`
`
`
`
`
`
`
`
`Paper No.
`Filed: January 17, 2020
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`
`
`
`
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`
`
`
`
`THE MANGROVE PARTNERS MASTER FUND, LTD., APPLE INC., and
`BLACK SWAMP IP, LLC,
`Petitioner
`v.
`VIRNETX INC.,
`Patent Owner
`
`
`
`Case IPR2015-010471
`Patent 7,490,151
`
`
`
`
`
`Patent Owner’s Sur-Reply Brief
`
`
`
`
`
`
`1 Apple Inc. and Black Swamp IP, LLC, who filed petitions in IPR2016-00063 and
`IPR2016-00167, respectively, have been joined as a Petitioner in the instant
`proceeding.
`
`

`

`
`
`Case No. IPR2015-01047
`
`TABLE OF CONTENTS
`Introduction ...................................................................................................... 1
`Claim Construction .......................................................................................... 1
`A.
`“Client” .................................................................................................. 1
`1.
`Petitioners’ Proposed Construction Finds No Support in
`the Claim Language .................................................................... 1
`The Specification Does Not Supports Petitioners’
`Construction ................................................................................ 5
`The Extrinsic Evidence Does Not Support Petitioners’
`Construction ................................................................................ 8
`“Between [A] and [B]” ........................................................................10
`B.
` Kiuchi Does Not Anticipate Claims 1, 2, 6-8, and 12-14 .............................11
` Kiuchi Combined with Rescorla Does Not Render Obvious Claims 1,
`2, 6-8, and 12-14 ............................................................................................15
` Kiuchi Combined with RFC 1034, With or Without Rescorla, Does
`Not Render Obvious Claims 1, 2, 6-8, and 12-14 .........................................17
` The Board Should Not Rely on Dr. Guerin’s Testimony ..............................26
` The Board Should Draw an Adverse Inference Regarding the RPI or
`Privity Relationship Between Mangrove and RPX, and Terminate the
`Proceeding .....................................................................................................26
` A New Panel Should Consider this Remand Proceeding To Avoid
`Constitutional Concerns ................................................................................28
` The Proceedings Should Be Terminated in View of 35 U.S.C.
`§ 315(b) ..........................................................................................................28
`Conclusion .....................................................................................................29
`
`
`
`
`
`2.
`
`3.
`
`i
`
`
`
`
`
`
`
`
`
`
`

`

`
`
`
`
`Case No. IPR2015-01047
`
`TABLE OF AUTHORITIES
`
` Page(s)
`
`Cases
`Arctic Cat Inc. v. Polaris Indus., Inc.,
`No. 19-1440, 2019 WL 7050133 (Fed. Cir. Dec. 23, 2019) .............................. 27
`Arthrex, Inc. v. Smith & Nephew, Inc.,
`941 F.3d 1320 (Fed. Cir. 2019) .......................................................................... 28
`Founding Church of Scientology of Washington, D.C. v. Webster,
`802 F.2d 1448 (D.C. Cir. 1986) .......................................................................... 26
`Omega Eng’g, Inc. v. Raytek Corp.,
`334 F.3d 1314 (Fed. Cir. 2003) ............................................................................ 4
`Phillips v. AWH Corp.,
`415 F.3d 1303 (Fed. Cir. 2015) (en banc) ............................................................ 5
`Samsung Elecs. Co. v. Infobridge Pte. Ltd.,
`929 F.3d 1363 (Fed. Cir. 2019) .......................................................................... 16
`VirnetX Inc. v. Cisco Sys. Inc.,
`767 F.3d 1308 (Fed. Cir. 2014 ...................................................................... 14, 24
`VirnetX Inc. v. Mangrove Partners Master Fund, Ltd.,
`778 F. App’x 897 (Fed. Cir. 2019) ................................................... 15, 17, 21, 28
`Statutes
`35 U.S.C. § 315(b) ................................................................................................... 28
`
`
`
`
`ii
`
`

`

`
`
`
`
`
`
`Case No. IPR2015-01047
`
`Introduction
`As VirnetX previously demonstrated, Petitioners’ grounds of unpatentability
`
`are deficient in a number of ways. Petitioners’ reply brief tries to argue otherwise.
`
`As explained below, those arguments are without merit.
`
` Claim Construction
`“Client”
`A.
`Petitioners’ Proposed Construction Finds No Support in the
`1.
`Claim Language
`Petitioners’ claim construction argument is based on a faulty premise—that
`
`the term “client,” in its “plain and ordinary meaning,” denotes “‘a device, computer,
`
`system, or program from which a data request to a server is generated.’”
`
`(Petitioner’s Reply Remand Brief, Paper 106 (“Reply”) at 1.) That argument is
`
`unsound. Petitioners’ original basis for their proposed construction—as reflected in
`
`their opening brief on remand—was the assertion that this is how a skilled artisan
`
`would have understood the term “client computer.” (Petitioners’ Remand Brief,
`
`Paper 104 (“PRB”) at 6-7.) In its opposition brief, VirnetX demonstrated that
`
`Petitioners’ support for that assertion—a claim that both their and VirnetX’s experts
`
`“agreed that a skilled person would have understood a conventional ‘client’ to be
`
`any application that generates a request for data from a server” (PRB at 7)—
`
`mischaracterized expert testimony. (See Patent Owner’s Opposition Brief,
`
`Paper 105 (“Opp.”) at 5-6.) Neither expert provided an opinion that would support
`
` 1
`
`
`
`

`

`
`Petitioners’ argument as to how a skilled artisan would have understood the term
`
`Case No. IPR2015-01047
`
`“client.” And as VirnetX demonstrated, one of skill in the art, reading the claims in
`
`context of the specification—as required—would have understood “client” to mean
`
`a “user’s computer.” (See Opp. 2-5.)2
`
`Petitioners’ reply scarcely defends their arguments based on the purported
`
`understanding of a skilled artisan. Instead, they retreat to a different position, urging
`
`that a “client” is a “device, computer, system, or program from which a data request
`
`to a server is generated” because claim 1 recites that the client sends a “DNS
`
`request.” (Reply at 2.) That argument is nonsensical. A claim term is not defined
`
`
`2 Petitioners are incorrect in contending that VirnetX’s expert, Dr. Monrose,
`
`“admitted the ’151 patent did not give ‘client’ a special definition.” (Reply at 1.)
`
`The portions of Dr. Monrose’s deposition that Petitioners cite (Ex. 1036, 74:15-
`
`75:16, 95:1-5) contain no discussion of that issue, much less any “admission.”
`
`Regardless, VirnetX’s argument is not that the ’151 patent advanced any “special
`
`definition” for the term “client,” but that the language of the claims and the
`
`specification make clear that the term “client” means a “user’s computer.” (See Opp.
`
`2-5.) Petitioners, moreover, do not even attempt to respond to VirnetX’s detailed
`
`explanations of how Petitioners’ brief mischaracterized the testimony of Dr. Guerin
`
`and Dr. Monrose. (See Reply at 9; Opp. 5-6.)
`
`
`
`2
`
`

`

`
`by the description of one of the claimed steps it performs; that would deprive the
`
`Case No. IPR2015-01047
`
`term “client” of any independent meaning. If a claim recites “a wire to transmit
`
`current,” that does not mean a “wire” can be defined as anything that can transmit
`
`current (such as a human body).
`
`Along similar lines, Petitioners argue that the claims of the ’151 patent “do
`
`not require checking a user’s status or credentials,” but “instead specify actions
`
`based on attributes of the request or the device making the request.” (Reply at 3.)
`
`That the client sends a request does not by itself define what the computer is in the
`
`context of the claims and specification of the ’151 patent. Claim 1 requires an
`
`encrypted channel between the client and secure server; an arrangement that the
`
`specification repeatedly describes as involving a user’s computer. (See Ex. 1001 at
`
`37:66-38:2 (describing that the client is a “user’s computer” because the channel is
`
`established “between [a] user computer” and a “secure target site”) (emphasis
`
`added)); see also Opp. 3.) Moreover, the clarification that the client is sending a
`
`“request” is contained within the language of claims 1, 7, and 13. This is not a
`
`situation where one claim recites the term “client,” and a different claim contains a
`
`clarification that could narrow the definition of “client.”
`
`Having failed to find adequate support for their preferred definition in the
`
`claim language of the ’151 patent, Petitioners resort to claim language of different
`
`patents, containing different terms. Thus, Petitioners quote—for the first time in this
`
`
`
`3
`
`

`

`
`proceeding—U.S. Patent No. 9,386,000 (not at issue here), which recites the term
`
`Case No. IPR2015-01047
`
`“client device,” and then “adds that ‘the client device is a user device.’” (Reply at 3
`
`(emphasis removed).) Petitioners claim that this shows that a “client” cannot be a
`
`“user’s computer” because otherwise “the additional language in this claim [of the
`
`’000 patent] would be superfluous.” (Id.) To the contrary, to the extent construction
`
`of “client device” in the ’000 patent is even relevant,3 that definition shows that the
`
`specification (which the ’151 patent shares with the ’000 patent) uses the term
`
`“client” to mean the “user’s computer.” (See Opp. 7-9.) If the “client device is a
`
`user device” (as stated in the ’000 patent), it follows that a “client” in the ’151 patent
`
`should be construed as “a user’s computer.”
`
`Finally, Petitioners contends that their construction “reflects the skilled
`
`artisan’s understanding that a device in a client-server system that acts as a server in
`
`one operation can act as a client in another.” (Reply at 2 (citing Ex. 1014 at 5-6).)
`
`As VirnetX already explained, Exhibit 1014 (RFC 1945) defines the user agent as
`
`
`3 Petitioners rely (Reply at 4) on the “presum[ption] … that the same claim term in
`
`the same patent or related patents carries the same construed meaning.” Omega
`
`Eng’g, Inc. v. Raytek Corp., 334 F.3d 1314, 1334 (Fed. Cir. 2003) (citation omitted).
`
`That doctrine does not apply, however, where the claim terms are different, as “client
`
`device” and “client” are.
`
`
`
`4
`
`

`

`
`the “client” that “initiates a request,” (Opp. 5 n.2, 21-22; Ex. 2039 at 73:14-19), and
`
`Case No. IPR2015-01047
`
`so supports VirnetX’s construction. Petitioners do not attempt to rebut this
`
`argument. Moreover, Petitioners apparently concede that their construction would
`
`result in a confusing ability of the claimed components to change whether they act
`
`as a “client computer” or a “target computer.” (See Reply at 2 (“nothing in the claims
`
`precludes a ‘secure server in a first transaction from being a ‘client in a second
`
`transaction involving a different ‘secure server”) (emphasis in original).) As
`
`explained below, infra at Section V, Petitioners’ interpretation of Kiuchi fails in light
`
`of this argument.
`
`2.
`
`The Specification Does Not Supports Petitioners’
`Construction
`As VirnetX demonstrated, the specification—the “single best guide to the
`
`meaning of a disputed term,” Phillips v. AWH Corp., 415 F.3d 1303, 1315 (Fed. Cir.
`
`2015) (en banc)—indicates that the proper construction of “client computer” is a
`
`“user’s computer.” (Opp. 3-4, 6-9.) The specification uses the terms “client” and a
`
`“user’s computer” synonymously because the two terms are consistent with each
`
`other. (See, e.g., Opp. 7-8.)
`
`Petitioners contend that the specification “refutes” this construction, but their
`
`effort falls short. Petitioners first accuse VirnetX of “ignor[ing]” a passage in the
`
`specification that “uses ‘client’ without any statement or suggestion it must be a
`
`user’s device.’” (Reply at 4 (quoting Ex. 1001 at 15:61-64, Fig. 8) (emphasis in
`5
`
`
`
`

`

`
`original).) As an initial matter, this is the first time Petitioners invoked this portion
`
`Case No. IPR2015-01047
`
`of the specification—it was not cited or discussed in their opening remand brief—
`
`and so VirnetX could hardly be accused of “ignoring” something Petitioners did not
`
`argue until now. In any event, the fact that in some passages the specification uses
`
`the term “client” without any further elaboration is hardly evidence that the term
`
`does not mean a “user’s computer.” After all, the claims themselves use the term
`
`“client” without further elaboration; otherwise, there would be no need to define the
`
`term. The specification does not have to expressly define the term “client” every
`
`time it uses that term. There is no implication—contrary to Petitioners’ argument
`
`(see Reply at 5-6)—that the computer at the starting end of the encrypted/secure
`
`channel is not a “user’s computer.” This is not—as Petitioners attempt to portray
`
`it—a situation where a patent uses different terms, which “connotes different
`
`meanings,” nor is it an attempt to “import[] limitations from the specification into
`
`the claim,” (Reply at 5, 6 (internal quotation marks, emphasis, and citations
`
`omitted)).
`
`Similarly unavailing is Petitioners’ argument that “[o]ther passages use
`
`‘client’ without any reference or context suggesting the necessity of a ‘user.’”
`
`(Reply at 6.) For starters, Petitioners’ argument is factually wrong. One of the
`
`figures to which they point—Figure 31, discussed at Ex. 1001 at 7:27-30—expressly
`
`refers to a “determin[ation] that client 3103 is a validly registered user.” (Ex. 1001
`
`
`
`6
`
`

`

`
`at 44:44.) The other passages cited by Petitioners (Figures 8 and 9) are also not
`
`Case No. IPR2015-01047
`
`inconsistent with VirnetX’s proposed definition. (See supra at 5-6 (discussing
`
`Figure 8).)
`
`Petitioners then try to counter VirnetX’s argument that the specification’s
`
`statement that “[a] user’s computer 2601 includes a convention client (e.g., a web
`
`browser) 2605” (Ex. 1001 at 37:51-52, quoted in Opp. 6-7) shows that the “client”
`
`is a “user’s computer,” since it is running applications like web browsers that would
`
`run on a computer being operated by a user, and not on a proxy. (Opp. 6-7.)
`
`Petitioners contend that the statement “expressly differentiates” between “user’s
`
`computer” and “client” because (they assert) “a web browser running on the user’s
`
`computer is identified as an example … of a client.” (Reply at 4-5 (emphasis
`
`removed).) At most, that argument can show that the term “client” (as used in the
`
`’151 patent) could mean “user’s application” (as opposed to only “user’s
`
`computer”); it does not show that the term “client” cannot mean “user.”
`
`As VirnetX pointed out, the specification consistently speaks about an
`
`encrypted/secure channel being established “between user computer … and secure
`
`target site.” (Ex. 1001 at 37:66-38:2, quoted in Opp. 3; see also Ex. 1001 at 38:60-
`
`61.) Petitioners argue that those passages merely describe “a single embodiment of
`
`the invention” that “imposes no restrictions on how” to make “a determination that
`
`a user has sufficient security privileges,” and that a user’s security privileges “can
`
`
`
`7
`
`

`

`
`be checked without a user’s credentials and can be delegated to devices other than
`
`Case No. IPR2015-01047
`
`the ‘user’s computer.’” (Reply at 5 (quoting Ex. 1001 at 37:60-66, 38:46-50).) This
`
`argument misses the mark. VirnetX never argued that the user’s computer must
`
`make that security determination. Moreover, this example supports VirnetX’s
`
`construction, for one would not be checking whether the user has sufficient security
`
`privileges if there were no user.
`
`Critically, Petitioners offer no effective response to VirnetX’s point that the
`
`specification expressly distinguishes the “client” from proxies like Kiuchi’s client-
`
`side proxy. (See Opp. 4; see also Opp. 6-7.) Petitioners accuse VirnetX of
`
`“ignor[ing] that one computer can play multiple roles in a client-server system.”
`
`(Reply at 6.) That observation, however, is not responsive to VirnetX’ point that, in
`
`the context of the ’151 patent, when the specification speaks of a “client” and a
`
`proxy, it is referring to two markedly distinct items.
`
`3.
`
`The Extrinsic Evidence Does Not Support Petitioners’
`Construction
`As VirnetX demonstrated, the extrinsic evidence, such as RFC 1945 and the
`
`dictionaries, support VirnetX’s proposed construction of “client,” and not that
`
`advanced by Petitioners. (See Opp. 4-5 & n.3.) Petitioners’ arguments to the
`
`contrary are unavailing. Singing a familiar tune, Petitioners accuse VirnetX of
`
`“ignor[ing] RFC 1945’s explanation that a proxy ‘acts as both a server and a client
`
`for the purpose of making requests on behalf of other clients.’” (Reply at 8 (quoting
`8
`
`
`
`

`

`
`Ex. 1014 at 5-6) (emphasis in original).) But the fact that a proxy can “act[] as both
`
`Case No. IPR2015-01047
`
`a server and a client” does not mean that the proxy is the server or the client. A
`
`proxy can perform the functions associated with a client, but that does not mean it is
`
`a client within the context of particular claim terms.
`
`Moreover, Petitioners themselves ignore RFC 1945’s definition of a “user
`
`agent” as “[t]he client which initiates a request.” (Ex. 1014 at 5 (emphasis added).)
`
`As VirnetX explained, that definition supports VirnetX’s—not Petitioners’—
`
`construction of “client.” And, as VirnetX also explained, Exhibits 1037 and 1043
`
`(which Petitioners again invoke, see Reply at 7-8) discuss “client” in a similar
`
`manner. (Opp. 5 n.2.) Petitioners characteristically offer not response to that
`
`argument.
`
`Finally, Petitioners offer no effective counter to VirnetX’s argument (Opp. 9)
`
`that its proposed construction is supported by the acknowledgement of Petitioner
`
`Apple’s own expert in parallel district-court litigation that Kiuchi’s user agent—i.e.,
`
`the web browser operated by the user, not the client-side proxy—“would be read
`
`naturally as a client computer.” (Ex. 2041 at 50:7-24.) Petitioners’ attempts to side-
`
`step the admission of their own expert (see Reply at 24-25) is unavailing. See infra
`
`at Section V.
`
`
`
`9
`
`

`

`
`
`Case No. IPR2015-01047
`
`“Between [A] and [B]”
`B.
`As VirnetX previously demonstrated, the phrase “between [A] and [B]”
`
`should be construed as “extending from [A] to [B].” (Opp. 9-10.) In reply,
`
`Petitioners do not present a single substantive reason why VirnetX’s construction
`
`is incorrect. (Reply at 9-10.) Petitioners assert that this is a case where the
`
`Phillips construction should be narrower than the BRI construction. (Id.)
`
`Petitioners do not try to justify this based on anything in the intrinsic or extrinsic
`
`record. Instead, Petitioners assert that “VirnetX urged the district court to reject
`
`this narrower construction” and argue that “VirnetX would not have proposed an
`
`unreasonable construction of ‘between [A] and [B]’ to a United States district
`
`court.” (Reply at 9-10.) This is an argument without any legal basis. VirnetX
`
`presented an argument in district court about the meaning of “extending from [A]
`
`to [B],” but the district court disagreed with VirnetX. (Ex. 2029 at 26.) VirnetX
`
`believes the district court’s reasoning was sound. Petitioners can hardly disagree
`
`given that Petitioner Apple has argued that the claims and specification support
`
`construing “between [A] and [B]” as “extending from [A] to [B].” (Ex. 2031 at
`
`25-26.)4
`
`
`4 Petitioners’ assertions about claim 13 (Reply at 10) are incorrect for the reasons
`
`discussed below.
`
`
`
`10
`
`

`

`
`
`Case No. IPR2015-01047
`
` Kiuchi Does Not Anticipate Claims 1, 2, 6-8, and 12-14
`Petitioners’ defense of their anticipation argument begins with the inapposite
`
`assertion that the Board is not foreclosed from considering anticipation on remand.
`
`(Reply at 11-13.) VirnetX never argued that the Board is foreclosed from
`
`considering anticipation. Rather, VirnetX explained that there was little reason for
`
`anticipation to be considered again since Petitioners’ mapping suffered from the
`
`same problems as those identified in the anticipation mapping rejected by the Federal
`
`Circuit (which is likely why the Federal Circuit only expressly instructed the Board
`
`to consider obviousness). (Opp. 12-13.)5 And on the merits, Petitioners have hardly
`
`any credible response at all.
`
`First, Petitioners assert that the client-side proxy “forward[s] the DNS request
`
`to a DNS function that returns an IP address of a nonsecure computer,” as recited in
`
`all of the independent claims, because the client-side proxy “performs a DNS
`
`lookup” in response to receiving what Petitioners’ purport is a DNS request. (Reply
`
`at 14.) In making this argument, Petitioners miss a crucial detail. The claims do not
`
`simply require accessing a DNS function in response to receiving a DNS request.
`
`
`5 Petitioners accuse VirnetX of “incorrectly quot[ing] the Federal Circuit.” (Reply
`
`at 13.) This accusation is puzzling; the passage Petitioners quote was never
`
`represented as a quotation from the Federal Circuit’s opinion.
`
`
`
`11
`
`

`

`
`The claims require forwarding the received DNS request to a DNS function. And
`
`Case No. IPR2015-01047
`
`on that point, Kiuchi is silent. Kiuchi simply discloses that “[i]f the client-side proxy
`
`receives an error status, then it performs DNS lookup, behaving like an ordinary
`
`HTTP/1.0 proxy.” (Ex. 1002 at 8.) In order to perform this DNS lookup, the client-
`
`side proxy necessarily generates a new DNS request. This is because the request
`
`sent to the client-side proxy is not disclosed as comporting with the DNS format. To
`
`the contrary, Kiuchi emphasizes that its system is an alternative to use of the DNS
`
`format. (See, e.g., Ex. 1002 at 7 (Kiuchi explains that the C-HTTP name service is
`
`used “instead of DNS,” the “DNS name service is not used for hostname resolution,”
`
`and a “DNS lookup” is only performed after a permission request to the C-HTTP
`
`name server fails); id. at 11 (explaining a different naming scheme is used in a C-
`
`HTTP based system); Ex. 2038 at ¶¶ 41-42.)6 Even if the Board finds this is not
`
`necessarily so, Kiuchi is at best silent, and thus cannot disclose forwarding any
`
`received DNS request to a DNS function.
`
`Second, Petitioners argue that Kiuchi’s client-side proxy “determin[es]
`
`
`6 To be sure, this is so even if the request sent to the client-side proxy satisfies the
`
`claim construction for the claimed “DNS request”—the point here is that the request
`
`sent to the client-side proxy follows a different protocol than what would be used to
`
`perform the “DNS lookup” in Kiuchi. (See also Opp. 13-14.)
`
`
`
`12
`
`

`

`
`whether [the/a] [intercepted DNS request / DNS request sent by a client] corresponds
`
`Case No. IPR2015-01047
`
`to a secure server,” as recited in all of the independent claims, because it asks the C-
`
`HTTP name server to perform this determination. (Reply at 15-16.) Petitioners do
`
`not dispute that the Board previously agreed that only the C-HTTP name server
`
`performs an operation that can map to the claimed “determining.” (Final Written
`
`Decision, Paper 80 at 8-9.) They nonetheless try to analogize the client-side proxy
`
`to the ’151 patent’s discussion of DNS proxy server 2610, which Petitioners say can
`
`make a determination “by communicating with gatekeeper 2603.” (Reply at 15-16
`
`(citing Ex. 1001 at 38:35-50).) Critically, however, the determination cited by
`
`Petitioners is not one corresponding to the claimed determination by the DNS proxy
`
`module. Specifically, the passage cited by Petitioners refers to a determination of
`
`whether a user has sufficient authorization (step 2704) that can be made by the DNS
`
`proxy server 2610 or by the gatekeeper 2603. (Ex. 1001 at 38:44-50.) On the other
`
`hand, the determination of whether the intercepted request corresponds to a secure
`
`host (step 2702) is disclosed as being performed by the DNS proxy server. (Ex.
`
`1001 at 38:35-43.) Moreover, Petitioners overlook that the client-side proxy is not
`
`even making a determination “by communicating with” the C-HTTP name server.
`
`It just acts in response to a determination by the C-HTTP name server. (See, e.g.,
`
`Ex. 1002 at 8.) Petitioners’ reliance on an assortment of Federal Circuit decisions
`
`suffers from the same problem—in each of those cases, while there was some
`
`
`
`13
`
`

`

`
`reliance on another component to acquire data relevant to a determination, the
`
`Case No. IPR2015-01047
`
`component that had to make the claimed determination actually made a
`
`determination. (See also Opp. 14-15.)
`
`Third, Petitioners’ do not dispute that claims 1 and 7 are not anticipated under
`
`VirnetX’s construction for “between [A] and [B].” (Reply at 17-18.) Petitioners
`
`argue that claim 13 is different because it can be satisfied with “a channel in which
`
`data is ‘encrypted on the insecure paths’ and is otherwise physically secure.” (Reply
`
`at 17.) According to Petitioners, “[t]hat is precisely what Kiuchi shows: encrypted
`
`data is sent over its C-HTTP connection between the proxies (Ex. 1002, 64), while
`
`data on that channel sent within each ‘member [institution] is protected by its own
`
`firewall.’” (Reply at 18.) Petitioners cite VirnetX Inc. v. Cisco Systems Inc.,
`
`767 F.3d 1308, 1321 (Fed. Cir. 2014), for the proposition that a “network is secure
`
`because it ‘has what’s called a firewall between its network and the public network
`
`[that] keeps the bad guys out.’” (Reply at 18.) Petitioners’ citation is misleading
`
`and incomplete. The quote, which comes from VirnetX’s expert in district court, in
`
`full provides that “[t]hat network is secure, because it’s been physically secured;
`
`and it also has what’s called a firewall between its network and the public network.
`
`So it keeps the bad guys out.” VirnetX, 767 F.3d at 1321. In other words, physical
`
`security is something that comes separate from the presence of a firewall. And
`
`Petitioners make no effort to show that Kiuchi discloses physical security. (See also
`
`
`
`14
`
`

`

`
`Opp. 15-16.) Indeed, Kiuchi is silent on the existence of any physical security. As
`
`Case No. IPR2015-01047
`
`such, Kiuchi cannot anticipate not only claims 1 and 7, but also claim 13, under
`
`VirnetX’s construction for “between [A] and [B].”
`
` Kiuchi Combined with Rescorla Does Not Render Obvious Claims 1, 2,
`6-8, and 12-14
`Petitioners newly argue that “any claims requiring end-to-end encryption are
`
`unpatentable, given that Rescorla would have made it obvious to modify Kiuchi’s
`
`system to provide end-to-end encryption.” (Reply at 18.) Petitioners did not present
`
`this argument in their original remand brief, and their attempt to add it now is
`
`gamesmanship. (See generally Paper 104.) In any event, this obviousness
`
`combination fails for several reasons.
`
`First, Rescorla does not qualify as a printed publication, and thus cannot be
`
`used in an obviousness combination. (Paper 48 at 40-45.) While the Board has
`
`previously disagreed (Paper 80 at 24-26)7, recent authority from the Federal Circuit
`
`has reaffirmed that VirnetX is correct that Rescorla—which is a mere “Internet-
`
`
`7 VirnetX did not challenge that finding on appeal because, as the Federal Circuit
`
`observed, “[t]he Board did not consider the disclosures of [this] reference[]
`
`because it did not identify any deficiencies of Kiuchi in its anticipation challenge.”
`
`VirnetX Inc. v. Mangrove Partners Master Fund, Ltd., 778 F. App’x 897, 908 (Fed.
`
`Cir. 2019).
`
`
`
`15
`
`

`

`
`Draft”—is not a printed publication. Specifically, the Federal Circuit recently held
`
`Case No. IPR2015-01047
`
`that “a work is not publicly accessible if the only people who know how to find it
`
`are the ones who created it.” Samsung Elecs. Co. v. Infobridge Pte. Ltd., 929 F.3d
`
`1363, 1372 (Fed. Cir. 2019). Petitioners introduced no evidence that those outside
`
`of the RFC development process would have known how to find Rescorla. (See
`
`Paper 48 at 40-45.) To the contrary, RFC 2026, which the Board previously relied
`
`on as “reflect[ing] ‘generally accepted practices’ for RFC documents and Internet-
`
`drafts” (Paper 80 at 25) states that Internet-Drafts are placed in a directory “for
`
`informal review and comment” for eventual publication as an RFC (Ex. 1010 at 8).
`
`In other words, Internet-Drafts like Rescorla are only meant to be known to those
`
`that are developing a particular draft into an RFC.
`
`Second, Petitioners do not even allege that Rescorla addresses the first two
`
`problems with Kiuchi identified by VirnetX. (See Opp. 13-15.) And Petitioners are
`
`simply wrong that Rescorla would have made it obvious to provide for “end-to-end
`
`encryption” in the context of Kiuchi. Kiuchi explains that its “approach is aimed at
`
`assuring proxy-proxy security” and is “fundamentally different” from “[o]ther
`
`secure protocols [that] are designed to be implemented in origin servers and user
`
`agents in order to assure ‘end-to-end’ security protection.” (Ex. 1002 at 10-11.)
`
`Kiuchi discourages end-to-end encryption, from a client to a target device, so that
`
`“no end-user has any chance to obtain keys for encryption or decryption” and
`
`
`
`16
`
`

`

`
`compromise security. (Id. at 11.) In this context, one of ordinary skill in the art
`
`Case No. IPR2015-01047
`
`would not have gone to Rescorla to modify Kiuchi to include precisely what Kiuchi
`
`says not to implement.
`
` Kiuchi Combined with RFC 1034, With or Without Rescorla, Does Not
`Render Obvious Claims 1, 2, 6-8, and 12-14
`As VirnetX previously explained (Opp. 16-25), Petitioners’ obviousness
`
`ground suffers from many defects. Petitioners’ arguments in reply are unavailing.
`
`First, Petitioners continue to urge the Board to defy the Federal Circuit by
`
`maintaining that, under VirnetX’s construction of “client,” “the Board correctly
`
`found Kiuchi’s client-side proxy was a ‘user’s computer’ because it is associated
`
`with the user agent.” (Reply at 23.) As the Federal Circuit recognized, “VirnetX’s
`
`proposed construction is a user’s computer, not any device that is associated with a
`
`user, however indirectly.” VirnetX, 778 F. App’x at 908. Petitioners are simply
`
`wrong in asserting that the Federal Circuit has not already rejected the notion that
`
`mere association with a user is enough under VirnetX’s construction.
`
`Petitioners further argue that Kiuchi’s client-side proxy is a “client” under
`
`VirnetX’s construction because Kiuchi’s client-side proxy has a “firewall
`
`administrator.” (Reply at 23-24.) At the outset, Petitioners concede that this is a
`
`new argument on remand, never presented before VirnetX’s appeal. (Reply at 24
`
`n.8.) Petitioners argue that the Federal Circuit did not prevent them from raising
`
`new arguments, but cite nothing in the Federal Circuit’s decision that includes any
`17
`
`
`
`

`

`
`invitation to present new arguments on the issue. VirnetX’s construction is not a
`
`Case No. IPR2015-01047
`
`new one—it was proposed before remand, in VirnetX’s Patent Owner’s Response,
`
`and Petitioners already had an opportunity to respond long ago when the evidentiary
`
`record was still open. It would be manifestly prejudicial to allow Petitioners to
`
`introduce new arguments at a time when the parties have agreed that it is improper
`
`to “introduce new evidence concerning the remanded merits issues.” (See Paper 87
`
`at 3.) If the Board nonetheless considers Petitioners’ arguments, it should reopen
`
`the record and provide VirnetX an opportunity to submit evidence on the issue (e.g.,
`
`expert testimony) so that VirnetX has a fair opportunity to respond to Petitioners’
`
`belated argument.
`
`Even with the record as it currently stands, it is clear that Petitioners’ position
`
`must fail. First, Kiuchi specifically identifies proxies as “servers,” not clients
`
`(Ex. 1002 at 8, 11, 12), and specifically differentiates a “proxy server” from “a user
`
`agent” (Ex. 1002 at 8). The client-side proxy is “specified as a proxy server for
`
`external (outside the firewall) access in each user agent within the firewall.” (Id.) It
`
`is just a manipulation of terms to characterize the