Ex Parte Ovadia

Patent Trial and Appeal Board

Aug 6, 2013

10242839 (P.T.A.B. Aug. 6, 2013)

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.
10/242,839 09/13/2002 Shlomo Ovadia P14961 3432
67861 7590 08/06/2013
COOL PATENT, P.C.
c/o CPA Global
P.O. BOX 52050
MINNEAPOLIS, MN 55402
EXAMINER
AHMED, JAMIL
ART UNIT PAPER NUMBER
2886
MAIL DATE DELIVERY MODE
08/06/2013 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 PATENT TRIAL AND APPEAL BOARD
________________

Ex parte SHLOMO OVADIA1
________________

Appeal 2010-004191
Application 10/242,839
Technology Center 2600
________________

Before MARK NAGUMO, MICHAEL P. COLAIANNI, and
DONNA M. PRAISS, Administrative Patent Judges.

NAGUMO, Administrative Patent Judge.

DECISION ON APPEAL
Shlomo Ovadia (“Intel”) timely appeals under 35 U.S.C. § 134(a)
from the final rejection2 of claims 1-12 and 14-38.3 We have jurisdiction.
35 U.S.C. § 6. We reverse.

1 The real party in interest is listed as Intel Corporation (“Intel”). (Appeal
Brief, filed 21 August 2009 (“Br.”), 1.)
2 Office action mailed 15 June 2009 (“Final Rejection”; cited as “FR”).
3 According to Intel, claim 39 has been canceled. (Br. 2.)
Appeal 2010-004191
Application 10/242,839

2
OPINION
A. Introduction4
The subject matter on appeal relates to an apparatus (some claims are
drawn to “a system”) and associated methods for routing data bursts in a
photonic burst-switched (“PBS”) network. (Spec. 5 [0022].) (Consistent
with the disclosure in the 839 Specification and the Brief, the terms
“photonic” and “optical” will be used interchangeably when referring to
bursts in this Opinion.) PBS networks are improvements on conventional
optical switched networks that are described as using wavelength routing
techniques that require slow (ten milliseconds) and expensive optical-
electrical-optical (“OEO”) conversion at each optical switching node of the
network. (Id. at 1 [0003].) Existing OEO technologies are also said to be
incapable of supporting the “bursty” traffic of packet communications, such
as the Internet. (Id.) In the claimed invention, all switching of the optical
data bursts is done optically.
Some background about networks is helpful to understand the claimed
invention. (The following discussion is oversimplified and should not be
mistaken for a detailed description of the invention.) Figure 1 of
the 839 Specification, reproduced on the following page, illustrates an
optical burst switching network 10 as a network of switching nodes. Small
local area networks 13 communicate with the metropolitan area network 11

4 Application 10/242,839, Method and apparatus of the architecture and
operation of control processing unit in wavelength-division-multiplexed
photonic burst-switched networks, filed 13 September 2002. The
specification is referred to as the “839 Specification,” and is cited as “Spec.”
Appeal 2010-004191
Application 10/242,839

3
(“MAN”) via ingress switching nodes 15, and data is routed among various
core switching nodes 17. The MAN is connected to a still larger Wide Area
Network (“WAN”) via egress switching node 18.
{Figure 1 is shown below}

{Fig. 1 shows a photonic burst-switched network}
In the words of the 839 Specification, “the ingress, egress and core
switching nodes of optical MAN 11 are configured to send and/or receive
optical control bursts, optical data bursts, and other control signals that are
wavelength multiplexed . . . .” (Spec. 8 [0031].) The wavelength
multiplexing is said to permit the signals to be sent on one or more pre-
selected wavelengths. (Id.) “In addition,” the Specification adds, “optical
Appeal 2010-004191
Application 10/242,839

4
data bursts can be time division multiplexed (TDM) on a given wavelength.”
(Id.) As will be seen, this last property or function is the critical argued
feature of the claimed invention.
In general, as illustrated below, a photonic burst includes a control
burst containing the header and other router information (Figure 4B),
followed by a payload (optical data burst) containing the data segments of
the packet (Figure 4B). (Id.)
{Figure 4A (left) and Figure 4B (right) are shown below}

{Fig. 4A: optical data burst format Fig. 4B: control burst format}
Variable duration time division multiplexing is depicted in Figure 6, below.

{Fig. 6 shows variable duration time division multiplexing of a
control burst at wavelength λC1, and of a data burst at λC2}
In the claimed apparatus, the control burst is TDM-propagated in E portions
on control wavelength λC1 (Fig. 6, upper). (Spec. 17 [0058].) The data burst
is TDM-transmitted in F portions on data wavelength λC2 (Fig. 6, lower).
Appeal 2010-004191
Application 10/242,839

5
(Id.) According to the Specification, “[e]ach of the F portions of the optical
data burst signal can have different time durations.” (Id.) Hence, the data
packets are said to be transmitted in a variable duration time division
multiplex channel.
An implementation of a core switching node 17 is illustrated in
Figure 8, which is reproduced below.

{Fig. 8 shows an implementation of a core switching node 17}
At the input of the node, a TDM-transmitted optical control burst,
followed by the data burst, is processed by demultiplexers 30 into the control
burst and the data burst. The control burst is routed to control unit 81, where
network resources are assessed by network processor 83 and compared to
the requirements of the associated data burst. The network processor then
reserves the appropriate photonic burst switches 32 and the corresponding
transmission channel(s). After the photonic burst switches have been set, the
Appeal 2010-004191
Application 10/242,839

6
data burst portions, delayed as necessary in optical buffer 37 (a set of fiber
delay lines), are routed through PBS 32. The network processor assembles a
new control burst, which is recombined with the data burst, and TDM-
transmitted via multiplexers 34, with appropriate variable time durations of
the F portions of the data burst. (Spec. 19 [0063] to 20 [0066].)
The appealed subject matter encompasses an apparatus capable of
performing certain functions of the network processor 83 in the core
switching node illustrated in Figure 8, supra. Figure 14, shown below,

{Fig. 14 shows a control burst processing unit of the invention}
illustrates a control burst processor 1400 having the features recited in
claim 1. The processor comprises a control burst parser 1406, that analyzes
the incoming control burst, a scheduler 1430 that assesses the routing
Appeal 2010-004191
Application 10/242,839

7
information for the associated data burst, as well as the network resources,
and, with the assistance of the PBS configuration and control block 1422 and
the contention resolution block 1424, sets the appropriate PBS switches 32
(see Fig. 8, supra), which switch the associated data burst to the appropriate
transmission channel(s). (An apparatus according to appealed claim 1 is not
required to have—nor is it prevented from having—the following elements
and functions; but they are added to indicate the role of the variable time
duration time division multiplexing in the invention.) A new control burst
for the outgoing burst is formed by control burst generator 1416, which is
appropriately combined with the optically burst-switched data burst and
time-duration TDM-transmitted to the next mode.
Claim 1 is representative and reads:
An apparatus [1400], comprising:
a parser [1406] to deframe an incoming control burst;
a scheduler [1430] coupled to said parser [1406]
to schedule
a configuration of a photonic burst switch (“PBS”)
and
a variable duration time division multiplexing channel
to transmit a data burst corresponding to said
incoming control burst,
the variable duration time division multiplexing
channel comprising a plurality of time slots,
each time slot comprising a duration that can be
different from at least one other time slot;
a contention resolution block [1424] coupled to said
scheduler [1430]
to determine an alternate routing of said data burst
when said scheduler [1430] determines a resource
contention of said PBS; and
Appeal 2010-004191
Application 10/242,839

8
a PBS configuration and control module [1422]
coupled to said scheduler [1430] to configure the
PBS in response to said scheduler,
the PBS configuration and control module [1422]
further coupled to the contention resolution
block [1424]
to configure the PBS in response to the contention
resolution block [1424] to route the data burst
according to the alternate routing when the
resource contention occurs.
(Claims App., Br. 28; indentation, some paragraphing, emphasis, and
bracketed labels to elements in Fig. 14, reproduced supra, added.)
The Examiner maintains the following grounds of rejection:5
A. Claims 1, 4-6, 14, 30, and 33-34 stand rejected under
35 U.S.C. § 103(a) in view of the combined teachings of
Zheng6 and Ciecierski.7
A1-A11. Claims 2-5, 7-11, 15-29, and 31-38 stand rejected in view
of Zheng and Ciecierski, further combined with
numerous other references. (Br. 7-8; Ans. 7-15.)8

5 Examiner’s Answer mailed 20 October 2009 (“Ans.”).
6 Si Q. Zheng et al., Unified associative memory of data channel schedulers
in an optical router, U.S. Patent Application Publication 2002/0118419 A1
(29 August 2002), based on an application filed 29 November 2001.
7 Walter V. Ciecierski et al., Synchronous programmable mixed format time
division multiplexer, U.S. Patent 3,632,882 (1972).
8 As none of these ‘supplemental’ rejections is contested separately, we
refer the interested reader to the Examiner’s Answer and the Brief for details
of these rejections.
Appeal 2010-004191
Application 10/242,839

9
B. Discussion
Findings of fact throughout this Opinion are supported by a
preponderance of the evidence of record.
Initially, we find that Intel has only specifically raised arguments
against the prima facie obviousness of limitations found in claim 1, and that
the separate patentability of no other claim has been argued. Intel has not
raised arguments based on secondary considerations such as unexpected
results. The issue dispositive of this appeal is whether the Examiner erred
harmfully in combining the teachings of Zheng and Ciecierski in arriving at
the legal conclusion of prima facie obviousness.
Briefly, the Examiner finds that Zheng, in Fig. 1b, describes an
apparatus meeting all the limitations of claim 1, but “Zheng does not
specifically disclose that the data channel is a variable duration time division
multiplexing channel.” (FR 3; Ans. 5.) The Examiner finds that it is well
known to use such a channel in a communication system for data
transmission, as illustrated by Ciecierski. On this basis, the Examiner
concludes that it would have been obvious to incorporate the concept of
using such a channel in the system of Zheng. In the Examiner’s words,
“[t]he motivation would have been to accommodate a plurality of different
data sources having both mixed rates and mixed data lengths (Ciecierski:
column 1, lines 26-33).” (FR 3; Ans. 5.) We understand the Examiner to be
arguing that it would have been obvious to use the known method of
variable duration time division multiplexing in place of the transmission
method particularly described by Zheng.
Appeal 2010-004191
Application 10/242,839

10
Intel does not expressly dispute the Examiner’s findings that every
component and connection between components recited in claim 1 reads on
particular components and connections in the core optical router described in
Zheng Fig. 1. (Br. 9; Intel refers to these findings as the Examiner’s “major
premise”.) Indeed, Intel emphasizes that Zheng “contemplates data bursts
and Burst Header Packets (BHPs) that have a variable length, and that fixed-
length data bursts and BHPs are special cases.” (Id. at ¶ 2.d.) Intel also
emphasizes that Zheng uses a slotted transmission technique between the
various routers and nodes in the disclosed optical burst switched (OBS)
network, and that a slot is a fixed-length time period. (Id. at ¶ 2.c, citing
Zheng [0067].) On this basis, Intel agrees with the Examiner that Zheng
“does not specifically suggest that the data channel is a variable duration
time division multiplexing channel.” (Id. at 10, ¶ 2.g.)
Intel argues that the Examiner’s reliance on Ciecierski “so simplifies
the actual context and disclosure of Ciecierski that the minor premise
essentially ignores the true context of the Ciecierski system.” (Br. 11, ¶ 5.b.)
As a result, according to Intel, the Examiner’s syllogism fails because the
“minor premise” (i.e., that Ciecierski teaches that variable duration time
division multiplexing of data is well known) is “essentially baseless.”
(Id. at ¶ 14 5.n.) Intel concludes that the Examiner’s conclusion of
obviousness must be reversed.
Among the findings of fact that underpin the conclusion of
obviousness are the teachings of the prior art, the knowledge and level of
ordinary skill, and whether the teachings of prior art would have been
combined by the ordinary worker. In re Gartside, 203 F.3d 1305, 1319
Appeal 2010-004191
Application 10/242,839

11
(Fed. Cir. 2000). In general, we evaluate findings of fact according to the
preponderance of the evidence of record.
The initial problem with the Examiner’s conclusion of obviousness
arises from the Examiner’s assumptions that a person having ordinary skill
in the art would have considered the substitution of any known data-
transmitting channel for the data-transmitting channel described by Zheng
both feasible and likely to work. It would then have been obvious to
“schedule” such a channel. We use the term “assumptions” deliberately as
the Examiner does not come forward with citations to evidence in the record
that this would have been so.
Zheng is concerned with routing data over networks, such as the
internet, at switching rates ranging from gigabits (1012) per second to terabits
(1014) per second. (Zheng 1 [0008]-[0009].) More particularly, Zheng
describes an optical burst-switched router in which incoming optical
information is routed from incoming optical transmission media to outgoing
optical media transmitting data over a plurality of channels. (Id. at [0013].)
Zheng provides a delay buffer (a set of N fiber delay lines (id. at 2 [0048]))
for delaying selected information between the incoming transmission
medium and one of the outgoing optical transmission media. (Id. at [0013].)
Scheduling circuitry stores information on unscheduled time and time gaps
on each channel of the outgoing medium. (Id.) In Zheng’s words, “[t]he
present invention provides an efficient architecture for identifying
unscheduled time and time gaps within which a data burst can be
scheduled.” (Id. at [0014].)
Appeal 2010-004191
Application 10/242,839

12
Ciecierski, in contrast, relates to “time division pulse multiplexing
systems in which intermittent pulse signals from a plurality of
unsynchronized transmitters are multiplexed over a common transmission
facility.” (Ciecierski col. 1, ll. 6-9.) More particularly, “[p]ulse signals of
relatively low pulse repetition frequency from various sources are
interleaved with one another to form a composite data stream that is
transmitted at high speed for some distance over a common link to a time
shared computer for example.” (Id. at ll. 9-14.) To implement this time
division pulse multiplexing system, Ciecierski provides “programmable
timing generators at the transmitter and receiver [of the common link] . . .
that can readily and efficiently accommodate a plurality of different data
sources having both mixed rates and mixed character lengths.”
(Id. at ll. 26-33.)
Much of the difficulty with the Examiner’s argument lies in the
differences between the optical, opto-electric and electronic technologies
used by Zheng (published in 2002), and the electronic technologies used by
Ciecierski (published in 1972). Moreover, as Intel points out (Br. 14, 5.l.),
Ciecierski teaches a channel in the context of data from plural sources to a
single destination, whereas Zheng is concerned with switching data traffic
over networks, in which data from plural sources is routed through plural
channels to plural destinations. The Examiner has not shown that the
practical problems solved by Ciecierski involving low-speed data
transmission would have been perceived as relevant in any practical way by
the ordinary worker to the problems confronted by Zheng of switching data
over networks at transmission rates of 109 to 1012 bits per second. In this
Appeal 2010-004191
Application 10/242,839

13
regard, Ciecierski has little if any more evidentiary power than would a
taking of Official Notice by the Examiner that variable duration time
division multiplexing channels were known to be used for transmitting data.
That fact, without evidence of motivation to apply variable duration TDM to
data transmission in the optical burst-switched network environment of
Zheng, is insufficient to sustain a challenged conclusion of prima facie
obviousness.
The Examiner, in the Response to Argument section of the Answer,
appears belatedly to attempt to bolster the rejection with argument that the
“‘data burst’ of Zheng actually reads [on] the terminology ‘time slot’ in
claim 1, since ‘time slot’ in the instant application in fact refers to duration
of TDM channels.” (Ans. 17.) The Examiner also finds that Zheng Fig. 17
shows that “the data burst of Zheng is time division multiplexing (TDM)
channels having variable lengths.” (Id. at 18.) Here, we understand the
Examiner to be arguing that Zheng “inherently” discloses variable duration
TDM transmission of data bursts.
Taken literally, however, these arguments are not persuasive because
applying these “definitions” to the language of the claim results in a
limitation that does not make sense. The function of the variable duration
time division multiplexing channel (“variable channel”) is to transmit a data
burst that corresponds to a particular control burst. The variable-channel
comprises a plurality of time slots, which time slots can have durations that
are not all the same. But, if a data burst of Zheng is a time slot in appealed
claim 1, it does not make sense to call a data burst of Zheng a variable-
channel because a variable channel, by definition, comprises plural time
Appeal 2010-004191
Application 10/242,839

14
slots. While we recognize that the Examiner may be interpreting the term
“variable duration time division multiplexing channel” in some broad
manner, it is improper to interpret a term in a claim that has not been
rejected as indefinite in a way that results in inconsistencies in the claim.
In this regard, Intel, by agreeing that Zheng does not expressly
disclose a variable channel, tacitly, but not affirmatively, implies that CH1 in
Fig. 17a, transmitting the variable length data bursts DB1, DB4, and DB2, is
not necessarily a variable channel. This argument appears to be consistent
with the description of variable duration time division multiplexing
providing in Figure 6, reproduced supra. The Examiner has not come
forward with evidence that the data bursts shown in Zheng Fig. 17a are
TDM-transmitted in F portions having variable durations. Nor has the
Examiner come forward with credible evidence that merely transmitting
variable duration data bursts would have been considered variable duration
TDM transmission. We conclude the weight of the evidence is against the
Examiner’s secondary argument.
The Examiner has not relied on the additional references for evidence
that cures the defects of Zheng and Ciecierski.
We conclude that Intel has shown harmful error in the obviousness
rejections maintained by the Examiner on appeal.
C. Order
We reverse the rejection of claims 1-12 and 14-38.
REVERSED
Appeal 2010-004191
Application 10/242,839

15
tc