Ex Parte Paul et al

Board of Patent Appeals and Interferences

Sep 15, 2010

10338762 (B.P.A.I. Sep. 15, 2010)

UNITED STATES PATENT AND TRADEMARK OFFICE
UNITED STATES DEPARTMENT OF COMMERCE
United States Patent and Trademark Office
Address: COMMISSIONER FOR PATENTS
P.O. Box 1450
Alexandria, Virginia 22313-1450
www.uspto.gov
APPLICATION NO. FILING DATE FIRST NAMED INVENTOR ATTORNEY DOCKET NO. CONFIRMATION NO.
10/338,762 01/08/2003 Manoj Paul P15462 2169
94221 7590 09/15/2010
Buckley, Maschoff & Talwalkar LLC/
Intel Corporation
50 Locust Avenue
New Canaan, CT 06840
EXAMINER
CHOU, ALAN S
ART UNIT PAPER NUMBER
2451
MAIL DATE DELIVERY MODE
09/15/2010 PAPER
Please find below and/or attached an Office communication concerning this application or proceeding.
The time period for reply, if any, is set in the attached communication.
PTOL-90A (Rev. 04/07)

UNITED STATES PATENT AND TRADEMARK OFFICE
_____________

BEFORE THE BOARD OF PATENT APPEALS
AND INTERFERENCES
_____________

Ex parte MANOJ PAUL and SRINIVASA R. BEEREDDY
_____________

Appeal 2009-005763
Application 10/338,762
Technology Center 2400
______________

Before JOHN C. MARTIN, JOSEPH F. RUGGIERO, and
KARL D. EASTHOM, Administrative Patent Judges.

MARTIN, Administrative Patent Judge.

DECISION ON APPEAL1

1 The two-month time period for filing an appeal or commencing a civil
action, as recited in 37 C.F.R. § 1.304, or for filing a request for rehearing,
as recited in 37 C.F.R. § 41.52, begins to run from the “MAIL DATE”
(paper delivery mode) or the “NOTIFICATION DATE” (electronic delivery
mode) shown on the PTOL-90A cover letter attached to this decision.

Appeal 2009-005763
Application 10/338,762

2
STATEMENT OF THE CASE
This is an appeal under 35 U.S.C. § 134(a) from the Examiner’s
rejection of claims 1, 4-11, 13-19, and 21-32, which are all of the pending
claims.
We have jurisdiction under 35 U.S.C. § 6(b). We reverse.

A. Appellants’ invention
Appellants’ invention relates to load distribution in H.323 networks.
Specification, title. In general, H.323 is a multimedia conferencing protocol
associated with voice, video, and/or data conferencing via a packet switched
network, such as an Internet Protocol (IP) network. Specification 1:5-7.2
Figure 1 is reproduced below.

2 References herein to Appellants’ Specification are to the Application as
filed rather than to corresponding Patent Application Publication
2004/0133677 A1.
Appeal 2009-005763
Application 10/338,762

3

Figure 1 is an illustration of a known H.323 network (id. at 4:7). Each
endpoint (EP) 130, which may comprise, for example, a telephone, video
phone, or Personal Computer (PC), registers with a gatekeeper 120 in its
administrative domain 110 (id. at 2:12-14). Specifically, after discovering
an available gatekeeper 120, EP 130 transmits a Registration Request (RRQ)
message to that gatekeeper 120, which responds with either (1) a
Registration Confirm (RCF) message if sufficient resources are available to
support that EP 130 or (2) a Registration Reject (RRJ) message if sufficient
resources are not available (id. at 3:6-14). After being registered, an EP 130
can request admission to make or receive a call by transmitting an
Admission Request (ARQ) message to its gatekeeper 120, which responds
Appeal 2009-005763
Application 10/338,762

4
with an Admission Confirm (ACF) message if it has enough resources to
support the call or an Admission Reject (ARJ) message if it does not have
enough resources (id. at 3:15-20).
Different gatekeepers 120 within an administrative domain 110 can
have different registration loads and/or call loads. For example, one
gatekeeper 120 might have registered a large number of EPs 130 while
another gatekeeper 120 registered only a few EPs 130 (id. at 2:21-24). As a
result, the distribution of traffic through the network 110 may not be
balanced, and the exchange of information between EPs 130 may be
degraded (id. at 2:24-26).
Figure 2 is reproduced below.

Appeal 2009-005763
Application 10/338,762

5
Figure 2 is a block diagram of an embodiment of a load distribution
element 200 that includes a storage device 220 to store load information
associated with a plurality of gatekeepers in the H.323 network (id. at 6:7-
12). For example, load information for gatekeepers within an administrative
domain may be centrally stored at a border element and used to balance
loads between gatekeepers (id. at 6:18-20)
A load associated with one gatekeeper to be adjusted based on load
information associated with at least one other gatekeeper (id. at 7:9-10). For
example, the centralized load distribution unit 210 may determine that a
gatekeeper's current registration load is high as compared to other
gatekeepers and therefore instruct that gatekeeper to reject further RRQ
message from endpoints (id. at 7:10-13). As another example, the
centralized load distribution unit 210 may determine that a gatekeeper's
current call load is low as compared to other gatekeepers and therefore
instruct that gatekeeper to accept further ARQ message from endpoints (id.
at 7:13-16).
Figure 4 is reproduced below.
Appeal 2009-005763
Application 10/338,762

6

Figure 4 illustrates a database 400 that can be stored in the storage
device 220 of load distribution element 200 (id. at 7:21-23). The stored
information includes, for each gatekeeper, its current registration load and
call load as well as an indication whether it can accept an additional RRQ
message or an additional call (i.e., ARQ) message. Specifically,
[f]or each gatekeeper, a current registration load 404 is
stored in the database 400. The current registration load 404
might represent, for example, the number of endpoints that are
currently registered with the gatekeeper divided by the maximum
number of endpoint registrations that can be supported by that
particular gatekeeper. Moreover, the database 400 indicates
whether or not further RRQ messages should be accepted 406 by
that gatekeeper. For example, as illustrated in FIG. 4, GK_101 has
a relatively high registration load (i.e., 85%) and therefore should
Appeal 2009-005763
Application 10/338,762

7
not accept further RRQ messages. On the other hand, GK_102
has a relatively low registration load (i.e., 55%) and therefore
should accept further RRQ messages.
Similarly, a current call load 408 is stored for each
gatekeeper. The current call load 408 might represent, for
example, the number of calls that are currently routed through the
gatekeeper divided by the maximum number of calls that can be
routed by that particular gatekeeper. Moreover, the database 400
indicates whether or not further ARQ messages should be
accepted 410 by that gatekeeper. For example, GK_103 has a
relatively high call load (i.e., 70%) and therefore should not
accept further ARQ messages. On the other hand, GK_102 has a
relatively low registration load (i.e., 60%) and therefore should
accept further ARQ messages.
Id. at 8:3-19.

B. The claims
The independent claims before us are claims 1, 9, 17, 25, 28, and 31,
of which claim 1 reads as follows:
1. A border element apparatus associated with an
administrative domain of an H.323 network, comprising:
a storage device to store load information associated with
a first message type and a second message type from a plurality
of gatekeepers of the H.323 network; and
a load distribution unit to adjust a first load associated
with the first message type for one gatekeeper based on load
information associated with at least one other gatekeeper and to
adjust a second load associated with the second message type
for one gatekeeper based on load information associated with at
least one other gatekeeper.
Claims App. (Br. 11.)
Appeal 2009-005763
Application 10/338,762

8
C. The references
The Examiner’s rejections are based on the following references:
Evslin et al. (“Evslin”) US 6,404,864 B1 Jun. 11, 2002
Li et al. (“Li”) US 6,591,301 B1 Jul. 8, 2003
Kliland et al. (“Kliland”) US 6,738,383 B1 May 18, 2004
Borella et al. (“Borella”) US 7,080,151 B1 Jul. 18, 2006

D. The rejections
Claims 1, 4, 5, 7, 9-11, 13, 15, 17-19, 21, 23, and 25-32 stand rejected
under 35 U.S.C. § 103(a) for obviousness over Kliland in view of Evslin and
Borella. Final Action 2, para. 2.
Claims 6, 8, 14, 16, 22, and 24 stand rejected under 35 U.S.C.
§ 103(a) for obviousness over Kliland in view of Evslin, Borella, and Li. Id.
at 8, para. 11.

THE ISSUES
The principal issues raised by Appellants’ arguments are whether the
Examiner erred in finding (1) that Evslin discloses the use of monitoring to
maintain a call quality standard and (2) that Borella discloses adjusting a
load associated with Borella’s call registration messages.3

3 See Ex parte Frye, 94 USPQ2d 1072, 1075 (BPAI 2010) (precedential)
(“If an appellant fails to present arguments on a particular issue — or, more
broadly, on a particular rejection — the Board will not, as a general matter,
unilaterally review those uncontested aspects of the rejection.”). Designated
as precedential at
(Continued on next page.)
Appeal 2009-005763
Application 10/338,762

9
ANALYSIS
Kliland discloses an H.323 system employing “real” (i.e.,
conventional”) gatekeepers, which know how to handle lots of types of
H.323 messages (col. 1, ll. 24-27), and “lightweight” gatekeepers, which can
handle only a limited range of the H.323 message set (col. 1, ll. 61-67).
In Kliland’s system, “supplementary achievements, comprising for
example load balancing, QoS (Quality of Service), information about cost,
etc. can easily be implemented” (col. 1, ll. 49-53 (emphasis added)).
Figure 1 of Kliland is reproduced below.

http://www.uspto.gov/ip/boards/bpai/decisions/prec/index.jsp.
Appeal 2009-005763
Application 10/338,762

10
Figure 1 is a sketch of a typical scenario employing endpoints (EPs),
an internal, lightweight gatekeeper located inside the domain or LAN, and
external, real gatekeepers located outside (col. 2, ll. 24-29).4 The
lightweight gatekeeper has knowledge of the real gatekeepers’ loads and, on
this basis, can distribute the traffic towards the least loaded gatekeeper
(col. 3, ll. 8-12). Load information can be communicated between real and
lightweight gatekeepers with RAI and RAC messages (col. 3, ll. 21-24.).
Alternatively, the proprietary non-StandardData field of GCF and GRJ can
be used for exchanging, e.g. QoS, information between the lightweight
gatekeeper and the real gatekeepers (col. 3, ll. 41-44). “If the lightweight
gatekeeper keeps track of each real gatekeeper’s e.g. QoS, e.g. in a table
form, these tables may be updated each time the lightweight gatekeeper
receives a GCF or GRJ from a real gatekeeper” (col. 3, ll. 44-48).
The Examiner finds that Kliland fails to satisfy two limitations in
claim 1: (1) the recited storage device and load distribution unit; and (2) the
recited first and second message types. The Examiner relies on Evslin for
the storage device and load distribution unit (Final Action 2, para. 3). Evslin
discloses a system and method for monitoring, evaluating and actively

4 Kliland’s claim 1 recites that the “subset” of H.3232 messages handled by
the internal gatekeeper includes: “GRQ (Gatekeeper Request), GCF
(Gatekeeper Confirm), GRJ (Gatekeeper Reject), IRR (Information Request
Response), IRQ (Information Request), RAI (Resource Availability
Indication), RAC (Resource Availability Confirm), RRQ (Registration
Request), RCF (Registration Confirm), and RRJ (Registration Reject)” (col.
6, ll. 10-18).
Appeal 2009-005763
Application 10/338,762

11
managing telephone-call quality in data-network-based telephony networks
(col. 1, ll. 8-12).5 Figure 2 of Evslin is reproduced below.

Figure 2 depicts a portion of Evslin’s data-network-based telephony
network (DNT) 200 (col. 4, ll. 50-52). Numeral 201 designated a distributed
monitoring, evaluation, and routing (DiMER) system (col. 4, ll. 53-54).
Evslin’s Figure 3B is reproduced below.

5 Evslin does not refer to H.323.
Appeal 2009-005763
Application 10/338,762

12

Figure 3B is a schematic diagram of basic functional elements for
implementing the operations of DiMER 201 (col. 7, ll. 34-40). The portfolio
monitoring and reporting element 310 receives call metrics CM (col. 7, l.
66–col. 8, l. 2) and provides consolidated call metrics to network quality
analysis and feedback element 320 for analysis (col. 11, ll. 6-8) in order to
obtain data representing call quality, such a call breakdown on a per-gateway
basis (col. 10, ll. 47-48). The Examiner reads the recited “storage device to
store load information” on portfolio monitoring and reporting element 310
and reads the recited “load distribution unit” on column 10, lines 22-30
(Final Action 2, para. 3). These lines explain that DiMER 201 “generates
revised routes by shifting call traffic between acceptable gateways” in
response to the consolidated processed metrics generated by the portfolio
generator.
Appeal 2009-005763
Application 10/338,762

13
The Examiner concluded that it would have been obvious “to
incorporate [Evslin’s] telephony network monitoring and management
system” into Kliland’s H.323 system in order “to meet [the] call quality
standard within the telephony network system.” Final Action 3. Appellants
argue that “[n]either Kliland, Evslin, nor the present application reference[s]
a call quality standard. Therefore, it would not have been obvious to
combine references since meeting an unknown call quality standard is not
suggested by either reference.” (Br. 6.) The Examiner responds (Answer
11) by directing Appellants’ attention to Evslin’s column 2, lines 48-60, of
which lines 54-57 state that “[i]n accordance with the present teachings, such
a system, and data-network-based telephony networks incorporating the
same, advantageously route calls to meet call-quality standards and/or cost
goals, among other targets” (emphasis added). Appellants responded by
arguing that
this statement indicates that call-quality standards are not always
met because it may, instead, meet cost goals, among other targets.
Appellants contend that the statement of general advantages fails
[sic] short of rising to a motivation to combine references since
neither Kliland, Evslin, nor the present application references a
specific call-quality standard, nor disclose[s] a problem of meeting a
specific call-quality standard.
(Reply Br. 2.) This new argument is entitled to no consideration because it
is not responsive to a new point in the Answer. See Ex parte Borden,
Appeal 2009-005763
Application 10/338,762

14
93 USPQ2d 1473, 1473-74 (BPAI 2010) (“informative”6) (absent a showing
of good cause, the Board is not required to address an argument newly
presented in the Reply Brief that could have been presented in the principal
Brief on Appeal). Although the Final Action does not identify the column
and line numbers where Evslin discloses “call quality standards,” that
disclosure is easy to locate and therefore should have been addressed in the
opening Brief.
For the above reasons, Appellants have not persuaded us that the
Examiner erred by combining the teachings of Kliland and Evslin in the
proposed manner.
The Examiner further finds that Evslin does not remedy Kliland’s
failure to disclose first and second message types, as required by claim 1,
such as messages associated with a current registration load and a current
call load, as recited in claim 4, and relies on Borella for such teachings.
Final Action 3. Borella discloses a method and system for mobile IP home
agent redundancy by using home agent control nodes for managing multiple
home agents. Borella, title. Borella’s invention is described as applicable to
H.323 networks (see, e.g., col. 7, ll. 5-6).
Figure 3 of Borella is reproduced below.

6 Designated as “Informative Opinion” at
http://www.uspto.gov/ip/boards/bpai/decisions/inform/index.jsp.
Appeal 2009-005763
Application 10/338,762

15

Figure 3 is a block diagram illustrating an exemplary Mobile IP
system 64, which includes one or more “immobile” network devices 66, 68,
70, 72, 74, and 76 and a mobile network device 78 (col. 8, ll. 49-52). The
home subnet 80 is connected to an external network 82 such as the Internet
or an intranet via a home agent (HA) 76, which can be a “gateway router”
for the home subnet 80 (col. 8, ll. 64-67).
Appeal 2009-005763
Application 10/338,762

16
When the HA 76 accepts a valid RRQ from a mobile node 78 that it
serves as an HA 76, the HA 76 creates or modifies an entry in an MBR
(Mobile Boot Record7) table for this mobile node 78 in its mobility binding
tables including: the mobile node’s care-of address; the Identification field
from the Registration Reply; and the remaining Lifetime of the registration
(col. 15, ll. 40-46).
Figure 6 is reproduced below.

Figure 6 is a block diagram illustrating an exemplary HA 138 with
plural HAs and plural HA control nodes (HACNs) used with HA 76 (col. 13,
ll. 60-62). HA 138 includes plural HACN cards 140, 142 and plural HA
cards 144-166 in an HA card chassis (col. 13, ll. 62-65). All active HAs
144-164 and standby HA 166 send periodic heartbeat messages to an active
HACN 140 (col. 16, ll. 46-47). These messages include the current call

7 Borella, col. 8, ll. 3-4.
Appeal 2009-005763
Application 10/338,762

17
load (e.g., data sessions including VoIP, H.323, etc.) and status of the
sending HAs (col. 16, ll. 47-49). When an HACN 140 receives a new RRQ,
it forwards the message to an HA (e.g., 144) chosen such that load is
balanced across all HAs 144-166 (col. 16, ll. 62-65).
Appellants do not deny that Borella discloses adjusting a load
associated with the current call load messages. Instead, Appellants argue
that in Borella
the registration request allows a home agent to create, modify, or
delete a mobility binding. However, nowhere does Borella
disclose that a registration request comprises load related
information.
Therefore, the proposed combination of prior art, taken in
any permissible combination fails to disclose or suggest both
adjusting a first load associated with a first message type and
adjusting a second load associated with a second message type.
(Br. 6 (emphasis added).) When the phrase “load related information” in the
first paragraph of this argument is read in light of the second paragraph, it is
clear Appellants are arguing that Borella fails to disclose or suggest
adjusting a load (e.g., a registration load) based on the registration
messages. This argument is not adequately addressed by the Examiner’s
response, which reads as follows:
Borella teaches a Registration Request to set up [a] mobility
binding record which is related to current load as claimed (see
column 15 line 25-32). Borella also discloses the mobility
binding record including information of the registration such as
the mobile node’s care-of address, and remaining lifetime of the
registration, which are all related registration request load
information (see column 15 line 40-47).
Appeal 2009-005763
Application 10/338,762

18
(Answer 11, para. B.) While we agree with the Examiner that the
the active registration requests in Borella can accurately be characterized as
“load related information,” that is not enough to demonstrate the
obviousness of adjusting a load associated with the active registration
requests.
For the above reasons, we will not sustain the rejection of claim 1 for
obviousness over Kliland in view of Evslin and Borella or the rejection on
this ground of the other independent claims (viz., claims 9, 17, 25, 28, and
31), which recite limitations like the above-discussed limitations of claim 1.
For the same reasons, we will not sustain the rejection on this ground of
dependent claims 4, 5, 7, 10, 11, 13, 15, 18, 19, 21, 23, 26, 27, 29, 30,
and 32.
We also will not sustain the rejection of claims 6, 8, 14, 16, 22, and 24
for obviousness over Kliland in view of Evslin, Borella, and Li, because the
subject matter relied on in Li does not cure the above-noted deficiency in the
other references.

DECISION
The Examiner’s decision that claims 1, 4-11, 13-19, and 21-32 are
unpatentable under 35 U.S.C. § 103(a) is reversed.

REVERSED

Appeal 2009-005763
Application 10/338,762

19

babc

Buckley, Maschoff & Talwalkar LLC/
Intel Corporation
50 Locust Avenue
New Canaan, CT 06840