Ex Parte Castelli et al

Board of Patent Appeals and Interferences

Mar 8, 2012

09916935 (B.P.A.I. Mar. 8, 2012)

UNITED STATES PATENT AND TRADEMARK OFFICE
____________

BEFORE THE BOARD OF PATENT APPEALS
AND INTERFERENCES
____________

Ex parte VITTORIO CASTELLI, PETER A. FRANASZEK,
and JOY ALOYSIUS THOMAS
____________

Appeal 2010-001595
Application 09/916,935
Technology Center 2100
____________

Before JOSEPH F. RUGGIERO, JOHN A. JEFFERY,
and THOMAS S. HAHN, Administrative Patent Judges.

JEFFERY, Administrative Patent Judge.

DECISION ON APPEAL
Appeal 2010-001595
Application 09/916,935

2
Appellants appeal under 35 U.S.C. § 134(a) from the Examiner’s
rejection of claims 1-40 and 42. We have jurisdiction under 35 U.S.C. §
6(b). We reverse.

STATEMENT OF THE CASE
Appellants’ invention predicts computer-system performance based
on the types and number of active devices in the system. See generally
Spec. 4-6. Claim 1 is illustrative:
1. A method for predicting a subsequent resource utilization in a
computer system having a plurality of devices, comprising the steps of:
monitoring, over a period of time, a contemporaneous resource
utilization and a number of active devices to obtain monitored values of the
contemporaneous resource utilization and the number of active devices; and
predicting the subsequent resource utilization, based upon the
monitored values of the contemporaneous resource utilization and the
number of active devices.
The Examiner relies on the following as evidence of unpatentability:
Sweet US 6,836,800 B1 Dec. 28, 2004
(filed Sept. 30, 1999)

THE REJECTIONS
1. The Examiner rejected claim 42 under 35 U.S.C. § 112, first
paragraph, as failing to comply with the written description requirement.
Ans. 4.1
2. The Examiner rejected claims 1-402 under 35 U.S.C. § 102(e) as
anticipated by Sweet. Ans. 5-12.

1 Throughout this opinion, we refer to (1) the Appeal Brief filed October 24,
2006; (2) the Examiner’s Answer mailed April 30, 2007; and (3) the Reply
Brief filed July 2, 2007.
Appeal 2010-001595
Application 09/916,935

3

THE WRITTEN DESCRIPTION REJECTION
The Examiner finds that claim 42 does not comply with the written
description requirement under § 112 since Appellants’ disclosure is silent on
(1) “segmenting . . . variable length historic data”; (2) comparing
“segmented historic data” and its “unequal length”; and (3) comparing “non-
homogeneous type data.” Ans. 4. The Examiner adds that while the
functionality of Appellants’ Figure 15 establishes a vector with values for
each device of a specific type, the claim requires monitoring over a specific
time period to obtain three distinct values over that period. Ans. 14-15.
Appellants argue that not only are the purported deficiencies in the
rejection not commensurate with the language of claim 42, that claim’s
subject matter is nonetheless supported by the disclosure at various cited
passages and figures. App. Br. 9-11; Reply Br. 2-3. The issue before us,
then, is as follows:

ISSUE
Under § 112, has the Examiner erred in rejecting claim 42 by finding
that it does not comply with the written description requirement? This issue
turns on whether Appellants’ disclosure reasonably conveys to skilled
artisans that Appellants possessed the claimed invention as of the filing date.

2 Although the Examiner’s rejection erroneously includes cancelled claim 41
(Ans. 5, 11-12), we present the correct claim listing here for clarity. Accord
App. Br. 9 (indicating that claims 1-40 are rejected under § 102); Ans. 3
(confirming this status as correct). Also, since the Examiner withdrew the
rejection of claim 7 under § 112 (id.), that rejection is not before us.
Appeal 2010-001595
Application 09/916,935

4
PRINCIPLES OF LAW
To satisfy the written description requirement, the disclosure must
reasonably convey to skilled artisans that Appellants had possession of the
claimed invention as of the filing date. Ariad Pharms., Inc. v. Eli Lilly &
Co., 598 F.3d 1336, 1351 (Fed. Cir. 2010) (en banc).
Although the description requirement under § 112 does not demand
(1) any particular form of disclosure, or (2) the Specification recite the
claimed invention verbatim, a description that merely renders the invention
obvious does not satisfy the requirement. Id. at 1352 (citations omitted).

ANALYSIS
Based on the record before us, we find error in the Examiner’s written
description rejection of claim 42. The claim recites, in pertinent part,
predicting subsequent resource utilization by monitoring over a time period:
(1) contemporaneous resource utilization; (2) a number of active devices;
and (3) each active device’s type to obtain corresponding values.
Subsequent resource utilization is then predicted based on these three
monitored values.
There is no dispute that the originally-filed disclosure supports
predicting subsequent resource utilization based on the first two monitored
values, namely (1) contemporaneous resource utilization, and (2) a number
of active devices—limitations commensurate with the invention recited in
claim 1. Compare claim 1 with claim 42. That the present application’s
Abstract and Brief Summary sections use language strikingly similar to
claim 1 (i.e., prediction based on the first two values of claim 42) only
Appeal 2010-001595
Application 09/916,935

5
further evidences that the originally-filed disclosure amply supports these
limitations.
The question, then, is whether the originally-filed disclosure supports
prediction based on these two values plus values of the type of each active
device. On this record, we find that it does.
We first note—as do Appellants (App. Br. 11)—that the Examiner’s
analysis in the rejection is somewhat puzzling, for it uses terminology
pertaining to comparing “segmented” and “non-homogeneous” data that is
not recited in claim 42.3 Although the Examiner attempts to clarify these
terms and their relationship to the recited limitations in the Response to
Arguments section (Ans. 15), we find this analysis unavailing.
We do note, however, that claim 42 was not originally filed, but rather
was added via amendment during prosecution. See Amd’t filed Nov. 7,
2005, at 10-12. Nevertheless, Appellants cite various passages and figures
in the disclosure that are said to support the recited limitations. App. Br. 10;
Reply Br. 2-3.
Although the originally-filed disclosure does not use the same
language in the claim in connection with prediction based on all three recited
values (unlike claim 1), we nonetheless find that the functionality of
Appellants’ Figure 15 considered with other cited functionality reasonably
conveys to skilled artisans that Appellants possessed the invention of claim
42 when the application was filed.

3 For example, the Examiner finds that the original disclosure lacks (1)
“segmenting . . . variable length historic data”; (2) comparing “segmented
historic data” and its “unequal length”; and (3) comparing “non-
homogeneous type data.” Ans. 4.
Appeal 2010-001595
Application 09/916,935

6
Appellants’ Figure 15 shows predicting resource utilization as a
function of the number of active devices in different device classes in a
computer system. To this end, Appellants’ system identifies, at particular
points in time, the number of active devices corresponding to a first type, or
“type 1” devices, and represents this number as “J1.” Spec. 29:3-9; Fig. 15.
Notably, there are two values associated with this representation: (1) “J”
representing the number of devices, and (2) the numeral “1” representing the
corresponding device type. See id. The system similarly identifies the
numbers of other types of devices (i.e., “type 2,” (J2), etc. for “d” types of
devices, and collects these numbers into a vector J = [J1, . . . , Jd]. Id. These
values are then used as a basis for model-based prediction. Id.
This functionality therefore enables prediction based not only on the
number of active devices, but also their type—an enhancement to
predictions based solely on the number of active devices and
contemporaneous resource utilization as in claim 1. That the very title of the
present invention indicates that prediction is based on both types and
numbers of active devices only bolsters this conclusion. Accord Spec. 1:5-8.
Thus, when considered as a whole, the originally-filed disclosure reasonably
conveys to ordinarily skilled artisans prediction based on the three recited
values.
To the extent that the Examiner argues that the functionality of Figure
15 does not support the monitored number of active devices being variable
over time (see Ans. 14), we disagree. Although the number and type of
active devices is determined at a particular point in time (Spec. 29),
Appeal 2010-001595
Application 09/916,935

7
Appellants’ disclosure is nonetheless replete with references to accounting
for the variability of such devices over time.4
We are therefore persuaded that the Examiner erred in rejecting claim
42.

THE ANTICIPATION REJECTION
The Examiner finds that Sweet discloses every recited feature of
independent claim 1 including predicting subsequent resource utilization
based on monitored values of (1) contemporaneous resource utilization, and
(2) a number of active devices which are said to be “integrated into the
system operation” of Sweet’s Figure 1. Ans. 5, 17-19.
Appellants argue that while Sweet’s signature detection software
derives data for “trending predictions,” these predictions pertain to network
performance regarding its capacity or configuration—not monitored values
of (1) contemporaneous resource utilization, and (2) the number of active
devices as claimed. App. Br. 17-18; Reply Br. 3-4. Appellants make similar
arguments regarding (1) identifying resource saturation in claim 21, and (2)
predicting the effects of adding a new device in claim 31—methods that are
likewise based on monitored values of resource utilization and the number of
active devices. App. Br. 18-20; Reply Br. 4. The issues before us, then, are
as follows:

4 See, e.g., Spec. 15:8-9 (noting that variations in the number of devices is
accounted for); see also Spec. 18:13-15 (same regarding variations of
distribution of numbers of active devices); Spec. 25:14-17 (noting that the
time-dependency of prediction parameters is accounted for).
Appeal 2010-001595
Application 09/916,935

8
ISSUES
Under § 102, has the Examiner erred by finding that Sweet:
(1) predicts subsequent resource utilization based on monitored values
of (a) contemporaneous resource utilization, and (b) a number of active
devices as recited in claim 1?
(2) identifies resource saturation (claim 21) or predicts the effects of
adding a new device (claim 31) based on monitored values of (a) resource
utilization, and (b) a number of active devices?

ANALYSIS
Based on the record before us, we find error in the Examiner’s
anticipation rejection of independent claim 1. Sweet’s system 10
automatically detects “signatures” (i.e., statistically-detectable patterns in
measurement data) to track network and computer resource usage for
analysis, thus allowing effective bandwidth management, capacity planning,
and development of realistic service-level expectations. Sweet, col. 1, l. 66
– col. 2, l. 17; col. 2, ll. 55-56; col. 3, ll. 24-30; Fig. 1. To this end, at least
one computer (e.g., 16b) on the network gathers (1) “utilization data” 32
(e.g., data regarding how much bandwidth of an individual network link is
being used), and (2) response time data. Sweet, col. 3, ll. 31-53; Figs. 2-3.
The computer’s signature detection software then analyzes this data to
derive signature data 40 and “other data” 42 that may be needed for (1)
subsequent calculations; (2) statistical correlation; or (3) “trending
predictions.” If a signature exists, it is used to (1) generate alarms for
unusual activity; (2) report on network status; and (3) plan changes (e.g.,
network upgrades). Sweet, col. 3, ll. 33-40; col. 4, ll. 42-52; Figs. 2-3.
Appeal 2010-001595
Application 09/916,935

9
Based on this functionality, we can see how Sweet’s system monitors
“contemporaneous resource utilization” as claimed, for the “utilization data”
that it gathers can indicate, among other things, bandwidth consumption
over a certain time period (see Sweet, col. 3, ll. 47-51)—certainly a
consumable resource consistent with Appellants’ description of the term.
See Spec. 2:4-6. And we can see how Sweet’s “trending predictions” are
based on this monitored resource utilization, for these predictions are based
on data derived from associated utilization data as noted above. See Sweet,
col. 4, ll. 42-46.
Despite Appellants’ and the Examiner’s respective characterizations
(App. Br. 17-18; Ans. 18-19), however, Sweet does not further describe
these “trending predictions.” But even assuming, without deciding, that
these predictions pertain to subsequent resource utilization as the Examiner
contends (Ans. 18-19), we still cannot say that these predictions are also
based on monitored values of the number of active devices as claimed.
While there would be a certain number of active devices in the network of
Sweet’s Figure 1, Sweet does not necessarily monitor the number of these
devices to obtain corresponding values, let alone use these values to arrive at
a prediction.
Nor does Sweet’s use of signature data change our conclusion. To be
sure, Sweet uses signatures to generate alarms for unusual activity, as well
as to plan network upgrades. Sweet, col. 3, ll. 34-40. In one example,
Sweet derives a particular signature from gathered data indicating light data
traffic from 2 to 3 AM. This signature is then used to establish an alarm
threshold to alert a network manager of unusually high data traffic in that
time period. Sweet, col. 2, l. 55 – col. 3, l. 3. Even assuming that using the
Appeal 2010-001595
Application 09/916,935

10
signature to establish a threshold could be considered predicting “subsequent
resource utilization”—at least in terms of what is normal and abnormal for
that time period—we still cannot say that this threshold-based “prediction”
is also necessarily based on monitored values of the number of active
devices as claimed. Although the established threshold is based on data
traffic, that does not mean that this traffic is necessarily tied to a particular
number of active devices: rather, an increased amount of data traffic could
result from an increased number of messages from a particular active device
during that period. We reach this conclusion noting that the term “traffic” in
this context is defined as “[t]he messages transmitted and received over a
communication channel.”5 In any event, even if it were probable that the
monitored data traffic was tied to a particular number of active devices on
which the threshold-based prediction is based, that is not enough for inherent
anticipation which requires that the reference necessarily disclose that
feature—which it does not.6
We reach a similar conclusion regarding planning network upgrades
based on a signature. Sweet, col. 3, ll. 34-40. Even assuming, without
deciding, that this planning involves a prediction based on values of
contemporaneous resource utilization, we cannot say that it is also based on
monitored values of the number of active devices as claimed.

5 NEIL SCLATER & JOHN MARKUS, MCGRAW-HILL ELECTRONICS
DICTIONARY 477 (6th ed. 1997).
6 “Inherency . . . may not be established by probabilities or possibilities. The
mere fact that a certain thing may result from a given set of circumstances is
not sufficient.” In re Robertson, 169 F.3d 743, 745 (Fed. Cir. 1999)
(citations and internal quotation marks omitted).
Appeal 2010-001595
Application 09/916,935

11
We are therefore constrained by the record before us to find that the
Examiner erred in rejecting (1) independent claim 1; (2) independent claims
21, 31, and 38-40 which contain commensurate disputed limitations; and (3)
the dependent claims for similar reasons.

CONCLUSION
The Examiner erred in rejecting (1) claim 42 under § 112, and (2)
claims 1-40 under § 102.

ORDER
The Examiner’s decision rejecting claims 1-40 and 42 is reversed.

REVERSED

babc