Notice: This opinion is subject to formal revision before publication in the
Federal Reporter or U.S.App.D.C. Reports. Users are requested to notify
the Clerk of any formal errors in order that corrections may be made
before the bound volumes go to press.
United States Court of Appeals
FOR THE DISTRICT OF COLUMBIA CIRCUIT
Argued March 15, 2004 Decided May 4, 2004
No. 03-1043
AT&T WIRELESS SERVICES, INC., ET AL.,
PETITIONERS
v.
FEDERAL COMMUNICATIONS COMMISSION AND
UNITED STATES OF AMERICA,
RESPONDENTS
AIRCELL, INC.,
INTERVENOR
On Petition for Review of an Order of the
Federal Communications Commission
L. Andrew Tollin argued the cause for petitioners. With
him on the briefs were Michael Deuel Sullivan, Douglas I.
Brandon, Carol L. Tacker, and John T. Scott III.
Bills of costs must be filed within 14 days after entry of judgment.
The court looks with disfavor upon motions to file bills of costs out
of time.
2
Roberta L. Cook, Counsel, Federal Communications Com-
mission, argued the cause for respondents. With her on the
brief were Robert H. Pate III, Assistant Attorney General,
Robert B. Nicholson and Robert J. Wiggers, Attorneys, John
A. Rogovin, General Counsel, Federal Communications Com-
mission, and Daniel M. Armstrong, Associate General Coun-
sel.
Michele C. Farquhar, Jonathan S. Franklin, and Lorane
F. Hebert were on the brief for intervenor.
Before: SENTELLE, ROGERS and GARLAND, Circuit Judges.
Opinion for the Court filed by Circuit Judge ROGERS.
ROGERS, Circuit Judge: In AT&T Wireless Services v. FCC,
270 F.3d 959 (D.C. Cir. 2001) (‘‘AirCell I’’), the court denied
all challenges except one to an order of the Federal Commu-
nications Commission granting AirCell, Inc., a waiver to
operate an aircraft-based analog cellular telephone system.
The court remanded to the Commission for an explanation of
its conclusion that AirCell’s system was unlikely to harmfully
interfere with the rights of ground-based cellular telephone
providers, specifically so that the Commission could explain
its choice of a particular threshold signal strength above
which AirCell would begin to cause such interference. In
petitioning for review of the Commission’s more detailed
explanation in AirCell, Inc., 18 F.C.C.R. 1926 (2003) (‘‘Re-
mand Order’’), AT&T Wireless Services, Inc., Cingular Wire-
less LLC, and CellCo Partnership contend that the Commis-
sion’s explanation on remand deviates from the reasoning in
the initial order in AirCell, Inc., 15 F.C.C.R. 9622 (2000)
(‘‘Initial Order’’), and, in any event, is unreasonable. We hold
that the petitioners misinterpret the Commission’s initial
order and the breadth of the remand in AirCell I, that they
have waived several of their challenges to the Remand Order
by failing to seek rehearing by the Commission of its explana-
tion on remand, and that their unwaived challenges are
unpersuasive. Accordingly, we deny the petition.
3
I.
The background to the Commission’s decision to grant to
AirCell, Inc. a waiver of the rules barring the use of analog
cellular telephones in aircraft, see 47 C.F.R. § 22.925 (2004),
is set forth in AirCell I, 270 F.3d at 961–63. A central
dispute in the initial AirCell proceeding before the Commis-
sion was how to predict whether, and if so how significantly,
AirCell would disrupt the networks of existing cellular tele-
phone providers, such as the petitioners. Because the peti-
tioners’ licenses entitle them to protection from harmful
interference within their geographic service areas, see 270
F.3d at 963–64; 47 C.F.R.§ 22.911(d) (2004), any waiver
allowing AirCell to operate depends upon a finding that such
operation will not cause harmful interference to license-
holders. The record contains data from several test flights,
in which receivers on the ground measured the strength of
AirCell’s signal while an aircraft, placing a call from an
AirCell telephone, flew overhead. The providers and AirCell,
based on this data, advocated different conclusions about the
likelihood of harmful interference from AirCell’s operation.
The tests were conducted in rural areas, where such harm-
ful interference is most likely because there is less back-
ground electromagnetic noise that a cellular call needs to
overcome. This permits cellular calls in rural areas to be of
acceptable quality even with a relatively weak signal, but
those weak signals are then more susceptible to interference
than calls in suburban or urban areas, which are required to
transmit at a higher strength. For the one day of test flights
the Commission considered representative of how AirCell
would operate in practice, July 10, 1997, AirTouch Communi-
cations submitted an analysis by Dr. William C. Y. Lee, who
concluded that AirCell’s calls would cause harmful interfer-
ence 30% of the time they were placed while flying over a
ground-based call using the same channel. He based this
prediction on the premise that harmful interference to a
ground call on the same frequency would occur whenever
AirCell’s signal strength exceeded a threshold of 124 decibels
below one milliwatt, i.e. –124 dBm. The Commission rejected
Dr. Lee’s –124 dBm threshold as ‘‘too conservative,’’ stated
4
that an interference threshold of 117 decibels below one
milliwatt, i.e. –117 dBm, was ‘‘more realistic for typical analog
systems,’’ and that ‘‘based on [ ] review of the evidence, it
appears [ ] that use of the latter threshold would have led to a
finding that AirCell would cause a significant level of harmful
interference 0% of the time.’’ Initial Order, 15 F.C.C.R. at
9631 n.67. On that basis, the Commission affirmed the grant
of the waiver to Aircell by the Bureau of Wireless Telecom-
munications.
In AirCell I, the court remanded on the harmful interfer-
ence threshold determination, holding that the Commission
had failed ‘‘to justify adequately its choice of an interference
threshold,’’ 270 F.3d at 968, and that it had not shown how it
‘‘translate[d] the raw signal data from the July 10, 1997 field
test into a finding that AirCell’s system ‘would cause a
significant level of harmful interference 0% of the time’ in the
real world’’ without conducting a probability study to deter-
mine how frequently the event studied, the simultaneous use
of the same channel by an AirCell call and a terrestrial call at
the same location, would occur. Id. at 968–69. The court
noted that the Commission’s ‘‘explanation for its conclusion in
AirCell’s favor may be relatively simple and briefly stated,’’
but that because the record contained conflicting explanations
of the data, the court was unable to discern ‘‘why the Com-
mission considered one interference threshold preferable to
another.’’ Id. at 968.
On remand, the Commission reaffirmed its prediction that
AirCell would not cause harmful interference with ground-
based cellular systems and explained its rejection of Dr. Lee’s
conclusion to the contrary. See Remand Order, 18 F.C.C.R.
1926. It explained that it had derived the –117 dBm figure by
starting from the strength of the weakest ground-based call
that could still be considered acceptable quality in a rural
area, which the Commission claimed was –100 dBm, and
subtracting a 17 dB ‘‘buffer,’’ which is the industry standard
for the amount by which a cellular call must exceed other
signals in order to be effective. It reasoned that the –124
dBm figure used by Dr. Lee had been derived by assuming
a –129 dBm environmental noise floor in quiet locations and
5
stipulating that AirCell would cause interference with ground
systems once it exceeded the noise floor by 4.76 decibels.
The Commission stated that this method was erroneous be-
cause it assumed too low a noise floor and failed to justify its
premise that interference to ground-based calls would become
harmful at 4.76 decibels above the noise floor. The Commis-
sion also stated that even if Dr. Lee’s conclusions about the
relative signal strengths were correct, it would show at most
that AirCell’s systems could cause interference, but not nec-
essarily harmful interference, because Dr. Lee had wrongly
assumed that all interference above his –124 dBm threshold
would be harmful. The Commission reasoned that the 17
dBm buffer was ‘‘significantly more buffer than is necessary
to prevent harmful interference,’’ and that an AirCell signal
would not cause harmful interference until it ‘‘seriously de-
grades, obstructs, or repeatedly interrupts’’ a ground-based
call, which, the Commission stated, would not occur until a
‘‘very substantial (e.g., 7 dB or more) excess over an [interfer-
ence threshold].’’ Id. at 1935–36.
Reviewing the test data from the July 10, 1997 test flights,
the Commission on remand showed that AirCell’s mean signal
strength exceeded the –117 dBm threshold on only on one of
the 24 test flights, and then only trivially. See id. at 1936–37.
The test data also ‘‘do not show AirCell received power
sustained at anywhere near the –110 dBm level that would
constitute harmful interference.’’ Id. at 1937. On the basis
of the fact that AirCell’s mean signal strength almost never
reached –117 dBm on the test flights, and that its sustained
signal strength came nowhere close to –110 dBm, the Com-
mission explained that it was unnecessary to conduct a proba-
bility study. The July 10 test flights measured AirCell’s
signal under the conditions when it was most likely to cause
interference (i.e., when an AirCell phone places a call while
flying over a ground-based phone using the same channel).
Therefore, because AirCell calls would not cause harmful
interference even under the worst-case scenario, it was un-
necessary to assess the probability of the worst-case scenario
occurring. The petitioners did not seek rehearing of the
Commission’s Remand Order, see 47 U.S.C. § 405(a) (2004),
6
but filed this petition for review pursuant to 47 U.S.C.
§ 402(a), and a motion for summary reversal and vacatur of
the Remand Order and the underlying waiver grant. The
court denied summary reversal. See AT&T Wireless, Inc. v.
FCC, No. 03–1043 (D.C. Cir. July 15, 2003) (per curiam).
II.
The sole question now before the court is whether the
Commission’s order satisfies the remand in AirCell I. The
court is generally the authoritative interpreter of its own
remand, see FCC v. Pottsville Broad. Co., 309 U.S. 134, 141
(1940), and in light of its previous opinion, the court owes no
deference to the Commission’s interpretation of its task on
remand. See City of Cleveland v. FPC, 561 F.2d 344, 346–47
(D.C. Cir. 1977). To the extent the Commission’s explanation
on remand encompasses technical predictions within its ex-
pertise, however, the court will defer to its judgment so long
as it is ‘‘not contrary to law, is rational, has support in the
record, and is based on a consideration of the relevant
factors,’’ NAACP v. FCC, 682 F.2d 993, 997 (D.C. Cir. 1982),
because ‘‘greater discretion is given administrative bodies
when their decisions are based upon judgmental or predictive
conclusions.’’ Id. at 1001; see also FCC v. WNCN Listeners
Guild, 450 U.S. 582, 594–96 (1981). Because the petitioners
did not seek rehearing of the Commission’s remand decision
before filing their petition for review, the court reviews only
whether the Commission has complied with the remand in-
struction by explaining its choice of an interference threshold
and its conclusion, based on the administrative record for the
initial order, that AirCell is unlikely to cause harmful inter-
ference with ground-based cellular calls. As this question
was presented to the Commission by the remand itself, see
AirCell I, 270 F.3d at 968–69, the Commission has had
opportunity to pass on it and the petitioners had no obligation
to seek rehearing before petitioning for review. See Time
Warner Entm’t Co. v. FCC, 144 F.3d 75, 79–81 (D.C. Cir.
1998). By bypassing review by the Commission, however, the
petitioners have waived any substantive challenges to the
Remand Order that were not presented by the court’s re-
7
mand. See 47 U.S.C. § 405(a); Time Warner Entm’t Co., 144
F.3d at 81.
The Commission, on remand, provided what the court
requested: an explanation of ‘‘why the Commission consid-
ered one interference threshold preferable to another.’’ Air-
Cell I, 270 F.3d at 968. It explained that the –117 dBm
threshold it considered ‘‘more realistic for typical analog
systems’’ in the initial AirCell order, 15 F.C.C.R. at 9631 n.67;
see also AirCell I, 270 F.3d at 968, was based on two
premises. First, an acceptable call in a rural area must be at
least 100 decibels below one milliwatt strong (–100 dBm) to
be of acceptable quality. See Remand Order, 18 F.C.C.R. at
1931–32. Second, as the parties did not dispute in the
original proceeding, a call, ‘‘to remain effective TTT must
sufficiently exceed the electromagnetic noise floor of the
location where the call is received, plus whatever interference
may be generated by a concurrent AirCell signal’’ by a ratio
of 17dB. See AirCell I, 270 F.3d at 966. Therefore, AirCell’s
signal risks interference with ground-based cellular calls
when it comes within seventeen decibels of –100 dBm, i.e.,
when it reaches a strength of –117 dBm. See Remand Order,
18 F.C.C.R at 1931–32.
The Commission similarly explained its reason for rejecting
the –124 dBm threshold proposed by AirTouch Communica-
tions in the report of Dr. Lee. Dr. Lee’s analysis was based
on an environmental noise floor in rural areas of –129 dBm,
–127 dBm when noise from AirTouch’s system is factored in,
and assumed that AirCell would cause harmful interference if
its signal exceeded that level by more than a ‘‘reliability
factor’’ of 4.76 dB. Dr. Lee’s calculations showed that an
AirCell signal would exceed the noise floor by the required
amount if it reached a strength of –124 dBm. The Commis-
sion did not consider Dr. Lee’s premise of a –127 dBm noise
floor credible; he assumed ‘‘zero environmental noise,’’ which
the Commission found was ‘‘not reflective of the typical noise
environment in which cellular carriers operate,’’ and his as-
sumption of only two decibels of system noise relied on high-
fidelity receiver equipment that was neither ‘‘universal or
standard practice for the cellular industry, even in rural
8
areas.’’ Remand Order, 18 F.C.C.R. at 1933–34. Second, the
Commission explained that the premise behind Dr. Lee’s
analysis — that AirCell would cause interference whenever it
exceeded the level of background noise by a 4.76 dB ‘‘reliabili-
ty factor’’ — had not been shown to have ‘‘any link’’ to
cellular call quality. See id. at 1935. Under the Commis-
sion’s reasoning, what matters is the ratio between AirCell’s
signal strength and the signal strength of terrestrial calls.
See id. AirCell’s signal might be undetectable if it falls below
the noise floor, but it does not need to; it can exceed the
noise floor without causing interference so long as it remains
at least seventeen decibels less powerful than any acceptable
ground-based call on the same channel.
Nothing in the explanation in the Remand Order deviates
impermissibly from the Commission’s reasoning in the Initial
Order. The petitioners contend that the –117 dBm threshold
was initially predicated, as was Dr. Lee’s –124 dBm threshold,
on reasoning ‘‘up’’ from the electromagnetic noise floor, and
that the Commission’s explanation on remand, which reasons
‘‘down’’ from the strength of an acceptable ground-based call,
is an attempt to evade record evidence that the noise floor at
the test site was –127 dBm, such that reasoning ‘‘up’’ would
have resulted in an interference threshold less permissive
than –117 dBm. Petitioners misread the initial order, howev-
er, for it does not suggest that the Commission derived the
–117 dBm figure by reasoning ‘‘up’’ from the noise floor.
Indeed, the court remanded because the Commission did not
explain the origin of the –117 dBm figure, which the initial
order simply stated was ‘‘more realistic for typical analog
systems.’’ 15 F.C.C.R. at 9631 n.67. While counsel for the
Commission at oral argument in AirCell I did represent to
the court that the –117 dBm figure was chosen because it was
the ‘‘average noise level in a quiet rural area,’’ the Commis-
sion is not bound to the post-hoc representations of counsel.
The court remanded on the interference threshold determina-
tion because it could not locate an explanation within the four
corners of the Commission’s order; on remand the Commis-
sion has supplied one. Put otherwise, the Commission’s
initial order did not mention the noise floor except in the
9
context of urban areas, as to which the Commission explained
that AirCell would not cause any interference because ‘‘the
higher noise floor in urban areas will ensure that the AirCell
units’ signal remains substantially below a detectable level.’’
15 F.C.C.R at 9637. There, the Commission explained, ‘‘very
marginal quality calls in quiet rural areas may be affected’’
notwithstanding the fact that the noise floor would mask
AirCell’s signal elsewhere. Id. at n.106. This discussion
contemplates that AirCell’s signal might exceed the noise
floor in rural areas, and is therefore consistent with the
Commission’s position on remand. But as it leaves unex-
plained the basis of the conclusion that AirCell will not cause
rural interference, it does not bind the Commission to the
petitioners’ insistence that the Commission previously con-
cluded that the noise floor would also mask AirCell’s signal in
rural areas.
The petitioners also challenge the merits of the Commis-
sion’s explanation on remand for the –117 dBm threshold.
First, they point out that the record showed an actual noise
floor of –127 dBm at the test site, confirming Dr. Lee’s figure
that the Commission considered too conservative. If there
was any such error, it was harmless. As explained, the actual
location of the electromagnetic noise floor is mostly irrelevant
to the interference threshold under the Commission’s reason-
ing; rather, what matters is the ratio between terrestrial
cellular calls and AirCell’s signal.
Second, the petitioners object to the Commission’s use of
–100 dBm as the starting point acceptable call from which to
reason ‘‘down’’ to the interference threshold. The Commis-
sion explained its choice of the –100 dBm figure on the basis
that some petitioners in the initial AirCell proceeding used
the –100 dBm figure, and that technical literature uses –100
dBm as the signal strength at which a call becomes so weak
that cellular sites in rural areas will ‘‘attempt to continue
carrying a call only if there were no cell site with better
reception to which to hand it off.’’ Remand Order, 18
F.C.C.R. at 1932. The petitioners, however, dispute the
relevance of the technical literature cited by the Commission,
and contend that record data show that many ground-based
10
cellular calls at the test site were weaker than –100 dBm, and
that absent any interference from AirCell, these calls would
function properly so long as they remain at least 17 decibels
above the measured noise floor, i.e, –110 dBm. Thus, the
Commission’s –100 dBm starting point effectively denies in-
terference protection to these weaker calls.
Whatever the merits of these objections might be, they are
not properly before the court. The only issue before the
court is to determine whether the Commission has complied
with the remand, not to review generally the merits of the
remand order where petitioners failed to petition for rehear-
ing, for the court may not consider any argument that ‘‘relies
on questions of fact or law upon which the Commission, or
designated authority within the Commission, has been afford-
ed no opportunity to pass.’’ 47 U.S.C. § 405(a). The peti-
tioners’ objections to the –100 dBm figure do not address
whether the Commission on remand explained its decision
based on the record. Indeed, their counsel conceded at oral
argument that the petitioners, by failing to seek rehearing,
have waived any objection that the Commission’s Remand
Order is arbitrary and capricious. While the court still must
ensure that the Commission’s decision on remand is explained
and has a basis in the record — both of which the Commis-
sion’s choice of a –100 dBm figure satisfies — the petitioners
have waived the right to challenge the substance of the
Commission’s determination beyond compliance with the re-
mand.
The petitioners also contend that the Commission’s expla-
nation on remand, despite lip service to the –117 dBm figure,
actually relies on a –110 dBm threshold. This miscompre-
hends both the purpose of the remand and the Commission’s
response. The court remanded so that the Commission could
explain ‘‘why [it] considered one interference threshold pref-
erable to another,’’ AirCell I, 270 F.3d at 968, not so that it
could defend the –117 dBm figure specifically. In rejecting
the challenges to the waiver on the grounds that the petition-
ers’ licenses had been modified by the waiver and that the
Commission had denied them the interference protection to
which they are entitled under 47 C.F.R.§ 22.911(d), the
11
court’s remand was not dependent on the Commission’s
reaching the conclusion that AirCell would not expose peti-
tioners to harmful interference by any specific technical
means: the court remanded because of the absence of a
satisfactory explanation for the Commission’s finding of non-
interference. See id. The terms of the court’s remand,
which focused only on the legal error of failing to provide an
adequate explanation, thus would not have prevented the
Commission from adopting a different interference threshold
on remand, provided the choice was based on the record and
adequately justified. Cf. Pottsville Broadcasting, 309 U.S. at
145. In any event, the Commission’s determination that
AirCell’s calls would not cause harmful interference was
based, as stated, on a threshold of –117 dBm. While the
Commission also discussed a more permissive –110 dBm
threshold, it did so to illustrate what it considered to be an
additional error in AirTouch Communications’ proposed –124
dBm threshold. See Remand Order, 18 F.C.C.R. at 1935–36.
The Commission’s review on remand of the July 10 test
flight data showed that AirCell’s mean signal strength ex-
ceeded –117 dBm only on one of the 24 test flights, and then
only by .62 decibels, a trivial amount. It concluded, there-
fore, that AirCell was unlikely to cause harmful interference.
See id. at 1937. This reasoning aligns with, and further
elaborates upon, the Commission’s conclusion in the initial
order. It differs only in the acknowledgment that one of the
test flights did show a signal strength slightly exceeding –117
dBm, as contrasted with its conclusion in the initial order that
AirCell would cause harmful interference ‘‘0% of the time.’’
Initial Order, 15 F.C.C.R. at 9631 n.67. Later in the order,
the Commission also noted that even if AirCell had exceeded
the –117 dBm threshold, the interference it would cause
would not necessarily become harmful until it reached a
higher level somewhere close to –110 dBm. See Remand
Order, 18 F.C.C.R. at 1935–36. But the Commission ex-
plained that AirCell’s signal would not exceed the –117 dBm
level in the first place. See id. at 1937. The context of the
Commission’s discussion makes clear that it is discussing a
hypothetical. Dr. Lee’s –124 dBm threshold assumed that
12
AirCell would cause harmful interference as soon as its
signal, by rising above the electromagnetic noise floor, was
strong enough to be detectable. The Commission was ex-
plaining that Dr. Lee’s analysis provided more buffer than
was necessary because not all detectable interference is so
severe that it can be considered harmful. See id. at 1935–36.
The petitioners point out that a –117 dBm interference
threshold can only lead to a finding of noninterference if one
considers AirCell’s mean signal strength, rather than its peak
signal strength, and that a finding of noninterference that
looks to AirCell’s peak signal strength during the test flights
can only be reached using a more permissive –110 dBm
threshold. The Commission relied upon mean signal
strength, under which the test flight data showed an AirCell
signal weaker than –117 dBm, and hence under the interfer-
ence threshold. The petitioners contend that this is errone-
ous, and that the Commission instead should have relied on
peak signal strength, which would have shown AirCell’s signal
to frequently exceed the –117 dBm threshold. This objection
is not properly before the court. Whether the average or
peak signal strength is what matters to an interference
analysis is the type of question that 47 U.S.C § 405(a)
requires parties to give the Commission opportunity to pass
upon before seeking judicial review, and the petitioners failed
to do so. Again, their counsel conceded at oral argument that
any objection that the use of mean signal strength was
arbitrary and capricious has been waived
The petitioners have not, however, waived their related
contention that the Commission’s use of mean signal strength
disregards the terms of the court’s remand, but that objection
is easily disposed of. They contend that irrespective of
whether it was arbitrary and capricious, use of mean rather
than peak signal strength is an attempted end-run around the
court’s instruction that the Commission demonstrate ‘‘how it
was able, in the absence of a probability study, to translate
the raw signal data from the July 10, 1997, field test’’ into a
determination of non-interference. See AirCell I, 270 F.3d at
968. The petitioners overread the court’s use of the phrase
‘‘raw signal data.’’ The context makes clear that the court
13
was using ‘‘raw signal data’’ to mean ‘‘actual data’’ from the
test flights as opposed to engineering theory and technical
literature, not taking sides in any dispute, which was nonexis-
tent on the first petition for review, as to the proper means of
interpreting the signal strength measured during the test
flights. To the extent the Commission on remand has made a
showing of which data it relied upon and how it interpreted it,
it has satisfied the court’s instruction to use ‘‘raw’’ data.
Finally, there was nothing ‘‘counterintuitive,’’ Petitioners’
Br. at 32, much less an abuse of discretion, in the Commis-
sion’s decision to rely on the record in the original proceeding
and to reject for filing the petitioners’ comments on the
court’s remand instruction. The Commission rejected a filing
of comments by AT&T Wireless, Cingular Wireless and Cell-
Co addressing the remand instruction, explaining that it
‘‘need not consider’’ the comments because it ‘‘neither solic-
ited nor granted leave to file additional pleadings’’ in the case,
and that the existing record was ‘‘a sufficiently adequate base
on which to rest the Commission’s decision’’ because the
Commission was ‘‘merely explain[ing]’’ its earlier decision. 18
F.C.C.R. at 1927. This is consistent with the court’s remand,
which sought an explanation from the Commission based of
the administrative record of the initial order. See AirCell I,
270 F.3d at 968. The petitioners acknowledge that all of the
data cited by their comments was already in the record, but
fail to appreciate that the time to submit comments was after,
not before, the Commission issued its remand order.
Accordingly, because the Remand Order provides an ade-
quate explanation in response to the remand in Aircell I, 270
F.3d at 968, we deny the petition for review of the Remand
Order and the motion for vacatur of the orders granting
waivers to AirCell.