Ex Parte John et al

Board of Patent Appeals and Interferences

Aug 12, 2010

10425695 (B.P.A.I. Aug. 12, 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.
10/425,695 04/28/2003 Rajesh John RSTN-031 5202
30139 7590 08/12/2010
WILSON & HAM
1811 Santa Rita Road
Suite 130
Pleasanton, CA 94566
EXAMINER
MESFIN, YEMANE
ART UNIT PAPER NUMBER
2444
MAIL DATE DELIVERY MODE
08/12/2010 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 RAJESH JOHN and MIKE MORRISON
____________________

Appeal 2009-006232
Application 10/425,695
Technology Center 2400
____________________

Before JOSEPH L. DIXON, JAY P. LUCAS, and THU A. DANG,
Administrative Patent Judges.

DANG, 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 2009-006232
Application 10/425,695

2
I. STATEMENT OF THE CASE
Appellants appeal under 35 U.S.C. § 134 (2002) from a final rejection
of claims 1-20. We have jurisdiction under 35 U.S.C. § 6(b) (2008).
We AFFIRM.

A. INVENTION
According to Appellants, the invention generally relates to computer
networking and more specifically to a network processor for use within a
network node (Spec. 1, ¶ [0002]).

B. ILLUSTRATIVE CLAIM
Claim 1 is exemplary and are reproduced below:
1. A network processor, comprising:

a plurality of processing engines; and

packet assignment logic operable to ascertain packet size
of received packets and to selectively assign the received
packets to the processing engines, wherein the packet
assignment logic distributes the received packets according to at
least in part packet size of previously distributed packets.

C. REJECTIONS
The prior art relied upon by the Examiner in rejecting the claims on
appeal is:
Bunce 6,836,808 B2 Dec. 28, 2004

Appeal 2009-006232
Application 10/425,695

3
Eakins, “Per-packet Load Balancing UPS IPSec Processor
Performance,” CommsDesign (April 22, 2002),
http://www.commsdesign.com/showArticle.jhtml?articleID=1650468
1.

Claims 1-20 stand rejected under 35 U.S.C. § 103(a) over the
teachings of Bunce in view of Eakins.

II. ISSUE
Did the Examiner err in finding that the combination of Bunce in view
of Eakins would have taught or suggested “packet assignment logic operable
to ascertain packet size of received packets” wherein “the packet assignment
logic distributes the received packets according to at least in part packet size
of previously distributed packets” (claim 1)? In particular, the issue turns on
whether combining Bunce’s load-balancing with Eakin’s detection of packet
size would have suggested the features of distributing according to packet
size of previously distributed packets as required by claim 1.

III. FINDINGS OF FACT
The following Findings of Fact (FF) are shown by a preponderance of
the evidence.
Bunce
1. Bunce discloses performing packet processing tasks on multiple
processors or threads concurrently and in a pipelined fashion, wherein
availability is determined based on the capacity of the input work
queue associated with each processor or thread (col. 1, ll. 45-60).
Appeal 2009-006232
Application 10/425,695

4
2. Before a processor assigns a protocol processing task to a subsequent
processor, the current available capacity of its input work queue is
checked, wherein if the input work queue is overloaded, the task is
assigned to another processor having a lower input processing load
(col. 4, ll. 18-22).
Eakins
3. Eakins discloses load-balancing on a per-packet basis, wherein a core
works sequentially to finish working on each packet before moving to
the next (pg. 2).
4. The system assigns packets from the same session to different cores
by allowing one core to work on a large packet while smaller packets
are processed by different cores (id.).
5. Eakins discloses assigning packets to specific processing cores based
on packet size, thus processing small 64-byte packets or large 1536-
byte packets, by measuring the size of the packet and then assigning it
to the core that is ready for that project (id.).
6. Thus, Eakins’ loading-balancing on a per-packet basis eliminates
waste occurring when processing small packets using power that was
allocated to process a large packet (pg. 3).

IV. ANALYSIS
Claims 1-17
Though Appellants admit that “Eakins describes load-balancing on a
per-packet basis” (App. Br. 5), Appellants contend that “Eakins is merely
Appeal 2009-006232
Application 10/425,695

5
concerned with measuring the size of a current packet” (id.). According to
Appellants, “he[sic] availability of the cores is not necessarily dependent on
the size of the packet(s) processed by each core” (App. Br. 6). Therefore,
Appellants contend that “Eakins does not teach determining the availability
of a core based [on] the size of a packet that is being processed by a core,
but rather teaches determining the availability of a core” (App. Br. 7).
However, the Examiner finds that “Bunce disclosed assigning the
received packet to an optimal thread that is determined to be optimal (not
overloaded)” (Ans. 11), and “Eakins determines which core is ready to
process a packet of a determined size (i.e., the predetermined size threshold)
which implies prior determination of the workload of the core with respect
to it’s[sic] status with previously assigned queued packets in the parallel
processing” (Ans. 12). Thus, the Examiner concludes that Bunce and Eakins
“both assigned packets based on the previously assigned packets queued in
the work queue of each processing core” (Ans. 12-13).
Thus, the issue we address on appeal is whether the combination of
the combination of Bunce in view of Eakins would have taught or suggested
“packet assignment logic operable to ascertain packet size of received
packets” wherein “the packet assignment logic distributes the received
packets according to at least in part packet size of previously distributed
packets” (claim 1), as Appellants contend. In particular, we address whether
combining Bunce’s load-balancing with Eakin’s detection of packet size
would have suggested the features of distributing according to packet size of
previously distributed packets as required by claim 1.
Appeal 2009-006232
Application 10/425,695

6
By contending that “Eakins is merely concerned with measuring the
size of a current packet” (App. Br. 5), Appellants appear to be arguing that
individually, Eakins does not disclose the features of claim 1. However, the
Examiner rejects claim 1 over the combined teachings of Bunce and Eakins,
and what the combined teachings would have suggested to one of ordinary
skill in the art. One cannot show nonobviousness by attacking references
individually where the rejections are based on combinations of references.
See In re Merck & Co., Inc., 800 F.2d 1091, 1097 (Fed. Cir. 1986).
Bunce discloses determining processor or thread availability based on
the capacity of the associated input work queue (FF 1), wherein the current
available capacity of its input work queue is checked, and if the input work
queue is overloaded, the task is assigned to another processor having a lower
input processing load (FF 2). A skilled artisan would have understood
Bunce to disclose assigning a task to a processor or thread based on its load,
i.e., the load of the previous tasks.
Eakins discloses load-balancing on a per-packet basis (FF 3), wherein
packets from the same session are assigned to different cores with one core
processing a large packet such as a 1536-byte packet, while different cores
processing smaller packets such as 64-byte packets (FF 4-5). Thus, in
Eakins, if one packet that is determined to be large is assigned to a core, the
smaller packets will be assigned to different cores.
In Eakins, each packet’s processing must finish before moving to the
next (FF 3). That is, if the previously distributed large 1536-byte packet is
Appeal 2009-006232
Application 10/425,695

7

determined as being processed by the core, the core will not be assigned
subsequent packets.
Further, in Eakins, the packets are distributed by size to ensure that
there is no waste (FF 6). We find Eakins to teach detecting the packet size
of a currently received packet and then, using load-balancing, distributing
the current packet to a core based on the packet size of the previously
distributed packets. That is, contrary to Appellants’ contention that “he[sic]
availability of the cores is not necessarily dependent on the size of the
packet(s) processed by each core” (App. Br. 6), Eakins teaches that
availability is in fact dependent on the size of packets to ensure that there is
no waste.
Accordingly, we conclude that combining Bunce’s load-balancing
with Eakins’ detection of packet size would have suggested the features of
distributing according to packet size of previously distributed packets as
required by claim 1.
Appellants do not provide separate arguments for independent claims
6 and 15, but merely contend that “the remarks provided above in regard to
the rejection of claim 1 also apply to the rejections of claims 6 and 15”
(App. Br. 7).2 Accordingly, claims 6 and 15 fall with claim 1.

2 Though Appellants provide new arguments concerning claims 6 and
15 in the Reply Brief, it is inappropriate for Appellants to discuss for the
first time in the Reply Brief matters that could have been raised in the
Appeal Brief. That is, “[t]he failure to raise all issues and arguments
Appeal 2009-006232
Application 10/425,695

8

Accordingly, we find that the Examiner did not err in rejecting
independent claims 1, 6 and 15, and claims 2-5, 7-14, 16 and 17 depending
respectively therefrom and falling therewith under 35 U.S.C. § 103(a).
Claims 18-20
As to claim 18-20, though Appellants admit that “Eakins uses the
example of processing a large packet and smaller packets by different cores”
(App. Br. 8), Appellants contend that “Eakins does not describe a category
for the large packet and a separate category for smaller packets” since “there
is no indication in the language of Eakins how those categories might be
delineated” (App. Br. 9).
However, the Examiner finds that Eakins discloses “measuring the
size of the packet and assign[ing] the packet to a packet-processing core that
is ready for that project and that the packet sizes being categorized in sizes
ranging from small to large packet sizes” (Ans. 14.) We agree with the
Examiner.
diligently, in a timely fashion, has consequences,” and thus, such newly-
raised arguments are technically waived. Ex parte Borden, 93 USPQ2d
1473, 1474-1475 (BPAI 2010) (informative decision) (“[T]he reply brief [is
not] an opportunity to make arguments that could have been made in the
principal brief on appeal to rebut the Examiner's rejections, but were not.”).
Cf. Kaufman Company v. Lantech, Inc., 807 F.2d 970, 973 n.* (Fed. Cir.
1986) and McBride v. Merrell Dow and Pharms., Inc., 800 F.2d 1208, 1211
(D.C. Cir. 1986).

Appeal 2009-006232
Application 10/425,695

9
Eakins discloses assigning packets to specific processing cores based
on packet size, such as small 64-byte packets or large 1536-byte packets (FF
4-5). We agree with the Examiner that Eakins does categorize the packets as
in at least two size categories of small (64-byte) or large (1536-byte). In
Eakins, the packets are compared to a threshold to ensure that the packets
are assigned without waste (FF 6). Thus, we agree with the Examiner that
Bunce in view of Eakins would have suggested this feature of claim 18.
Accordingly, we find that the Examiner did not err in rejecting claims
18-20 under 35 U.S.C. § 103(a).

V. CONCLUSIONS OF LAW
(1) The Examiner did not err in concluding that claims 1-20 are
unpatentable under 35 U.S.C. § 103(a) over the teachings of Bunce in view
of Eakins.
(2) Claims 1-20 are not patentable.

VI. DECISION
We affirm the Examiner’s rejection of claims 1-20 under 35 U.S.C.
§ 103(a).
No time period for taking any subsequent action in connection with
this appeal may be extended under 37 C.F.R. § 1.136(a).

AFFIRMED

Appeal 2009-006232
Application 10/425,695

10
peb

WILSON & HAM
1811 SANTA RITA ROAD
SUITE 130
PLEASANTON, CA 94566