comply
with the general field strength limits set forth in Section 15.209. Finally, for any U-NII devices
that use an AC power line, we proposed to require such devices to comply also with the
conducted limits set forth in Section 15.207.
52. Comments. Only a few parties commented on the emission limits proposed in
the NPRM. Several commenters supported the proposal. For example, Mulcay agrees with the
proposal to limit emissions pursuant to Section 15.209. However, other commenters argue that
the Commission should permit industry to develop limits on emissions that fall outside the bands
of operation. WINForum supports reliance on emission limits and measurement methods that
would be developed by industry and argues that the rules regarding unwanted emissions should
be stated in terms of burst average power and should be independent of the power of the
fundamental emission.
53. Decision. Limits on emission levels outside the bands of operation and
frequency stability requirements are necessary to protect adjacent spectrum occupants and
sensitive operations that may operate on harmonic frequencies. However, in view of the higher
and different power limits we are adopting for U-NII devices in each of these bands, we are
making appropriate adjustments to the limits we proposed in the NPRM on the permissible
emission levels outside the band. Specifically, we will require U-NII devices operating in the
upper band to attenuate emissions below the maximum power spectral density by a factor of at
least 40 dB for frequencies from the band edge to 10 megahertz from the band edge and by a
factor of at least 50 dB for frequencies greater than 10 megahertz from the band edge. For the
other two bands which have lower maximum power limits we will take this limit as an absolute
limit. This will provide the same level of interference protection outside all three bands.
Accordingly, the attenuation of peak levels of emissions outside of the frequency bands of
operation below the maximum peak power spectral density contained within the bands of
operation must be in accordance with the following limits:
i) For transmitters operating in the band 5.15-5.25 GHz: all emissions within the
frequency range 5.14-5.15 GHz and 5.35-5.36 GHz must be attenuated by a factor of at
least 27 dB; within the frequency range outside these bands by a factor of at least 37 dB.
ii) For transmitters operating in the 5.25-5.35 GHz band: all emissions within the
frequency range from the band edge to 10 MHz above or below the band edge must be
attenuated by a factor of at least 34 dB; for frequencies 10 MHz or greater above or below
the band edge by a factor of at least 44 dB.
iii) For transmitters operating in the 5.725-5.825 GHz band: all emissions within the
frequency range from the band edge to 10 MHz above or below the band edge must be
attenuated by a factor of at least 40 dB; for frequencies 10 MHz or greater above or
below the band edge by a factor of at least 50 dB.
As already specified in the rules, the measurements of such emissions shall be performed using a
minimum resolution bandwidth of 1 MHz. Regardless of the attenuation levels shown above,
we will not require emissions outside the frequency range of operation to be attenuated below the
general radiated emission limits in Section 15.209 of our rules. Further, we will not specify
these emission limits as a maximum power spectral density of the operating band, as requested
by WINForum, because such a limit would have to be adjusted with changes in antenna gain in
order to maintain a consistent interference potential. The emission limits being adopted are
based on the peak power spectral density within the band of operation, and the power spectral
density is varied to reflect changes in the gain of the antenna. We recognize that changes to the
gain of the antenna at harmonic frequencies may not directly correlate with changes to the
antenna gain at the fundamental frequency. However, we believe that the limits being adopted
for spurious emissions are sufficient to reduce the probability of harmful interference. Further,
the provisions in Section 15.205 of our rules will ensure that harmful interference does not result
to critical safety services regardless of antenna gain.
54. Further, we will adopt our proposal to remove the 5.15-5.25 GHz band from the
restricted bands listed at Section 15.205 of the rules. We note that U-NII devices will have to
comply with the provisions of Section 15.205 in order to protect sensitive operations. We also
note that the 4.5-5.15 GHz and 5.35-5.46 GHz bands remain restricted; therefore, U-NII devices
operating close to the band edges at 5.15 GHz or 5.35 GHz will be required to sharply attenuate
their signal at the band edge or avoid using the spectrum close to the band edge. We do not
believe that this requirement will significantly affect U-NII operations overall. In any event, this
requirement is needed to protect sensitive and safety-of-life operations in adjacent bands.
Additionally, we adopt our proposal to require that emissions comply with the general field
strength limits set forth in Section 15.209. Finally, any U-NII devices that use an AC power line
must comply with the conducted limits set forth in Section 15.207.
4. Channeling Plan & Modulation Efficiency
55. In the NPRM, we did not propose to adopt limits on channelization or modulation
efficiency, but did request comment on whether we should specify a channeling plan or a
minimum modulation efficiency requirement for U-NII devices to ensure efficient use of the
spectrum. We specifically requested comments on whether a 20 or 25 MHz channeling plan
and/or a 1 bit/second/Hz ("bps/Hz") modulation efficiency should be adopted and whether these
regulations would be beneficial in facilitating unlicensed broadband high data rate use of these
bands.
56. Comments. Commenting parties disagree as to whether the Commission should
adopt a mandatory channelization plan or minimum bandwidth requirement for unlicensed U-NII
devices. Those supporting a channelization plan and/or minimum bandwidth requirement argue
that the 5 GHz unlicensed bands should be dedicated for wideband systems. These parties
contend that other unlicensed bands such as the U-PCS spectrum are already available for
narrower bandwidth applications. They claim that some channel limitations are needed. They
contend that wideband devices with high signaling speed requirements suffer disproportionately
from harmful interference caused by narrow bandwidth devices with low signaling speed
requirements, and that complex rules would be required to correct this imbalance. WINForum,
Lucent, and Nortel support a minimum channel spacing of 20-25 megahertz and suggest that the
rules not prohibit U-NII devices from combining channels to enable very wide bandwidth
communications. In addition, Nortel states that a 20-25 megahertz minimum channel
bandwidth would simplify any industry-developed access protocol by limiting the number of
channels that would need to be scanned in order to detect the absence of communications from
other devices before transmitting. Further, it notes that such a channeling plan would enable U-
NII devices to be compatible with HIPERLAN equipment.
57. On the other hand, several NII proponents argue that, with so many open
questions about the future needs for unlicensed wireless networking capabilities, it is premature
and technically unwise to specify a channeling plan or a maximum channel bandwidth. The
channelization for these bands should be flexible, they state, because the bandwidth required for
a given application is dependent on the data rate, communications distance, type of modulation,
and specific error correction coding involved. They also claim that a minimum channel width or
channelization requirement may limit both technical innovation and flexibility and therefore may
increase costs and retard development of new communications options. Finally, although
Apple opposes mandatory channelization standards, it states that, in the bands used for high data
rate systems, voluntary channelization plans or more informal channelization etiquettes could be
developed by industry to promote efficient spectrum use.
58. Most U-NII proponents oppose the imposition of any requirement for modulation
efficiency. They claim that such a requirement would increase system complexity and preclude
certain modulation techniques, which would in turn increase costs and development time, and
delay implementation of U-NII devices. Several commenters oppose the 1 bps/Hz modulation
efficiency suggested in the NPRM on the grounds that it would preclude spectrum efficient
technologies such as spread spectrum, which they observe is spectrally efficient because of its
high interference rejection and spectral reuse but may not meet a 1 bps/Hz requirement.
Further, several parties claim that efficiency can only be measured meaningfully when
geographic frequency re-use (cell area) is also considered, such as bps/Hz/unit-area. These
parties argue that a robust system with low modulation efficiency that is capable of operating in
the presence of higher potential interference may nevertheless have higher throughput per unit
area than a less robust system. Finally, they argue that it is unnecessary to mandate a standard
for spectrum efficiency, since the market will decide what efficiency is needed. In this regard,
WINForum recommends forgoing the adoption of a modulation efficiency standard at this time
in favor of allowing industry groups to consider the development of a more flexible spectral
efficiency measure that would take into account frequency reuse characteristics.
59. A few U-NII proponents do support adoption of a modulation efficiency standard.
For example, Hewlett-Packard Company ("Hewlett-Packard") recommends a minimum
modulation efficiency standard of 0.66 bps/Hz, arguing that, though specification of a high
bandwidth efficiency does not guarantee a high spectral efficiency, it can nevertheless prevent
low transmission rate systems from using the spectrum inefficiently. Lucent recommends a
minimum modulation efficiency standard of 0.5 bps/Hz based on the use of a 3-dB bandwidth, as
opposed to use of the full emission bandwidth. If, however, the emission bandwidth were used,
Lucent recommends a higher minimum modulation efficiency standard. NTIA also recommends
adoption of a bandwidth efficiency standard, but claims that imposition of a strict efficiency
limitation at the outset may dampen rapid implementation. Therefore, NTIA recommends that
the Commission adopt an effective bandwidth efficiency standard that would come into effect at
some reasonable future date, such as three years after conclusion of this rule making
proceeding.
60. Finally, some parties, particularly incumbents, argue that a modulation efficiency
standard should be required in order to ensure that spectrum is not wasted. They state that highly
efficient technologies currently exist and that it is not unreasonable to require U-NII devices to
have modulation efficiencies higher than 1 bps/Hz. Alstatt Associates, for example, argues that,
since digital television set-top boxes have a modulation efficiency of 6.66 bps/Hz, and Part 21
and 94 devices have a minimum modulation efficiency of 4.46 bps/Hz, U-NII devices should be
required to have a minimum efficiency of 3 bps/Hz. Larus Corporation ("Larus") agrees that
we should adopt a modulation efficiency standard of no less than 3 bps/Hz, while the Northern
Amateur Relay Council of California, Inc. ("NARCC") argues that a spectrum efficiency of 2
bps/Hz is appropriate and has in fact been achieved for years.
61. Decision. One of our goals in this proceeding is to provide rules which
permit maximum technical flexibility in the design and development of U-NII devices capable of
providing high data rate communications for a variety of multimedia applications in a shared
spectrum environment. Such devices have not yet been designed, built, or tested. Accordingly,
we believe that adopting a rigid channelization plan or mandating a modulation efficiency
standard at this time would not meet this goal, and could delay implementation of U-NII devices
by precluding certain technologies or applications. Further, we believe that the low power limits
we are adopting will ensure efficient use of the spectrum by providing for high frequency reuse,
which will allow for large numbers of U-NII devices to share the spectrum in any geographic
area. We also believe that establishing a channelization plan or modulation efficiency at this
early stage in the technological development of the devices might have several undesirable
effects, such as increasing costs and delaying the benefits of U-NII devices to the public.
Accordingly, we will not adopt a channeling plan or a modulation efficiency standard at this
time.
62. Nevertheless, we note that the focus of this proceeding is to make available
spectrum for broadband high data rate unlicensed devices capable of meeting the
communications requirements of new multimedia applications. We therefore agree with those
commenting parties that suggest the purpose of making these bands available is to support use of
high data rate devices. Accordingly, we are adopting a definition for the type of devices that will
be approved for this band. Specifically, the Part 15 rules will state that unlicensed U-NII
operations in the 5.15-5.35 GHz and 5.725-5.825 GHz bands will be limited to wide bandwidth,
high data rate digital operations. Unlicensed devices accessing the 5.725-5.825 GHz band under
other Part 15 rules would not be subject to this definition. This will give equipment
manufacturers the flexibility to design and manufacture a variety of broadband devices using
different technologies and modulation techniques, while ensuring that this spectrum is used for
its intended purpose. This definition will be enforced through the Commission's equipment
certification process.
D. Spectrum Etiquette
63. In the NPRM, we proposed a basic "listen-before-talk"
("LBT") spectrum sharing
etiquette, similar to that established for U-PCS devices, to ensure that the U-NII spectrum is
used by devices in a manner that would permit them to share with one another. We suggested
that the proposed etiquette could serve as an interim protocol standard until industry developed a
spectrum sharing etiquette. In this regard, the NPRM encouraged industry to develop appropriate
etiquette protocols for these devices through a consensus process and stated that, if appropriate,
we would consider those protocols in this or a further rule making proceeding. Finally, we
solicited comments on whether such an etiquette should be required at all, or whether the
minimal technical requirements would be sufficient to ensure spectrum sharing among U-NII
devices.
64. Comments. The commenters overwhelmingly oppose the LBT spectrum
etiquette proposed in the NPRM for U-NII devices. Several parties argue that the LBT protocol
is unnecessary and would be detrimental to U-NII devices at 5 GHz. For example, Motorola
states that LBT would be ineffective in controlling interference among U-NII devices,
particularly in buildings with many rooms and hallways. Several commenters also assert that
LBT would be detrimental because it would preclude isochronous multimedia applications and
other technologies such as Asynchronous Transfer Mode ("ATM") that would not be able to
comply with strict transmission time-frame requirements. Motorola claims that LBT would
unduly restrict the utilitarian choices which manufacturers of U-NII devices could offer to
consumers.
65. Additionally, some commenters oppose establishing any interim etiquette on the
grounds that devices developed under such an interim etiquette could be rendered useless once a
permanent etiquette is adopted. In this regard, Lucent asserts that adoption of an interim
etiquette would hinder introduction of future systems and would inhibit the process of
developing an industry consensus for spectrum sharing rules. Cylink contends that interim rules
would harm the competitiveness of small businesses. It claims that only larger companies could
afford to cover the risk of betting on the eventual outcome of industry working group
deliberations aimed at adopting a consensual etiquette. Similarly, WINForum expresses
concerns about the compatibility of interim devices with any subsequent permanent spectrum
etiquette and suggests that any interim operations should be constrained to 50 megahertz in the
upper band with a date-certain changeover mandate.
66. Several commenters, while not supporting the proposed LBT etiquette, do support
the development of a spectrum etiquette, or of multiple etiquettes, by industry consensus in order
to help minimize interference among U-NII devices. For example, WINForum states that
high-level protocols, like that adopted for U-PCS, may be excessively complex for U-NII
devices, but simple RF rules (e.g., power limits, channelization, unwanted emission limits) may
prove insufficient to ensure fair, efficient, and open access. It adds that the development of
such standards is appropriately left to voluntary standards organizations. NTIA also supports
some type of channel monitoring protocol or U-NII etiquette to minimize interference, both to
and from radar systems. Further, several other commenters aver that industry should be
permitted to develop etiquettes within a time frame mandated by the Commission.
67. Some commenters oppose the adoption of any spectrum etiquette, stating that a
required etiquette would inevitably limit innovation in the development of new U-NII products,
and that the use of etiquettes has not always been proven to avoid interference. The Connectivity
for Learning Coalition ("Coalition") asserts that while such protocols mandate a manner in which
some technologies may share the spectrum, use of those technologies may or may not meet the
needs of the education or library communities. Metricom, Inc. ("Metricom") states that, in
theory, an etiquette may appear to allow for spectrum sharing, but there is no practical evidence
that complex etiquettes prevent interference. Metricom states that creative engineers guided
by minimal technical standards will best be able to design communications solutions to match
consumer needs. Finally, 3Com Corporation ("3Com") claims that a formal spectrum etiquette
would limit ingenuity and development of U-NII devices, and it urges the Commission to
encourage the development of voluntary spectrum etiquettes to permit interoperability.
68. Decision. In general, we believe that a spectrum etiquette can provide
benefits by facilitating compatibility among devices and allowing for equal access to the
spectrum by devices that use different technologies. However, we do not believe that the interim
LBT etiquette proposed in the NPRM would provide such benefits for unlicensed U-NII devices
in the 5 GHz band. As pointed out in the comments, that LBT etiquette would be ineffective in
controlling interference among devices and would preclude some technologies that may be
desirable for U-NII devices. Accordingly we will not adopt our proposed etiquette.
69. We also note that the record does not provide an alternative spectrum etiquette to
our proposed LBT etiquette that we could adopt at this time. Additionally, we do not think that it
would be in the public interest to wait for an industry group to develop a satisfactory new
etiquette as suggested by WINForum. We are concerned that it could take industry a substantial
period of time to develop an etiquette for unlicensed U-NII devices at 5 GHz, because of the
wide range of interests that would have to be accommodated in establishing a single etiquette for
all the broad multimedia applications envisioned for U-NII devices. Further, after such an
etiquette is developed, we would have to conduct a rule making proceeding to adopt the etiquette
as a mandatory standard. Completing these activities would take at least a year, and possibly
considerably longer. We do not believe that such a delay in implementing rules permitting U-NII
operations would serve the public interest.
70. We recognize that there are trade-offs in adopting any etiquette and that the
benefits of an etiquette must be weighed against its drawbacks. For example, an etiquette could
beneficially facilitate compatibility among devices and thus promote spectrum sharing, inter-
communications among different devices, and equal access to the spectrum by devices built by
various manufacturers. Drawbacks of an etiquette include an increase in the complexity of
equipment design and, hence, an increase in cost to the manufacturer and the user, as well as a
potential limitation on access to the spectrum by some technologies and equipment. In the
instant case, it is early in the developmental stage for equipment to operate in these shared bands
as intended. Therefore it may be very difficult to develop a spectrum etiquette at this time that
will not limit the types of equipment that could most efficiently or effectively provide the desired
broadband communications.
71. Accordingly, we are not adopting a spectrum sharing etiquette at this time, nor
will we delay access to the 5 GHz bands by U-NII devices until industry develops an etiquette.
We believe the minimal technical rules we are adopting, particularly the maximum power limits
discussed above, will generally allow for equal access and sharing of these bands by U-NII
devices and thereby accomplish the intent of our proposed spectrum etiquette. Finally, our
course of action will not preclude industry from developing any voluntary standards that it deems
appropriate in the future. In this regard, we continue to encourage industry to develop
appropriate etiquette protocols through a cooperative consensus process. If standards are
developed that would better facilitate sharing of this band without precluding U-NII devices or
technologies, we would consider adopting those protocols in a further rule making proceeding.
We note that WINForum states that it has already begun setting the foundation for joint industry
action in this area. We encourage all interested parties to take part in this process and to
cooperate in good faith.
E. Spectrum Sharing Considerations
72. In the NPRM, we recognized that a number of primary services now use, or soon
will use, the spectrum which we proposed for U-NII devices. The existing operations include
Government radiolocation systems; mobile satellite feeder links; amateur operations; industrial,
scientific, and medical operations; other unlicensed Part 15 operations; and proposed ITS. We
stated in the NPRM that it would be necessary to develop spectrum sharing criteria between
primary operations and the new U-NII devices. We tentatively concluded in the NPRM that
sharing would be feasible, but requested comments on this issue.
73. Comments. NTIA, the Government agency responsible for the spectrum
management for Government operations, supports our proposal to permit U-NII devices to share
these bands with primary Government operations. However, NTIA urges us to adopt sharing
protocols and power limitations to facilitate sharing. NTIA states that the success of community
networks will depend on their geographic separation from high powered radar systems operating
in these bands. NTIA adds that Federal radar systems serve the interests of national security and
that, therefore, all efforts should be made to avoid operating community network links near
military radar sites. Additionally, NTIA states that compatibility analyses of long range links
with existing radar needs to be completed for both U-NII and spread spectrum systems before
higher powers are authorized.
74. In the 5.15-5.25 GHz band, parties with MSS interests argue that sharing is not
feasible between MSS feeder links and new U-NII devices. In particular, L/Q asserts that only
1070 simultaneous users of U-NII devices could operate in the 5.15-5.25 GHz segment in the
continental United States before unacceptable interference would be caused to the feeder links
for Globalstar, its proposed mobile satellite system. Airtouch Communications, Inc.
("Airtouch"), a limited partner in Globalstar, claims that its analysis indicates that U-NII
operations in the 5.15-5.25 GHz band would reduce the capacity of Globalstar in the U.S. by
over 27%. Further, Airtouch and L/Q argue that the European sharing analysis for
HIPERLAN, addressed in the NPRM, cannot be applied in this proceeding because U-NII
devices and HIPERLAN do not have similar technical parameters and the International
Telecommunication Union ("ITU") analysis is not based on current data.
75. On the other hand, U-NII proponents claim that U-NII devices would be able to
share with Government radiolocation and MSS feeder uplinks operations because of the very low
power with which U-NII devices will operate in the 5.15-5.25 GHz band and because of the
attenuation characteristics of radio signals in the 5 GHz range. With regard to sharing with MSS
feeder uplink operations, they conclude that MSS feeder operations would also be able to share
with U-NII devices. They base this conclusion on the ITU study, which predicted that
HIPERLAN systems would be able to co-exist with the MSS feeder links in the 5.15-5.25 GHz
band in Europe. Further, they note that HIPERLAN devices will be operating at 1 W, a power
level substantially higher than the power limit proposed for U-NII devices in this band, and that
global MSS systems must be built to be robust enough to share with HIPERLAN. WINForum
and Solectek Corporation ("Solectek") also counter L/Q's claim that only 1070 U-NII devices
could use the band in the United States before causing interference to MSS feeder links, pointing
out that L/Q made strict assumptions that are not representative of the U-NII device
environment. Specifically, WINForum notes that ITU's studies assume a more reasonable
performance margin of 0.41 dB rather than the 0.004 dB that L/Q used. Based upon these more
realistic criteria, WINForum claims that over 540 million U-NII devices could be deployed in the
United States without causing harmful interference to the FSS uplink systems.
76. Additionally, some parties argue that use of directional antennas will decrease the
radiation perceived by a satellite above the users in the vertical plane. Further, Lace, Inc.
("Lace") argues that a 10% power increase in the MSS feeder link would easily resolve the
interference problem, if indeed that problem ever occurs. Lace and Solectek argue that there
are other means to mitigate interference such as power spectral density limits, transmitter on-time
limits, station antenna directivity, relay link antenna directivity, out-of-band noise rejection, and
positioning long range outdoor links above 5.25 GHz.
77. In the 5.725-5.825 GHz band, incumbent operators either oppose allowing U-NII
operation due to interference concerns or urge that sharing studies be completed before that band
is made available to U-NII devices. For example, the San Bernardino Microwave Society
("SBMS") argues that sharing between U-NII devices and amateur operations is not possible.
However, the ARRL states that the proposed U-NII maximum power limit of 100 mW EIRP
appears to be sufficient to avoid significant interference to the amateur service, but it argues that
the ubiquitous nature, mobility, and potential aggregate interference potential of these devices
necessitates that sharing studies be performed. Additionally, Section 15.247 spread spectrum
interests oppose U-NII operations in this band and argue that without a means to control usage,
operations in this band would rapidly degrade and become unusable. Further, incumbents
oppose high power U-NII operations in this band because they argue it is more likely to cause
interference to incumbent operations. The ARRL also claims that higher power U-NII
operations should not be permitted because the Commission is unlikely to enforce the
requirement that unlicensed device users cease operation if they are causing harmful interference
to allocated services. Finally, Metricom states that to avoid interference, U-NII devices in the
upper band should be required to operate in spread spectrum mode.
78. On the other hand, U-NII proponents argue that U-NII devices can share with
Government radiolocation, amateur operations, ISM devices, other Part 15 devices and proposed
ITS operations. They argue that these incumbent operations already share this band with other
types of unlicensed devices. In this regard, Mulcay notes that a substantial number of devices,
including Part 15 direct sequence spread spectrum radios with 1 W output power and antenna
gains of 30 dBi, frequency hopping radios with omni-directional antennas and non-
communication devices under Part 18 with no limit on radiated power, already share the 2.4 and
5.8 GHz bands on an unlicensed, non-coordinated basis without causing interference.
WINForum likewise argues that U-NII devices operating under equivalent technical standards
can also share this band without causing harmful interference. Apple notes that U-NII devices
will only share a part of the 275 megahertz wide amateur band at 5.65-5.925 GHz and, therefore,
claim that U-NII operations will not significantly affect the amateur radio service. With regard
to sharing with other Part 15 devices, U-NII proponents contend that the record demonstrates that
U-NII devices can be designed to coexist with spread spectrum devices. CEMA argues that
industry can develop technical guidelines and methodologies to allow community network
systems and other systems to share unlicensed bands. Further, Apple claims that directional
antennas will reduce the probability that multiple devices will compete for spectrum in
overlapping areas.
79. Decision. We continue to believe that U-NII devices can share these bands
with existing and future operations. Specifically, we believe that the power limits, power
spectral density requirements and emission limits that we are adopting herein will permit the
robust development of U-NII devices without a significant impact on other spectrum users. With
regard to Government operations, we agree with NTIA that MLS operations below 5.15 GHz
must be protected. Accordingly, we are not allowing U-NII devices access to spectrum below
5.15 GHz. We believe that this decision, along with the power limits and out-of-band emission
limits, will adequately protect MLS operations. We also agree with NTIA that co-channel
sharing with Government radiolocation is possible. We believe the power limits we are adopting
will allow for this sharing as detailed below. Further, we believe that there will be no
interference from U-NII devices to possible ITS operations, since we are not allowing U-NII
devices access to the 5.85-5.875 GHz band.
80. In the 5.15-5.25 GHz band, we note that the sharing analyses completed to date
often reach different results because they are based on different assumptions. For example, since
the CEPT studies were made, Globalstar has changed some of the parameters of its system, and,
therefore, MSS feeder links potentially could be more susceptible to interference than those
studies concluded. Based upon the information before us, we conclude that the limits we are
adopting will ensure that U-NII devices do not cause harmful interference to MSS feeder link
operations.
81. In the 5.25-5.35 GHz band, we believe that the 1 W EIRP limit and the power
spectral density requirements we are adopting for U-NII devices will adequately protect the
primary radiolocation operations. We note that Government radiolocation systems are limited in
number and generally located at remote military sites, on board ships, in aircraft and in
spacecraft, and that these considerations in conjunction with the U-NII power limits should
adequately protect the radiolocation service. Further, U-NII devices will have to accept
interference from the radiolocation service.
82. In the 5.725-5.825 GHz band, we believe that the 4 W EIRP limit and the power
spectral density requirements we are adopting for U-NII devices will adequately protect the
primary radiolocation operations and amateur operations. These limits provide U-NII devices
with power levels equivalent to Part 15 spread spectrum devices that already share this band with
incumbent services. Therefore, U-NII devices should likewise be able to share this band without
causing interference to the primary services. Further, with regard to spectrum sharing with the
amateur service, we note that the amateur service has access to 275 megahertz of spectrum in the
5.65-5.925 GHz band. We believe amateur licensees will, if necessary, be able to operate around
U-NII devices, which only have access to 100 megahertz in this portion of the 5 GHz spectrum.
Additionally, we note that we are not at this time providing spectrum above 5.825 GHz for U-NII
devices. This eliminates any sharing concerns with users or potential users of the 5.825-5.875
GHz band, which includes lower power Part 15 devices such as hearing aid devices, as well as
ITS operations, and FSS operations.
83. We also believe our power spectral density requirements will permit U-NII
devices to share this spectrum with unlicensed spread spectrum devices as the potential for
interference to these devices from new U-NII devices will be no greater than that which would be
expected from additional spread spectrum devices. Thus we see no reason to restrict U-NII
devices in this band to spread spectrum technologies as requested by some commenters.
Accordingly, we will allow U-NII devices in this band to operate on a technology-neutral basis.
We believe this will provide manufacturers flexibility in designing U-NII products and thus will
provide consumers with greater choices.
F. Alternative Regulatory Structure
84. In the NPRM, we proposed to allow U-NII devices to operate on an unlicensed
basis. We tentatively concluded that the low power and limited range of U-NII devices would
make licensing administratively difficult for users and the Commission. Further, we noted that
this spectrum may be of very limited use to licensed services due to the presence of incumbent
operations. Nonetheless, we requested comment on whether new U-NII operations should be
provided on a licensed basis. We also solicited comments with regard to whether we should
license higher power community networks if we were to allow such operations. We also asked
whether, in the case of mutually-exclusive applications, we should use competitive bidding to
award such licenses.
85. Comments. Most U-NII proponents support our proposal to allow U-NII
devices to operate on an unlicensed basis. They oppose licensing and auctions of any U-NII
operations, arguing that the benefits of authorizing Part 15 devices and systems would be
undermined completely if licensing were required. They state that licensing -- even expedited
licensing -- would impede deployment, reduce innovation, reduce spectrum efficiency, increase
costs, undermine the development of community networking and deny the benefits of low cost
and flexible alternatives to existing media. Apple also contends that unlicensed community
networks would not create problems of regulatory parity for common carriers and other profit
making service providers. It states that those electing to use unlicensed bands would accept the
fact that they will not control their spectrum environment and will be limited to low power
operations; in exchange, they would be freed from the costs and burdens associated with
licensing.
86. However, AT&T, PacTel, TIA and some microwave equipment manufacturers
state that if the Commission permits the operation of longer range community networks, those
networks should be subject to licensing and auctions. AT&T states that allowing unlicensed
community networks would be unfair to the holders of existing spectrum licenses, particularly
those who received their licenses through the auction process, and would undermine the
Congressional objective of promoting regulatory parity among wireless services. AT&T
further states that the type of operation envisioned for community networks requires a degree of
reliability and quality that can only be realized through licensed services. PacTel argues that
unlicensed community networks would create an inequitable regulatory structure where
unlicensed service providers operate in competition with licensed service providers without the
common carrier obligations of a licensee. TIA states that implementing long range networks
requires frequency coordination, use of narrow beam antennas and other fundamental
components of licensing in order to succeed.
87. Decision. We continue to believe that low power U-NII devices and
associated operations are more amenable to an unlicensed structure and should be regulated
under the Part 15 rules. Specifically, the rules governing U-NII devices are similar in their low
power and flexible regulatory nature to those governing Part 15 devices. While some U-NII
devices in the upper band could have ranges of several kilometers, we believe that most devices
will have typical communication ranges of a few meters to a few hundred meters. Additionally,
like other existing unlicensed devices, we believe that trying to license U-NII devices
individually would be administratively difficult if not impossible for both the Commission and
the consumer and would greatly delay the implementation and use of this band by U-NII devices.
Further, we do not think it would be advisable at this time to license spectrum blocks and large
service areas to providers.
88. We also are unpersuaded by the arguments that U-NII devices and associated
operations need to be licensed in order to provide regulatory parity with licensed services. With
regard to unlicensed U-NII devices that are used for community networks in the upper band, we
note that these will also be of very limited range in comparison to the distances of fixed point-to-
point operations, will have to operate in a Part 15 sufferance mode and may not always be able to
provide the same grade of service as the licensed operations. That is, they will receive no
protection from other users of the spectrum. Further, we note that in the upper band unlicensed
devices are already providing point-to-point links for data transmissions, typically of up to 1.5
Mbits/sec. Further, we believe that the vast majority of U-NII devices will provide
communications that are complementary to, rather than competitive with, the licensed services.
We believe that the relationship between U-NII devices and the licensed point-to-point services
will be analogous to the relationship between cordless telephones and PCS or the cellular
telephone service. That is, U-NII devices will provide a variety of broadband high data rate
services but only in a very limited range and generally on the premises of the users, while
licensed fixed point-to-point microwave services provide communication links that are
substantially longer, up to 30 and 40 miles, and in a controlled radio environment where the
licensee has the right of protection from interference.
89. We do believe, however, that this proceeding has raised a number of spectrum
issues that warrant further attention. Users and manufacturers of unlicensed devices, for
example, may have little incentive to make the investment necessary to improve spectrum
efficiency and thus allow more users to benefit. As we continue to implement spectrum policies
that promote competition and efficiency we may also need to consider how to harmonize these
policies with those for unlicensed devices.
G. New Part 16 Regulations
90. In the NPRM, we tentatively concluded that the technical and operational flexibility
afforded under Part 15 is the appropriate structure for regulating U-NII devices, rather than a new
Part 16 regulatory scheme. Under the Part 16 concept, unlicensed devices could be treated as a
recognized radio service with spectrum rights, including interference protection. Alternatively,
we proposed to establish a "safe harbor" or clear technical operating parameters under which
users of unlicensed U-NII devices could operate without being considered sources of harmful
interference. Consistent with Part 15 operations, we also proposed that U-NII devices have to
accept any interference.
91. Comments. NII proponents support the principles underlying either "safe
harbor" or the Part 16 approach. Apple argues that for U-NII devices to become viable, these
devices must be treated as a recognized radio service, and their operations must be in protected
spectrum reflected in Section 2.106 of the rules, the Table of Frequency Allocations. Further,
Apple states that the Commission should make clear that it will not introduce new, incompatible
services into the NII bands in the future. Apple argues this is fully consistent with both the
Communications Act and Commission precedent; in that, it is identical to the approach adopted
for unlicensed-PCS and millimeter wave bands. Further, Apple argues that this approach is
consistent with the Commission's obligation under Section 303(g) of the Communications Act to
'study new uses for radio... and generally encourage the larger and more effective use of radio in
the public interest.' CEMA also argues that the Commission has the authority to elevate the
status of unlicensed devices and suggests upgrading the status of U-NII devices to co-primary
within the allocated bands. It claims that otherwise these devices will remain, by regulatory
design, second class citizens in the RF environment. Further, WINForum claims that some
rural and educational users may not be willing to risk investment in equipment absent some
reassurance that their communication needs will be met now and in the future. Finally,
Cylink, Metricom and existing unlicensed spread spectrum device providers argue that if
additional protection is provided to unlicensed U-NII devices in this range of the spectrum, then
this protection also should be extend to unlicensed spread spectrum devices.
92. However, Airtouch and other parties with interests in the 5.15-5.35 GHz and
5.725-5.825 GHz bands state that the "safe harbor" concept conflicts with the Part 15 regulatory
scheme and would relieve unlicensed users of their obligation to avoid interference to licensed
users. L/Q argues that an analogy cannot be made to the protection provided to unlicensed
Data-PCS devices because unlicensed Data-PCS devices received an allocation of exclusive
spectrum, but U-NII devices will not operate on exclusive spectrum. SBMS and other amateur
interests oppose "safe harbor" rules because there will be no means of enforcement to prevent U-
NII devices from causing interference.
93. Decision. We generally have provided spectrum for low power unlicensed
devices on a non-interference basis, meaning that unlicensed devices must not cause interference
to licensed users and must accept any interference they receive. This regulatory approach to
accommodating unlicensed devices has protected licensed use while permitting the development
of a wide variety of low power unlicensed devices. While we seek to encourage the important
and valuable telecommunication operations which will be provided by U-NII devices, we find
that the current record does not provide a compelling reason to believe that such devices require
higher or more protected status than we have provided for low power unlicensed devices in the
past. Accordingly, we do not believe that it is necessary to create a new Part 16 or "safe harbor"
rules to provide additional protection for U-NII devices. We therefore, as discussed below, will
regulate these devices in the same manner that we regulate other low-power unlicensed devices.
We do conclude, however, that some special consideration is warranted with regard to the use of
unlicensed devices in the lower band, 5.15-5.25 GHz, which will be shared with MSS.
94. In the 5.25-5.35 GHz and 5.725-5.825 GHz bands, where the radio environment is
well established with mature services, we can adopt rules in Part 15 for U-NII devices in which
all parties can have confidence that sharing is possible with little or no threat of interference. In
both of these bands, we believe U-NII device manufacturers and users can feel confident that
their operations will not cause interference to primary operations, because in the 5.25-5.35 GHz
band the only party authorized to use this spectrum is Government radar operations, with which
we believe low power U-NII devices can share spectrum without causing interference, and
because the U-NII devices in the 5.725-5.825 GHz band will operate with powers equivalent to
those of existing unlicensed operations that currently share this band without causing
interference. Additionally, if interference problems did occur in these bands they would be
localized and could probably be identified and resolved. In these cases we believe that the
current Part 15 at sufferance rules are appropriate.
95. We recognize that it is likely that two new uses of the 5.15-5.25 GHz band, MSS
feeder link operations and U-NII devices, will be developing at the same time. In view of this
fact, as indicated above, we are adopting relatively conservative operating parameters for U-NII
devices. We believe that the very low power limits and indoor use restriction on unlicensed
operations will ensure that millions, or even tens of millions, of U-NII devices can successfully
co-exist and share the spectrum with MSS feeder links. Further, we note that interference from
U-NII devices to MSS operations could potentially occur only as a result of the cumulative effect
of many millions of U-NII devices and not by any single device. To the receiver on the MSS
satellite, the operation of many low power U-NII devices looks like an increase in the ambient
noise floor. This has the effect of decreasing the desired signal-to-noise ratio received from the
higher power MSS feeder link and can ultimately reduce the capacity of or cause interference to
MSS operations.
96. While we believe that this approach for U-NII devices is technically conservative
and will fully protect MSS operations, we note that MSS interests have also suggested that we
limit the aggregate EIRP density of emissions from unlicensed devices on the Earth's surface to
the MSS satellite to 10 dBW/MHz. They argue that MSS operations could begin to be affected
when emissions from unlicensed devices approach such a level. Alternatively, they suggest
that the Commission should review the technical parameters for U-NII operations in a future rule
making as such a limit is approached. They state this would allow the Commission to review, for
example, whether some future reduction in permitted power of U-NII devices in this band should
be imposed. They state that all existing U-NII devices would be grandfathered. We concur that
such an approach would provide further assurance that future potential conflicts between U-NII
devices and MSS operations are taken into account and that MSS operations are protected
appropriately. Accordingly, we invite MSS parties to monitor the emissions from U-NII devices
in the 5.15-5.25 GHz band and, if emissions approach the 10 dBW/MHz level, to request that we
initiate a rule making to reassess the use of this band. At that time the Commission could
determine if future U-NII devices should be required to operate at different technical standards.
In this regard, we note that it may also be appropriate to reassess the technical parameters
governing U-NII devices in light of second generation MSS systems. For example, second
generation MSS systems may be more sensitive and therefore more susceptible to interference
from U-NII devices. On the other hand, if European HIPERLAN systems proliferate and operate
at more power than U-NII devices, second generation MSS systems may of necessity be
designed to be more robust and immune to interference from such devices.
97. We believe that this approach will provide both MSS feeder link and U-NII
operations with an appropriate level of protection and assurance for the continuation of their
operations. While we think it unlikely that an interference situation will arise, this approach will
permit us to develop regulatory solutions that will adequately protect the investments of both
services, if such a situation were to develop. Accordingly, we believe that this approach will
provide both the MSS community and the U-NII device manufacturers with adequate certainty
concerning their operations, and we do not believe that a "Part 16" or "safe harbor" rule is
necessary for U-NII devices at this time.
H. Equipment Authorization
98. In the NPRM, we proposed that U-NII devices would be subject to our
certification requirements pursuant to Section 15.201(b), prior to marketing. Motorola
recommends that we take this opportunity to streamline our equipment approval process so that
all products, including U-NII devices, may be approved and provided to the public with minimal
costs and delays. However, Motorola made no specific suggestion in reference to this
proceeding and its comments in reference to PP Docket No. 96-17 will be considered therein.
We do not believe that applying the certification process to U-NII devices will significantly delay
the provision of this equipment to the public. We believe this process helps prevent non-
compliant devices from interfering with other devices or services. Accordingly, we are adopting
our proposal to require U-NII devices to comply with the existing certification requirements for
intentional radiators under Part 15.
99. Finally, we will require U-NII devices to comply with the RF hazard requirements
set forth in Sections 1.1307(b), 1.1310, 2.1091, and 2.1093 of our rules. For purposes of these
rules, all U-NII equipment will be deemed to operate in an "uncontrolled" environment. Any
application for equipment certification for these devices must contain a statement confirming
compliance with these requirements. Technical information showing the basis for this statement
must be submitted to the Commission upon request.
ORDERING CLAUSE AND EFFECTIVE DATE
100. Accordingly, IT IS ORDERED that Part 15 of the Commission's Rules, 47 C.F.R.
Part 15 IS AMENDED as set forth in the attached Appendix, effective 60 days after publication
in the Federal Register. This action is taken pursuant to Sections 4(i), 303(c), 303(f), 303(g) and
303 (r) of the Communications Act of 1934, as amended, 47 U.S.C. Sections 154(i), 303(c),
303(f), 303(g) and 303(r).
101. Regulatory Flexibility Analysis. As required by Section 603 of the
Regulatory Flexibility Act ("RFA"), an Initial Regulatory Flexibility Analysis ("IRFA") was
incorporated in the NPRM in this proceeding. The Commission sought written public comments
on the proposals in the NPRM including on the IRFA. The Commission's Final Regulatory
Flexibility Analysis ("FRFA") in this Report and Order is attached as Appendix B.
