UNITED STATES PATENT AND TRADEMARK OFFICE
`
`___________________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`___________________
`
`TALARI NETWORKS, INC.
`Petitioner
`
`v.
`
`FATPIPE NETWORKS INDIA LIMITED,
`Patent Owner
`
`
`
`
`
`___________________
`
`Case No. IPR2016-00976
`Patent 6,775,235 B2
`___________________
`
`PETITIONER’S REPLY TO PATENT OWNER’S RESPONSE
`
`
`
`
`
`

`

`
`
`
`
`I.
`
`II.
`
`TABLE OF CONTENTS
`
`IPR2016-00976
`Patent No. 6,775,235 B2
`
`Page(s)
`
`INTRODUCTION .......................................................................................... 1
`
`FATPIPE’S CLAIM CONSTRUCTIONS ARE IMPROPER ....................... 1
`FatPipe’s Narrow Construction Of “Selects Between Network
`A.
`Interfaces On A Per-Packet Basis” / “Make Network Path
`Selections On A Packet-By-Packet Basis” Is Legally Flawed ............. 2
`1.
`The Specification Does Not Support FatPipe’s
`Constructions ............................................................................. 3
`The Board Should Adopt The Broadest Reasonable
`Interpretation .............................................................................. 6
`FatPipe’s Construction Improperly Narrows “Dynamic Load-
`Balancing” ............................................................................................ 7
`1.
`FatPipe’s Construction Is Unsupported ..................................... 7
`2.
`The Board Should Adopt The Broadest Reasonable
`Interpretation .............................................................................. 9
`
`B.
`
`2.
`
`III. CLAIM 4 IS ANTICIPATED OR RENDERED OBVIOUS ...................... 10
`A. Karol Discloses The Claimed “Packet Path Selector Which
`Selects Between Network Interfaces On A Per-Packet Basis” ........... 10
`Karol Discloses The Claimed “Site Interface” .................................. 12
`
`B.
`
`IV. CLAIMS 5 AND 7-15 ARE ANTICIPATED OR RENDERED
`OBVIOUS ..................................................................................................... 15
`A.
`Claim 5 Is Invalid Over Karol Alone Or In Combination With
`Stallings Which Disclose “Obtaining At Least Two Known
`Location Address Ranges Which Have Associated Networks” .......... 16
`Claim 7 Is Invalid Over Karol Which Discloses “The Method
`Of Claim 5, Wherein The Forwarding Step Forwards The
`Packet Toward The Internet When The Packet’s Destination
`Address Does Not Lie Within Any Known Location Address
`Range” ................................................................................................ 20
`
`B.
`
`
`
`-i-
`
`

`

`C.
`
`D.
`
`E.
`
`Claim 8 Is Invalid Over Karol Which Discloses “The Method
`Of Claim 5, Wherein The Destination Address Identifies A
`Destination Location To Which Only A Single Associated
`Network Provides Connectivity From The Current Location,
`And The Forwarding Step Forwards The Packet To That Single
`Associated Network” .......................................................................... 21
`Claim 9 Is Invalid Over Karol Which Discloses “The Method
`Of Claim 5, Wherein Repeated Instances Of The Selecting Step
`Make Network Path Selections On A Packet-By-Packet Basis” ........ 22
`Claims 11-13 Are Invalid Over Karol Alone Or In Combination
`With Stallings Which Disclose “Dynamic Load-Balancing” ............ 23
`
`CLAIM 19 IS ANTICIPATED OR RENDERED OBVIOUS BY THE
`INSTITUTED GROUNDS ........................................................................... 24
`A. Karol Discloses The Claimed “Site Interface” ................................... 24
`Karol Discloses The Claimed “A Packet Path Selector Which
`B.
`Selects Between The Network Interfaces On A Per-Session
`Basis To Promote Load-Balancing” .................................................. 24
`Karol Discloses The Claimed “Step Of Sending A Packet To
`The Controller Site Interface Is Repeated As Multiple Packets
`Are Sent, And The Controller Sends Different Packets Of A
`Given Message To Different Parallel Networks.” ............................. 26
`
`C.
`
`V.
`
`VI. CONCLUSION ............................................................................................. 26
`
`
`
`-ii-
`
`

`

`IPR2016-00976
`Patent No. 6,775,235 B2
`
`TABLE OF AUTHORITIES
`
`
`
`CASES
`Chef America, Inc. v. Lamb-Weston, Inc.,
`358 F.3d 1371 (Fed. Cir. 2004) ............................................................................ 2
`
`Page(s)
`
`Gart v. Logitech, Inc.,
`254 F.3d 1334 (Fed. Cir. 2001) .......................................................................... 13
`
`Genentech, Inc. v. Chiron Corp.,
`112 F.3d 495 (Fed. Cir. 1997) ............................................................................ 13
`
`In re Bigio,
`381 F.3d 1320 (Fed. Cir. 2004) ............................................................................ 7
`
`In re NTP, Inc.,
`654 F.3d 1279 (Fed. Cir. 2011) ............................................................................ 6
`
`In re Paulsen,
`30 F.3d 1475 (Fed. Cir. 1994) .............................................................................. 4
`
`Oatey Co. v. IPS Corp.,
`514 F.3d 1271 (Fed. Cir. 2008) ............................................................................ 3
`
`Perreira v. Dep’t of Health and Human Serv.,
`33 F.3d 1375 (Fed. Cir. 1994) .............................................................................. 9
`
`Phillips v. AWH Corp.,
`415 F.3d 1303 (Fed. Cir. 2005) (en banc) ............................................................ 1
`
`Verizon Servs. Corp. v. Vonage Holdings Corp.,
`503 F.3d 1295 (Fed. Cir. 2007) ............................................................................ 4
`
`OTHER AUTHORITIES
`
`37 CFR 1.75. ............................................................................................................ 22
`
`
`
`-iii-
`
`

`

`IPR2016-00976
`Patent No. 6,775,235 B2
`
`1001
`
`1002
`
`1003
`
`1004
`1005
`
`1006
`
`1007
`
`1008
`
`
`Exhibit No. Description
`U.S. Patent No. 6,775,235 by Sanchaita Datta and Ragula Bhaskar
`entitled “Tools and Techniques for Directing Packets Over
`Disparate Networks” (“the ’235 Patent”)
`File History for U.S. Patent No. 6,775,235
`U.S. Patent No. 7,406,048 by Sanchaita Datta and Ragula Bhaskar
`entitled “Tools and Techniques for Directing Packets Over
`Disparate Networks” (“the ’048 Patent”)
`File History for U.S. Patent No. 7,406,048
`Declaration of Dr. Kevin Negus
`U.S. Patent No. 6,628,617 by Mark John Karol and Malathi
`Veeraraghavan entitled “Technique for Internetworking Traffic on
`Connectionless and Connection-Oriented Networks” (“Karol”)
`W.R. Stevens, “TCP/IP Illustrated Volume 1, the Protocols,”
`Addison-Wesley Professional Computing Series, 1994, ISBN-0-
`201-63346-9 (“Stevens”).
`February 1, 2016 Order granting Motion to Transfer to the Western
`Division of the Eastern District of North Carolina, D.I. 57 in 6:15-
`cv-00458-RWS in the Eastern District of Texas
`U.S. Patent No. 6,748,439 by David R. Monachello et al. entitled
`“System and Method for Selecting Internet Service Providers from a
`Workstation that is Connected to a Local Area Network”
`(“Monachello”)
`FatPipe’s Infringement Contentions
`William Stallings, “Data and Computer Communications,” Prentice-
`Hall, 5th Edition, 1997, ISBN-81-203-1240-6, (“Stallings”)
`Office Action dated 4/13/2012 for U.S. Application No. 10/034,197
`Office Action dated 2/2/2012 for U.S. Application No. 10/034,197
`FatPipe’s Proposed Modifications to Claim Construction
`U.S. Patent No. 6,317,431 by Terence G. Hodgkinson and Alan W.
`O’Neill entitled “ATM Partial Cut-Through” (“Hodgkinson”)
`Adaptive Private Networking Configuration Editor User’s Guide,
`APNware Release 2.5 (FATPIPE-001374-1448)
`Deposition transcript of Mr. Joel Williams, IPR2016-00976 and
`IPR2016-00977, May 10, 2017
`“Routing Lookups in Hardware at Memory Access Speeds” by
`Pankaj Gupta, Steven Lin, and Nick McKeown, April 1998
`U.S. Patent No. 7,047,313 by Keith Broerman entitled “Method For
`
`1009
`
`1010
`1011
`1012
`1013
`1014
`1015
`
`1016
`
`1017
`
`1018
`1019
`
`
`
`-iv-
`
`

`

`
`Exhibit No. Description
`Redirecting Packetized Data Associated With A Destination
`Address In A Communication Protocol Layer To A Different
`Destination Address In A Different Protocol Layer”
`U.S. Patent No. 7,554,930 by Michael Gaddis et al. entitled
`“Internet Route Deaggregation And Route Selection Preferencing”
`IEEE Xplore Record of “Routing Lookups in Hardware at Memory
`Access Speeds” by Pankaj Gupta, Steven Lin, and Nick McKeown,
`April 1998
`
`1020
`
`1021
`
`IPR2016-00976
`Patent No. 6,775,235 B2
`
`
`
`
`
`-v-
`
`

`

`IPR2016-00976
`Patent No. 6,775,235 B2
`
`
`I.
`
`INTRODUCTION
`
`Talari’s Petition and FatPipe’s Patent Owner's Response (“POR”) show a
`
`consistent pattern—Talari relies on the text of the patent and prior art references,
`
`while FatPipe ignores the text and relies instead on the ipse dixit of Mr. Williams’s
`
`opinion. And while Talari details why the challenged claims are anticipated by
`
`Karol and are nothing more than obvious combinations of prior art methods used in
`
`a predictable manner, FatPipe repeatedly recites Mr. Williams’s conclusions that
`
`ignore the plain meaning of the challenged claims and the disclosures of the
`
`applicable references and elements that indisputably existed in the prior art. But
`
`FatPipe cannot amend its claims simply by proffering Mr. Williams’s
`
`declaration—particularly when that declaration contradicts what the patent and the
`
`references plainly say.
`
`II.
`
`FATPIPE’S CLAIM CONSTRUCTIONS ARE IMPROPER
`
`In the district court litigation, FatPipe argued that no construction was
`
`necessary for the phrases it seeks to construe here. (Petition, 7-8; See Ex. 1014.)
`
`The Board also agreed no construction is needed. (Paper No. 7, 6.) Now, in an
`
`about-face calculated to dodge the prior art, FatPipe argues that the claims need to
`
`be construed and narrowed by improperly importing limitations into the claims
`
`and excluding embodiments described in the specification. Phillips v. AWH Corp.,
`
`
`
`-1-
`
`

`

`IPR2016-00976
`Patent No. 6,775,235 B2
`
`
`415 F.3d 1303, 1320 (Fed. Cir. 2005) (en banc). But FatPipe cannot re-write the
`
`claims to say something they do not.
`
`The proposed terms need no construction because they are made up of easily
`
`understood terms, and their plain meaning is entirely consistent with the
`
`specification of the ‘235 Patent. Chef America, Inc. v. Lamb-Weston, Inc., 358
`
`F.3d 1371, 1373 (Fed. Cir. 2004).
`
`A.
`
`FatPipe’s Narrow Construction Of “Selects Between Network
`Interfaces On A Per-Packet Basis” / “Make Network Path
`Selections On A Packet-By-Packet Basis” Is Legally Flawed
`
`FatPipe argues the phrase “selects between network interfaces on a per-
`
`packet basis” in claim 4 and the phrase “make network path selections on a packet-
`
`by-packet basis” in claim 9 (“Disputed Phrases”) mean “for each packet, makes a
`
`discrete choice between network paths/interfaces.” (POR, 9.) As shown in the
`
`Petition, Karol’s gateway makes a routing selection (i.e., “discrete choice”)
`
`between the connection oriented or connectionless network (i.e., “network
`
`paths/interfaces”) for each datagram (i.e., packet). (See, e.g., Petition, 14-16; Ex.
`
`1006, 5:23-25; Paper No. 7, 11-12.)
`
`In an attempt to avoid Karol, FatPipe seeks to narrow its construction such
`
`that “discrete choice” is construed to exclude the embodiment in which packets are
`
`routed “based on a single selection that applies to multiple subsequent packets.”
`
`(POR, 11-12, 26.) FatPipe also argues that the only way to make a discrete choice
`
`
`
`-2-
`
`

`

`IPR2016-00976
`Patent No. 6,775,235 B2
`
`
`(or network path selection) is to update a routing database. (POR, 27; Ex. 1017,
`
`64:12-21.) FatPipe’s constructions are not supported by the intrinsic record.
`
`1.
`
`The Specification Does Not Support FatPipe’s
`Constructions
`
`First, the claims do not exclude the embodiment in which packets are routed
`
`“based on a single selection that applies to multiple subsequent packets.” (Ex.
`
`1001, 17:47-52.) Claim 4 requires a selection for each packet (“per-packet”) in
`
`order to “send the packet through the network interface that was selected.”
`
`Whether or not the selection applies to a single packet or multiple subsequent
`
`packets is not specified. Claim 9 allows “repeated instances” of “selecting,” but
`
`does not exclude the instance in which a single selection applies to multiple
`
`subsequent packets. The “inventive method” contemplates a range of solutions,
`
`including the embodiment in which a packet path selector makes an initial path
`
`selection that applies to each packet (“per-packet”) without the need to repeat the
`
`selecting step 908. (Ex. 1001, 16:22-27, Fig. 9.) FatPipe’s new and narrow
`
`construction excludes this embodiment. Oatey Co. v. IPS Corp., 514 F.3d 1271,
`
`1276 (Fed. Cir. 2008).
`
`Second, the specification does not ascribe any limited meaning to the
`
`Disputed Phrases and never uses the term “discrete choice.” (Ex. 1017, 63:8-64:3.)
`
`An inventor may provide a meaning for a term that is different from its ordinary
`
`
`
`-3-
`
`

`

`IPR2016-00976
`Patent No. 6,775,235 B2
`
`
`meaning, but that was not done for the Disputed Phrases. In re Paulsen, 30 F.3d
`
`1475, 1480 (Fed. Cir. 1994). The inventors knew how to define terms. (See, e.g.,
`
`Ex. 1001, 1:56-2:33.) They did not provide definitions for the Disputed Phrases,
`
`and there is no basis for FatPipe’s requirement of a “discrete choice” (i.e., an
`
`update to the routing database prior to the network path selection) for each packet.
`
`Third, the specification identifies the “inventive configurations” as methods
`
`and systems to coordinate traffic over parallel networks without manual
`
`intervention. (Ex. 1001, 9:30-33.) Figure 9, which “is a flowchart further
`
`illustrating methods of the present invention…” describes embodiments to do just
`
`that. (Ex. 1001, 13:32-38 (emphasis added).) The specification makes clear that
`
`the “path selecting step 908 may be performed once per packet, or a given
`
`selection may pertain to multiple packets.” (Ex. 1001, 14:44-46 (emphasis
`
`added).) FatPipe’s improper construction excludes the embodiment in which “a
`
`given [path] selection may pertain to multiple packets.” (POR, 12.) Verizon Servs.
`
`Corp. v. Vonage Holdings Corp., 503 F.3d 1295, 1305 (Fed. Cir. 2007).
`
`FatPipe admits the specification discloses that a given selection may pertain
`
`to multiple packets (POR, 13-14), but mischaracterizes this embodiment of the
`
`invention as “prior art.” The stated problem in the prior art was the need for
`
`“manual reconfiguration” to switch data between disparate networks. (See, e.g.,
`
`Ex. 1001, 9:30-33 (“Unlike the configuration shown in [prior art] FIG. 1, the
`
`
`
`-4-
`
`

`

`IPR2016-00976
`Patent No. 6,775,235 B2
`
`
`inventive configurations in FIGS. 6 and 10 do not require manual intervention by
`
`network administrators to coordinate traffic flow over parallel networks.”
`
`(emphasis added)), 2:56-65, 4:17-24, 7:38-42.) One such disclosed solution
`
`includes making selections that pertain to multiple packets, but FatPipe’s
`
`construction excludes this embodiment.
`
`Fourth, FatPipe argues that the only way to make a “discrete choice” is to
`
`make a router update, but this argument is not supported by the specification.
`
`Figure 9 identifies the process of the claimed invention, which includes steps 900
`
`and 902 that gather information used in routing packets. (Id., 13:39-14:3.) These
`
`steps correlate to the updating of a routing database. (See Ex. 1007, 37-39, 112-
`
`113.) FatPipe’s construction contradicts the intrinsic record because it would
`
`require steps 900 and 902 to occur before every routing decision so that the routing
`
`table/information is updated. Figure 9 does not require steps 900 and 902 to occur
`
`before every routing decision, and the specification makes clear that “steps may be
`
`omitted unless required by the claims….” (Ex. 1001, 14:14-17.) Claims 4 and 9
`
`do not require routing update steps 900 and 902 prior to each packet routing
`
`decision, and FatPipe’s own annotations highlight the omission of these steps.
`
`(POR, 13-14 (annotating Fig. 9); Ex. 1001, 16:15-21.)
`
`Fifth, when cross-examined about where the claims specify how to make a
`
`“discrete choice,” Mr. Williams pointed to dependent claim 12 reciting “dynamic
`
`
`
`-5-
`
`

`

`IPR2016-00976
`Patent No. 6,775,235 B2
`
`
`load-balancing.” (Ex. 1017, 66:6-67:5.) Claim 12 does not depend from claim 4
`
`or 9, and does not support his attempt to introduce limitations in the Disputed
`
`Phrases in claims 4 and 9.
`
`Finally, FatPipe’s reliance on In re NTP, Inc. is misplaced. (POR, 12.) In In
`
`re NTP, Inc., the background section of the patent presented a definition of
`
`“electronic mail” that excluded prior art technologies. In re NTP, Inc., 654 F.3d
`
`1279, 1289 (Fed. Cir. 2011). In contrast, the ‘235 Patent makes no such definition
`
`and describes that “routing of packets based on a single selection that applies to
`
`multiple subsequent packets” is an embodiment. The Disputed Phrases should not
`
`be limited to exclude this embodiment.
`
`2.
`
`The Board Should Adopt The Broadest Reasonable
`Interpretation
`
`The Disputed Phrases should be construed as “for each packet a network
`
`interface/path is chosen,” which is the broadest reasonable interpretation (“BRI”).
`
`This construction is supported by the specification. The patent discloses that the
`
`“path selecting step 908 may be performed once per packet, or a given selection
`
`may pertain to multiple packets.” (Ex. 1001, 14:44-46.) In either case, a network
`
`interface/path is chosen for each packet.
`
`
`
`-6-
`
`

`

`IPR2016-00976
`Patent No. 6,775,235 B2
`
`
`
`B.
`
`FatPipe’s Construction Improperly Narrows “Dynamic Load-
`Balancing”
`
`FatPipe construes “dynamic load-balancing” to mean “distributing packets
`
`based on actual traffic assessed after the packet arrives.” (POR, 15.) The
`
`inventors did not provide any such definition (Ex. 1017, 121:1-4), and the
`
`specification does not support FatPipe’s narrow construction. If a construction is
`
`necessary, it should be construed as “sending packets in distributions that balance
`
`the load of a given network, router, or connection relative to other networks,
`
`routers, or connections available to the controller without manual intervention.”
`
`(Ex. 1001, 9:30-33, 11:21-24, 11:33-38.)
`
`FatPipe’s Construction Is Unsupported
`
`1.
`“[T]he PTO should only limit the claim based on the specification or
`
`prosecution history when those sources expressly disclaim the broader definition.”
`
`In re Bigio, 381 F.3d 1320, 1325 (Fed. Cir. 2004). Here, none of the passages
`
`cited by FatPipe amount to an express disclaimer limiting “dynamic load-
`
`balancing” to the distribution of packets based on actual traffic assessed after the
`
`packet arrives. FatPipe’s construction is a blatant, improper attempt to read
`
`additional limitations into the claims.
`
`FatPipe cites a description of the prior art in which traffic could not readily
`
`switch between networks without manual intervention. (POR, 15.) FatPipe argues
`
`
`
`-7-
`
`

`

`IPR2016-00976
`Patent No. 6,775,235 B2
`
`
`this passage means “dynamic load-balancing” must be limited to balancing traffic
`
`in response to “actual traffic”—traffic that existed at or shortly after the time of the
`
`packet’s arrival. (POR, 16.) FatPipe misstates the meaning of this passage. The
`
`passage only highlights the difference between the prior art (manual switchover)
`
`and the alleged invention (no manual intervention). It is unrelated to balancing
`
`traffic based on traffic loads that existed at the time of a packet’s arrival or at some
`
`undefined period thereafter. The specification uses “dynamic load-balancing” to
`
`mean that no manual switchover is required, as discussed above in Section II.A.1.
`
`FatPipe also misinterprets a passage disclosing that the packet path “is
`
`determined after the packet arrives, e.g., using per-packet dynamic load
`
`balancing.” (POR, 16.) This reference to dynamic load-balancing is to the alleged
`
`invention being able to route packets on parallel networks without manual
`
`intervention. The passage does not refer to “actual traffic” or limit when the traffic
`
`must be assessed.
`
`FatPipe’s reliance on its declarant is equally unavailing. Mr. Williams cites
`
`only to a dictionary definition of “actual” to conclude that “actual traffic” must be
`
`“traffic existing at or shortly after the time of the packet’s arrival, as opposed to
`
`traffic conditions that existed sometime in the past.” (Ex. 2003, ¶44.) His
`
`dictionary defines “actual” as “existing” but offers nothing to support his timing
`
`requirements. “Shortly after the time of the packet’s arrival” is “sometime in the
`
`
`
`-8-
`
`

`

`IPR2016-00976
`Patent No. 6,775,235 B2
`
`
`past.” Mr. Williams offers no explanation as to how this claim element is met.
`
`(Ex. 1017, 131:12-17.) When asked how much time after the packet’s arrival
`
`traffic should be measured, Mr. Williams said it is “undefined.” (Ex. 1017,
`
`128:20-129:12.) Furthermore, Mr. Williams fails to identify where the packet
`
`arrives, offering only a “loose answer” of “the controller.” (Id., 123:1-7.) Mr.
`
`Williams’s construction should be rejected. See Perreira v. Dep’t of Health and
`
`Human Serv., 33 F.3d 1375, 1377 n.6 (Fed. Cir. 1994); Trial Practice Guide 48763.
`
`FatPipe’s references to Talari’s own documents are also unavailing because,
`
`among other things, they are not contemporary (from 2017 and 2015), and do not
`
`disclose that “dynamic load-balancing” must be based on actual traffic assessed
`
`after the packet arrives as alleged by FatPipe. (Exs. 2008 and 2009.)
`
`2.
`
`The Board Should Adopt The Broadest Reasonable
`Interpretation
`
`FatPipe’s argument ignores the patentees’ definition of load-balancing in the
`
`specification, which states: “send packets in distributions that balance the load of a
`
`given network, router, or connection relative to other networks, routers, or
`
`connections available to the controller 602.” (Ex. 1001, 11:21-24.) Prior art load-
`
`balancing “use a per-department and/or per-router basis for dividing traffic” that
`
`could be changed only with manual intervention. (Id., 11:33-38.) The
`
`specification uses the term “dynamic” to denote load-balancing using the alleged
`
`
`
`-9-
`
`

`

`IPR2016-00976
`Patent No. 6,775,235 B2
`
`
`invention which can distribute packets between networks without the need for
`
`manual intervention. (Id., 9:30-33.) The BRI of “dynamic load-balancing” is
`
`“sending packets in distributions that balance the load of a given network, router,
`
`or connection relative to other networks, routers, or connections available to the
`
`controller without manual intervention.”
`
`III. CLAIM 4 IS ANTICIPATED OR RENDERED OBVIOUS
`FatPipe concedes Karol discloses all of the elements of claim 4 except for
`
`“packet path selector which selects between network interfaces on a per-packet
`
`basis” and “site interface.” But these elements are also disclosed by Karol.
`
`A. Karol Discloses The Claimed “Packet Path Selector Which Selects
`Between Network Interfaces On A Per-Packet Basis”
`
`Under the BRI, Karol discloses packet path selection on a per-packet basis.
`
`(Petition, 10-17; Section II supra; Paper No. 7, 11-12.) But even under FatPipe’s
`
`construction, Karol discloses the claimed elements.
`
`FatPipe makes two arguments to suggest that Karol does not select a
`
`network on a per packet basis. (POR, 24.) Both are wrong. First, FatPipe
`
`contends that Karol routes packets based on the forwarding database’s pre-
`
`computed route resulting in “no distinct selecting for each particular packet.”
`
`(POR, 24, 26.) This argument ignores Karol’s distinct selection for each particular
`
`packet: at the CL-CO gateway “a determination is made if that packet should be
`
`
`
`-10-
`
`

`

`IPR2016-00976
`Patent No. 6,775,235 B2
`
`
`carried by CO network 160.” (Ex. 1006, 5:23-25.) The determination is not based
`
`on a “pre-computed route.” It is based on service requirements and “the traffic
`
`situation in the CL and CO networks” to evaluate whether a particular packet may
`
`be redirected. (Id., 5:35-38; Paper No. 7, 15.) Even assuming Karol were limited
`
`to OSPF (which is not required (Ex. 1006, 14:20-22; Ex. 1017, 101:14-18)),
`
`Stallings discloses that OSPF does not use “pre-computed” routes but rather
`
`“computes” routes. (Ex. 1011, 557.)
`
`Second, FatPipe argues that Karol cannot select a network on a per packet
`
`basis (i.e., make a “discrete choice”) because Karol cannot update the forwarding
`
`database at the same rate that packets are arriving. (POR, 24, 27.) Mr. Williams
`
`testified that for Karol to make a discrete choice, the router update must occur
`
`faster. (Ex. 1017, 64:12-21.) FatPipe reasons that packets typically arrive every
`
`few microseconds or less and because the forwarding database is only updated at
`
`the very quickest every 1 second, there cannot be a network selection for each
`
`packet. (POR, 27.)
`
`Nothing in either the specification or Karol requires a specific packet rate.
`
`(Ex. 1017, 103:6-12.) Even under FatPipe’s post hoc construction, packets in
`
`Karol arriving at a rate of 1 packet per second (or slower) enable Karol to make a
`
`“discrete choice” because the routing table will be updated before each packet.
`
`Alternatively, Karol meets FatPipe’s construction by using routing protocols other
`
`
`
`-11-
`
`

`

`IPR2016-00976
`Patent No. 6,775,235 B2
`
`
`than OSPF that enable routing lookups and updates according to the fastest
`
`commercially available packet transmission rate. (Ex. 1006, 14:20-22; See, e.g.,
`
`Ex. 10181, Abstract and Section 5.)
`
`B. Karol Discloses The Claimed “Site Interface”
`Karol discloses a “site interface” in two ways:
`
`(1) when the “controller” is viewed to be Karol’s
`disclosure of the CL-CO gateway in combination with
`one or more routers and/or switches (e.g., the routers
`and/or switches in node 111), then the site interface is the
`interface between source 101 and node 111; or
`
`(2) when the “controller” is viewed to be Karol’s CL-CO
`gateway alone, then the site interface is the interface with
`the routers and/or switches in node 111. (Petition, 12.)
`
`Regarding (1), FatPipe argues that Karol’s “controller” is limited to the CL-
`
`CO gateway. (POR, 28.) Regarding (2), FatPipe argues that the CL-CO gateway
`
`interface with CL network 111 is a “network interface” instead of a “site
`
`interface.” (Id.) Both arguments fail.
`
`First, the claims do not require the controller to be a single device. The
`
`claims explicitly define the controller as being a plurality of devices. (Ex. 1001,
`
`17:39-55.) The use of “comprising” in Claim 4 when listing the elements that are
`
`included with the controller is inclusive or open-ended and does not exclude
`
`1 Ex. 1018 is an IEEE paper published in INFOCOM on April 1998 that was
`publicly accessible prior to the critical date. (Ex. 1021.)
`
`
`
`-12-
`
`

`

`IPR2016-00976
`Patent No. 6,775,235 B2
`
`
`additional, un-recited elements such as routers and/or switches in node 111.
`
`Genentech, Inc. v. Chiron Corp., 112 F.3d 495, 501 (Fed. Cir. 1997)
`
`(“Comprising” is a term of art used in claim language which means that the named
`
`elements are essential, but other elements may be added and still form a construct
`
`within the scope of the claim.) The controller is not precluded from comprising the
`
`CL-CO gateway with the routers and/or switches in node 111.
`
`Routers and/or switches are common in the field of computer networking to
`
`connect devices. (Ex. 1001, 1:17-24; Ex. 1017, 143:4-9.) The claims to do not
`
`preclude the use of routers and switches as part of the “controller” to connect to the
`
`site because there is no requirement of a “direct connection” between the site and
`
`controller. Mr. Williams relies on Figure 7 of the specification to opine that the
`
`claims require a “direct connection” and that the embodiment in which the
`
`controller is spread across multiple routers or switches is excluded. (Ex. 1017,
`
`135:5-10, 136:20-137:3.) But the figure is not limiting. Gart v. Logitech, Inc., 254
`
`F.3d 1334, 1342 (Fed. Cir. 2001) (“drawings are not meant to represent ‘the’
`
`invention or to limit the scope of coverage defined by the words used in the claims
`
`themselves”). Nothing in the specification or claims limits the claimed
`
`“controller” to a single device. (See Ex. 1001, 7:21-25, 10:59-65, 11:64-12:4.)
`
`And FatPipe offers no reason that switches could not be used as part of the
`
`
`
`-13-
`
`

`

`IPR2016-00976
`Patent No. 6,775,235 B2
`
`
`controller. (POR, 29-30 (opining only that routers are not typically thought of as a
`
`single device).)
`
`Regarding (2), even assuming the “controller” is the CL-CO gateway alone,
`
`Karol discloses a “site interface.” (Petition, 12.) Although FatPipe admits Karol’s
`
`site (source 101) has a site interface connected to node 111 (Ex. 1017, 132:5-
`
`133:11), FatPipe argues there must be a “direct connection” to the CL-CO
`
`gateway. (Ex. 1017, 133:14-17.) This is contradicted by the ‘235 Patent which
`
`discloses that network devices such as a LAN can interface between the site and
`
`the controller. (Ex. 1001, 13:1-3 (“One then connects the site interface 702 to a
`
`site 102 to receive packets from a computer (possibly via a LAN) at the site 102.”
`
`(emphasis added)); Ex. 1017, 143:4-9.) Karol’s node 111, which can consist of
`
`routers and switches, interfaces between source 101 and the CL-CO gateway in the
`
`same manner as that disclosed in the ‘235 Patent.
`
`FatPipe’s position that the site must “directly” interface with the controller is
`
`also inconsistent with its position in the district court litigation where FatPipe’s
`
`infringement contentions relied on the Talari “controller” situated between a LAN
`
`switch and the alleged site:
`
`
`
`-14-
`
`

`

`
`
`
`IPPR2016-000976
`
`Patent No
`
`. 6,775,2355 B2
`
`
`
`Talaari Device
`
`
`
`Sitte
`
`
`
`LANN switch
`
`
`
`(Ex. 1010, 1-2.)
`
`
`
`
`
`AA POSITA would undderstand frrom the dissclosure in n Karol thatt under thee
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`BRI, Kaarol discloses a site innterface. ((Ex. 1005,
`
`
`
`
`
`
`
`
`
`IV. CCLAIMS 55 AND 7-115 ARE ANNTICIPAATED OR
`
`
`
`
`
`
`
`OBVIOUSS
`
`¶169.)
`
`
`
`RENDERRED
`
`
`
`for “obttaining at least two knnown locaation addre
`
`
`
`
`
`
`
`
`
`
`
`
`
`ss ranges wwhich havee associateed
`
`
`
`networkks.” But ass the Boardd already ddeterminedd, Karol disscloses thiss element ttoo.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`(Paper NNo. 7, 13-114.)
`
`
`
`
`
`-15-
`
`O F
`
`
`
`FatPipe conncedes thatt Karol disccloses all oof the elemments of claaim 5 exceept
`
`
`
`
`
`
`
`
`
`
`
`
`
`

`

`IPR2016-00976
`Patent No. 6,775,235 B2
`
`
`
`Regarding the dependent claims, FatPipe makes no separate argument as to
`
`claims 10, 14, and 15, and the only disputes relate to claims 7-9 and 11-13.
`
`A. Claim 5 Is Invalid Over Karol Alone Or In Combination With
`Stallings Which Disclose “Obtaining At Least Two Known Location
`Address Ranges Which Have Associated Networks”
`
`Consistent with the specification, one or more non-contiguous addresses
`
`comprise an “address range.” (Petition, 17-19; Ex. 1005, ¶¶97, 99, 223.) FatPipe
`
`construes “address range” to require “contiguous addresses” and argues Karol’s
`
`CO database does not have an “address range.” (POR, 36.)
`
`FatPipe’s requirement of “contiguous addresses” is contradicted by the
`
`specification, which discloses that “a network may have more than one associated
`
`contiguous range of addresses which collectively constitute the address range for
`
`that network.” (Ex. 1001, 13:46-49 (emphasis added).) An “address range” may
`
`be comprised of a collection of addresses that are not contiguous with each other.
`
`No disclosure exists in the specification for excluding a single address from
`
`being an “address range.” (Ex. 1017, 50:10-21.) FatPipe argues that the
`
`specification defines “address range” by its use of “.x.x” when referring to
`
`example addresses, but this is not the BRI. (POR, 36; see also, Ex. 1019, 8:43-46
`
`(referring to an address class (or range) as consisting of “one or more addresses”),
`
`Ex. 1020, 16:43-47 (referring to an address range of a single address –
`
`140.150.0.0).) The use of “.x.x” denotes the optional use of a subnet mask. (Ex.
`
`
`
`-16-
`
`

`

`IPR2016-00976
`Patent No. 6,775,235 B2
`
`
`1001, 13:40-43 (“Address ranges may be specified as partial addresses….”)
`
`(emphasis added); Ex. 1017, 47:3-5.) An address range in the form of “.x.x” is not
`
`required.
`
`No technical reason