Trials@uspto.gov
`571-272-7822
`
`
` Paper 41
`
`Date: September 28, 2022
`
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`__________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`__________
`
`CLOUDFLARE, INC., and SPLUNK, INC.,
`Petitioners,
`v.
`SABLE NETWORKS, INC.,
`Patent Owner.
`__________
`
`IPR2021-00909
`Patent 8,243,593 B2
`__________
`
`Record of Oral Hearing
`Held: September 7, 2022
`__________
`
`
`Before STACEY G. WHITE, GARTH D. BAER, and
`JULIET MITCHELL DIRBA, Administrative Patent Judges.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`

`

`IPR2021-00909
`Patent 8,243,593 B2
`
`APPEARANCES:
`
`ON BEHALF OF THE PETITIONER:
`
`
`DANIEL CALLAWAY, ESQ.
`JAMES L. DAY, ESQ.
`of: Farella Braun & Martel, LLP
`Russ Building
`235 Montgomery Street
`17th Floor
`San Francisco, California 94104
`(415) 954-4414
`dcallaway@fbm.com
`jday@fbm.com
`
`MEHRAN ARJOMAND, ESQ.
`ALEX S. YAP, ESQ.
`of: Morrison & Foerster, LLP
`707 Wilshire Boulevard
`Los Angeles, California, 90017-3543
`(213) 892-5200
`ayap@mofo.com
`marjomand@mofo.com
`
`
`ON BEHALF OF THE PATENT OWNER:
`
`
`DANIEL LOWENSTEIN, ESQ.
`KENNETH J. WEATHERWAX, ESQ.
`of: Lowenstein & Weatherwax, LLP
`1016 Pico Boulevard
`Santa Monica, California 90405
`(310) 307-4500
`dlowenstein@lowensteinweatherwax.com
`kweatherwax@lowensteinweatherwax.com
`
`
`
`
`The above-entitled matter came on for hearing Wednesday,
`September 7, 2022, commencing at 10:00 a.m. EDT, via Videoconference.
`
`
`
`2
`
`

`

`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`IPR2021-00909
`Patent 8,243,593 B2
`
`
`
`
`
`P-R-O-C-E-E-D-I-N-G-S
`
`
`
`10:00 a.m.
`JUDGE DIRBA: Good morning. This is the oral hearing for
`IPR2021-00909, relating to U.S. Patent 8,243,593. My name is Judge Dirba
`and I'm joined today by my colleagues, Judges White and Baer. We'll begin
`with appearances. Petitioner.
`MR. DAY: This is James Day on behalf of Petitioner, Cloudflare.
`And I'd like to mention my colleague, Dan Callaway is here with me, also on
`behalf of Cloudflare. And I understand the Counsel for Splunk, Petitioner
`Splunk, is on the public line. And that is Alex Yap and Mehran Arjomand,
`Morrison Foerster.
`JUDGE DIRBA: And Mr. Day, who will be presenting on
`Petitioner's behalf, today? Is that just you, or will that be your colleague,
`Mr. Callaway as well?
`MR. DAY: Just me, Your Honor.
`JUDGE DIRBA: Okay. And Patent Owner.
`MR. WEATHERWAX: Can you hear me, Your Honor?
`JUDGE DIRBA: I can, not very well.
`MR. WEATHERWAX: All right. Just let me know if I'm too soft,
`because I am hard of hearing. And so, I do my best to modulate my own
`volume. But if I'm not, please remind me. This is Kenneth Weatherwax, for
`the Patent Owner. Patent Owner is also appeared at this meeting, by Daniel
`Epstein and Erin McCracken.
`(Simultaneous speaking.)
`JUDGE DIRBA: And Mr. Weather -- I apologize.
`MR. WEATHERWAX: I will --
`
`
`
`
`3
`
`

`

`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`IPR2021-00909
`Patent 8,243,593 B2
`
`
`
`
`
`
`JUDGE DIRBA: Go ahead, Mr. Weatherwax. I'll let you finish
`your appearances.
`MR. WEATHERWAX: That was all my appearances, Your Honor.
`JUDGE DIRBA: Excellent, and Mr. Weatherwax, will anyone other
`than you be presenting on Patent Owner's behalf today?
`MR. WEATHERWAX: No, Your Honor.
`JUDGE DIRBA: All right. I'll begin with a few general reminders
`for the hearing today. The hearing order provided that each party will have
`45 minutes of total argument time. Petitioner will open the hearing by
`presenting its arguments regarding the alleged unpatentability of the
`challenged claims. Petitioner may reserve rebuttal time, but no more than
`half of its total argument time.
`Thereafter, Patent Owner will response to Petitioner's arguments.
`Patent Owner may reserve surrebuttal time of no more than half of its total
`argument time to respond to Petitioner's rebuttal. Each side's rebuttal must
`respond to the arguments presented by the other side, but otherwise, each
`side may use its allotted time to discuss the case as it chooses.
`No new arguments may be presented at this hearing that have not
`been argued in the briefs. Counsel should hold any objections regarding the
`other side's arguments until it is their turn to speak. Also, I'll note that we've
`received Petitioner's objection to Slide 12 of Patent Owner's demonstrative.
`We will not rule on that objection at this time. But if Patent Owner presents
`that slide, we will ask Counsel to specifically identify where in the briefs the
`objected to argument was made.
`A few reminders for video. Please mute your microphone when
`you're not presenting. We may have an unexpected interruption in the
`
`
`
`4
`
`

`

`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`
`
`
`
`IPR2021-00909
`Patent 8,243,593 B2
`
`connection. If that happens to one of the video participants, or the Court
`Reporter, we will pause the argument and resume once the connection can
`be reestablished.
`If it is your connection that is lost, please spend up to five or so
`minutes attempting to reestablish the connection. If attempts are
`unsuccessful, at that point, please rejoin the hearing by audio, or reach out to
`one of the hearing staff.
`Are there any questions before we begin? Petitioner?
`MR. DAY: No, Your Honor.
`JUDGE DIRBA: Patent Owner?
`MR. WEATHERWAX: No, Your Honor.
`JUDGE DIRBA: Very well. We'll begin with Petitioner's
`presentation. Mr. Day, how long would you like to reserve for your
`rebuttal?
`MR. DAY: I'd like to save 10 minutes for rebuttal, Your Honor.
`JUDGE DIRBA: And you may begin when you're ready.
`MR. DAY: Well, good morning, Your Honors. We are here today
`to discuss the 593 Patent. I've provided a set of slides. And I'd refer you to
`Slide 2 to begin. And you'll see that this 593 Patent is entitled, Mechanism
`for Identifying and Penalizing Misbehaving Flows in a Network.
`And what the patent is talking about, is you have these flows of data
`packets going through a network. And you want to identify the type of
`traffic, so you can manage it. And in some cases, you can drop some of the
`packets. You can discard the flow entirely. And what the patent talks about
`is, at the time, there were existing techniques that looked for signatures. In
`
`
`
`5
`
`

`

`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`
`
`
`
`IPR2021-00909
`Patent 8,243,593 B2
`
`different techniques, but including looking for signatures in the data, in the
`packets, or in the flow, and using that to identify the type of traffic.
`But there were problems with that approach. And so, the 593 Patent
`says, what you should do to address those problems, is instead collect
`behavioral statistics. Look at the behavior of the flow. Look at how it acts.
`Look at what it does, and use that information to determine what type of
`traffic it is.
`Well, we filed our IPR petition. I'm now going to look at Slide
`Number 3. Filed an IPR petition, it was based on a primary reference to the
`patent, called Yung. And Yung discloses and describes a bandwidth
`management device that similarly analyzes flows of data packets through a
`network. Identifies the type of traffic and then, can applies things that Yung
`called, polices, to the flow, including a discard policy that would allow you
`to drop some or all of the packets in a flow.
`And like the 593, Yung talks about how there were existing
`techniques, including signature-based techniques. And that there were
`problems with those, they could, the signature changed. The signature could
`be encrypted. And so, it's better to look at what Yung's called, behavior
`patterns of the flow. And use those behavior patterns, what the flow does,
`how it acts, use that to identify the type of traffic.
`It's -- the review was instituted, and the Patent Owner voluntarily
`disclaims all the claims that we had challenged under Yung alone, plus some
`others. And what we're left with are the grounds on Slide 3.
`Ground 2 addresses Independent Claim 9, and Independent Claim
`29, and their dependents. And then Ground 3, is limited just to Independent
`
`
`
`6
`
`

`

`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`
`
`
`
`IPR2021-00909
`Patent 8,243,593 B2
`
`Claim 3. Of course, there is a dispute about a few of these dependent claims
`that I show in Ground 2, and I will get to that dispute.
`But first, and we can sort of skip over Slide 4, which just presents
`the topics I want to talk about. I want to start by talking about the Ground 2,
`and the combination of Yung and Copeland. And how Copeland discloses
`the badness factor recited in Claim 9 and 29.
`And Slides 5, 6, and 7, I'm going to move through very quickly,
`because the point of those is just that Yung alone, discloses the preamble of
`Claim 9 and the step of maintaining a set of behavioral statistics for the flow.
`That's undisputed, so we can move quickly to Slide 8, which focuses on the
`last step of Claim 9. Computing, based at least partially upon that set of
`behavioral statistics, a badness factor for the flow.
`And we identified, in Copeland, in the secondary reference,
`something that Copeland called a concern index, as disclosing what's recited
`here in Claim 9, in the last, in the computing step. And if you look on Slide
`8, we pulled out a few of the quotes from Copeland.
`On the upper right, it says, Copeland says, statistics are collected for
`each determined flow. And you analyze the flow. If the flow is a value,
`referred to as a concern index, it's assigned to each flow that appears
`suspicious. What Copeland is talking about, is identifying network probes,
`and network attacks. And so, we identified that.
`In response, what the Patent Owner has said is, well, you don't have
`any evidence supporting that. You're just relying on the ipse dixit of your
`expert. So, if you look at Slide 9. It shows some of our expert's declaration.
`And I've underlined the sentence that Patent Owner points to, and says, that's
`
`
`
`7
`
`

`

`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`
`
`
`
`IPR2021-00909
`Patent 8,243,593 B2
`
`the only evidence you have, which of course ignores all of the other
`evidence that we have.
`So, our expert, Dr. Jeffay, talked about what the concern index is,
`immediately after the sentence that they focus in on. He compares the
`concern index of Copeland to the disclosures in the 593 Patent, and what is
`says about this badness factor. If you move onto Slide 10, he, our expert
`continues. And he works with the Claim language, and I've highlighted and
`underlined on the left here, on Slide 10, that the claim actually tells us
`something about this badness factor. And it says that the badness factor
`provides an indication of whether the flow is exhibiting undesirable
`behavior.
`And our expert said, that's exactly what Copeland's concern index
`value is doing. Copeland identifies suspicious activity with the concern
`index. In other words, it provides an indication of whether the flow is
`exhibiting undesirable behavior. Our expert explains, that network probes
`and network attacks are undesirable behavior. He didn't just discern it. He
`explained why that is so.
`On Slide 11, we just show a little more of his testimony, and I'll
`move quickly ahead to Slide 12. And the point really, of his testimony is,
`that on Slide 12 we've got some of the language where the 593 Patent
`describes a badness factor. It's a factor that provides an indication of
`undesirable behavior. And then some quotes from Copeland on the right,
`that describe the concern index that does the same thing. It indicates
`suspicious flows that may be undesirable behavior, network probes, and
`network attacks.
`
`
`
`8
`
`

`

`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`IPR2021-00909
`Patent 8,243,593 B2
`
`
`
`
`
`
`But I want to address here, Patent Owner's argument that Cloudflare,
`Petitioner, has taken the position in district court, that this term, badness
`factor, is indefinite. And we cited a number of cases in our reply brief,
`explaining that, that issue is not before the Board. That, you know, unless
`it's impossible to apply Copeland to the claim language, it just doesn't come
`up in this proceeding. It's not an issue in this proceeding.
`Theirs, was their response to that case law in the sur-reply, and really
`the argument, is essentially that our expert, Professor Jeffay, doesn't know
`what the term means.
`Well, first of all, Professor Jeffay did not offer any opinion in the
`district court. So, it's not as though he has said at one time, this term is
`indefinite, and now is not saying that. His testimony has been clear. And
`you'll see in Patent Owner's slides, they quote him as saying two things
`about the badness factor. One, it's not a term of art. It's not disputed. And
`two, that a person of ordinary skill reading the 593, would not understand
`the bounds. So, that the full scope, the outer bounds of the term. And the
`problem with this term is that at the outer bounds, there may be some
`ambiguity in what is bad, and what is not bad.
`But he also testified, that whatever, you know, there may be some
`ambiguity at the margins, but Copeland describes exactly what the 593
`Patent is talking about, when it describes both in the specification, and the
`claims, that a badness factor is an indication of whether the flow is
`exhibiting undesirable behavior.
`That's exactly what the concern index is doing. So, if there's a
`question at the margins, we're not talking about the margins, we're right,
`Copeland is dead center in the claim language.
`
`
`
`9
`
`

`

`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`IPR2021-00909
`Patent 8,243,593 B2
`
`
`
`
`
`
`Now, I want to ask you to look at Slide Number 13, which has
`Figure 1 from Copeland. And it's, there's a lot going on here, so I want to
`take just a minute to orient you. If you look, it's Figure 1 from Copeland,
`Slide 13. In the upper left there's a picture of a hacker. So, this a guy sitting
`at computer, doing bad things. And directly below, is a legitimate user.
`Both of them are sending flows across a network. Not surprisingly the
`legitimate user is sending legitimate normal packet flows. And the hacker is
`sending what Figure 1, what Copeland refers to as bad flows. And he used
`the concern index to identify those bad flows.
`So, if there's some, again, if there is some ambiguity about what
`someone might consider bad or not, Copeland is very clear, that the concern
`index is about, intended to identify the bad flows. It corresponds to the
`badness factor.
`When you look at Slide 14, Patent Owner tries to distinguish the
`concern index by saying, well -- and I quoted on the left side of Slide 14,
`some passages from the sur-reply. Patent Owner says, the concern index
`doesn't work because it's not, the calculation of a badness factor, it does not
`calculate the badness factor for each flow, as recited in the claim. So, the
`argument is, the claims say you have to calculate the badness factor for each
`flow, which is simply inconsistent with what the claims actually say.
`And here on Slide 14, we've highlighted that Claim 9 is, recites a
`method for processing a flow. And down in the computing stuff, you
`compute a badness factor for the flow. This method is fully performed, as
`soon as you operate on one flow. There's nothing in the claim that says you
`have to repeat it once. There's certainly nothing that says you have to repeat
`it for every flow you may come across.
`
`
`
`10
`
`

`

`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`IPR2021-00909
`Patent 8,243,593 B2
`
`
`
`
`
`
`Now, in the slides, in Patent Owner's slides, it invites while, that we
`objected to, you see what I think, and what we said in the new argument, and
`that is, they try to distinguish the concern index by saying that it's not stored,
`as to individual flows. So, we object to that, the new argument.
`But in any event, if you look at Slide 16, which again, shows us the
`language of Claim 9. Claim 9 doesn't require storing the concern -- or the
`badness factor. It doesn't require that. What is says is, in the first step
`you're maintaining a set of behavioral statistics. So, arguably, you store
`those. But then all you have to do is compute the badness factor. Copeland
`discloses computing a concern index. And that matches the claim language.
`And the next Slides, 17 and 18, simply make the point that Claims 9
`and 29 are highly paralleled. And so, the arguments regarding 9, also apply
`to 29. So, let's move ahead. We can go to Slide 20 and talk a little bit about
`the reasons for combining Yung and Copeland. On Slide 20, I've underlined
`some language that the Patent Owner focuses in on.
`Patent Owner says, the entire basis for your motivation to combine,
`comes from one quotation in Yung. Where we cite Yung, as saying -- and
`Yung provides a number of different exemplary patterns of behavior to look
`for. And then says, the application behavior pattern can incorporate other
`factors as well. And our expert, Dr. Jeffay said, yes, and a person that's
`skilled in the art, would be motivated to look at the concern index in Yung,
`and apply it as another factor. He did, he does say that.
`But the Patent Owner says, this is the entire basis, which again
`requires you to ignore, in this case, five pages of detailed analysis from Dr.
`Jeffay, explaining exactly how, and why a person of ordinary skill in the art,
`would take the concern index from Copeland, and apply it in Yung's
`
`
`
`11
`
`

`

`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`
`
`
`
`IPR2021-00909
`Patent 8,243,593 B2
`
`bandwidth management device. It would improve Yung. It would make life
`easier on network administrators. Dr. Jeffay also explains just how you
`could do that. How this would work.
`And I have a few slides here, 20 and 21, and 22, where we pull out
`some of that language, but I think we can move ahead to the next argument
`and look at Slide 23.
`So, one of the arguments that Patent Owner makes, and if -- I've
`quoted on Slide 23, on the left side, language from the sur-reply, where it's
`challenging our motivation to combine evidence. And it says, Copeland
`concern index value, is not a behavioral attribute of a flow that there is any
`preponderant evidence that the POSITA would have been motivated to
`combine with Yung. Copeland's concern index value is reflective of packet
`header information, not the behavior of a flow.
`Kind of a mouthful, but there are a couple of points to make here.
`Seems to be that Patent Owner is saying, Copeland only uses behavioral
`statistics from packet headers. And that, that somehow causes a problem
`with motivation to combine. Well, there's a few problems there.
`One, if you look back at the language of Claim 9 and 29, there's no
`mention anywhere of behavioral statistics excluding packet header data. It
`doesn't say, you know, maintain a set of behavioral statistics obtained, you
`know, from some source other than packet headers. It doesn't say that.
`Patent Owner hasn't suggested any sort of a claim construction that would
`require that. So, the premise, the implicit premise here, is not supported by
`the claim. That's problem one.
`Problem two, is this argument appears to presume that Copeland
`only uses packet header data as behavioral statistics, which is not correct.
`
`
`
`12
`
`

`

`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`
`
`
`
`IPR2021-00909
`Patent 8,243,593 B2
`
`And again, if we're looking as Slide 23, our expert, Professor Jeffay, said,
`Copeland computes the concern index based at least upon a set of behavioral
`statistics. He says how Copeland does that.
`And then as highlighted at the bottom, on the right-hand side of Slide
`23, one of the examples that he pulls from Copeland Figure 6. And it is,
`Copeland uses the number of packets in the flow as the concern index. And
`the packet header data, does not include a number that is the total number of
`packets in a flow. That is not packet header data. That's data that's
`aggregated as you analyze the flow and as you analyze the packet.
`And we, so it's, Copeland is not using packet header data in this
`instance. And if you look at Slide 24, we know that this data, the number of
`packets that Copeland is using for the concern index, qualifies as a
`behavioral statistic, because the 593 Patent tells us it does. So, on Slide 24,
`we have two excerpts from the 593 Patent. And in both cases, it says, the
`number of packets in the flow is a behavioral statistic. So, Copeland is not
`limited merely to packet header data for its behavioral statistics. That's
`problem two with this argument.
`Third, if the idea here, is that Copeland wants its packet header data,
`but Yung does not, and so you wouldn't, you know, want to combine those
`two. That fails both because Copeland is not limited just to packet header
`data. But also, because Yung does not, does use packet header data, as well
`as other behavioral data to analyze packets.
`And on Slide 25, I cited one example where Yung explicitly says,
`here's a pattern you can look for, and the information that will show you that
`pattern, is in the header of packets.
`
`
`
`13
`
`

`

`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`IPR2021-00909
`Patent 8,243,593 B2
`
`
`
`
`
`
`And if we move ahead to Slide 28, I think the main point here is,
`Yung and Copeland are both looking at the same types of data. And with
`regard to the packet count, the number of packets in a flow, and the total
`bytes, so the full size of the flow, both Yung and Copeland are identifying
`exactly the same thing. And we know those are behavioral statistics
`because, and if you look at the bottom of Slide 28, the 593 Patent tells us
`they are.
`So, wherever that data is coming from, the packet header or
`somewhere else, those are behavioral statistics. They're collected by Yung
`and Copeland. That this argument offers no reason why you would not
`combine. And I guess there's a final point, that I'll also note. There's no
`evidence behind any argument in response to our motivation to combine
`evidence. There's no contrary expletive or other evidence.
`So now, I want to move ahead. And if you look at Slide 29, I want
`to move onto this Four-Steps Whitepaper. And the question about whether it
`was publicly accessible? If you look at Slide 30, the point there is simply
`that Yung discloses most of Claim 3. So, we're now talking about Grounds
`3. It applies only to Claim 3. Yung discloses most of the claim, that's
`undisputed.
`If you go to Slide 31, this Four-Steps Whitepaper discloses the one
`limitation, the one other limitation, that's also undisputed. If you look at
`Slide 32, we put in evidence with our petition, from the Internet Archive. A
`declaration from the Archive, authenticating and showing that this document
`was on the Packeteer website no later than, as of March 17th, 2003. That's
`undisputed.
`
`
`
`14
`
`

`

`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`IPR2021-00909
`Patent 8,243,593 B2
`
`
`
`
`
`
`If we move to Slide 33, now, we put in a fair amount of additional
`evidence with our reply brief, and I'll get to that. But I want to start with just
`the evidence that was submitted with our petition. And what do we know
`about this Four-Steps Whitepaper, and whether it was publicly accessible?
`Well, on Slide 33, we know that Yung, our primary reference, is a
`Packeteer patent, and it explicitly identifies Packeteer and its PacketShaper
`product. Now, the Whitepaper is a technical product overview of
`Packeteer's PacketShaper product. And Yung not only identifies it, but says
`this is exactly the type of device you can use to implement this invention I'm
`describing to you.
`What else do we know? On Slide 34, our expert explained that
`Packeteer was known in the industry. That was based on his personal
`experience in the 2000s. He at one time bought and used a PacketShaper in
`his lab. He also cited a few documents to support the fact that Packeteer was
`known, and that the PacketShaper product was on the market.
`Now, if we move ahead to Slide 35, I want to talk about what
`looking at the Four-Steps Whitepaper itself, tells us. And the point is, it tells
`us it's a marketing document. This was a marketing document put on the
`Packeteer website in order to promote the PacketShaper product. How do
`we know that?
`Well, if we read on Slide 35 it says, you know, if you have, or do
`any of these problems sound familiar to you? And then it lists off a few
`problems that network administrators might have. Says, if you do,
`PacketShaper, you know, can solve these problems. And the paper has
`underlined some language at the bottom, explicitly says, this paper is
`
`
`
`15
`
`

`

`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`
`
`
`
`IPR2021-00909
`Patent 8,243,593 B2
`
`intended for system and network administrators. It's intended for customers
`who might want to buy this product.
`If you look at Slide 36, there's more marketing language. Ease of
`deployment, you can deploy PacketShaper all at once, or you can do it in
`phases. There are a variety of models to choose from. If there was any
`question at all, on the last page it says, for more information, you know, call
`us. Or if you want to buy this product, call us or go to our website. It's a
`marketing piece that was on the website at least, well, well over a year
`before the priority date.
`Okay. If you look at Slide 38, now, what happened is, after the
`Institution Decision, the Patent Owner objected to the Four-Steps
`Whitepaper, saying it was not authenticated. And that it was, had not been
`shown to be prior art. So, we sent them a bunch of supplemental evidence,
`authenticating and showing that in fact, it was publicly accessible.
`One of the things we did, is we asked our expert, Dr. Jeffay, to use
`the Internet Archives' Wayback machine, and go to an archive page from
`early 2003 of the Packeteer website. Go to the Home Page, and try to
`navigate through the Home Page, to find this Whitepaper. And we've cited a
`little, some snippets from his supplemental declaration here on Slide 38. He
`says, that's what I did.
`On Slide 39, he walks through, first you click the link to the
`Product's Page. On Slide 40, he says, on the Product's Page there's a link
`that takes you to the PacketShaper product. Slide 41, he says on the
`PacketShaper Product Page, under the heading, Whitepapers, the very first
`thing you see is a link to the Four-Steps Whitepaper. And when you click
`
`
`
`16
`
`

`

`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`
`
`
`
`IPR2021-00909
`Patent 8,243,593 B2
`
`that link, Slide 42, confirms that it is the same document we submitted as
`Exhibit 1006. All right.
`JUDGE DIRBA: Counsel, I have a tangential question about the
`Whitepaper. You mentioned that this evidence was provided to Patent
`Owner shortly after the Institution Decision, after they objected to the
`evidence, the Whitepaper itself, and before Patent Owner's response. Is
`there any evidence in the record about when that information was provided
`to Patent Owner?
`MR. DAY: Yes, we included the service email evidencing when that
`was sent. I will have to look up the exhibit number, but I recall we put the
`service email in with our reply brief.
`JUDGE DIRBA: I see, helpful. If you could give me that, during
`your reply presentation.
`MR. DAY: Will do. Thank you, Your Honor.
`Okay, if we move ahead to Slide 45. The other thing we did is with
`our, this supplemental evidence that we provided in December, months
`before the Patent Owner deposed Dr. Jeffay, and months before the Patent
`Owner's response, in addition to his supplemental declaration, we also gave
`them a number of other documents, including what's shown on Slide 45.
`Which is an IBM Patent based on an application that was filed back in 2003,
`that cites to the same Four-Steps Whitepaper.
`
`Again, in our reply brief, we gave them the supplemental evidence in
`December. We added it to the record in this proceeding with our reply brief.
`At that time, we explained and we cited cases, saying that, that was timely.
`There's been no response. There was no response in the sur-reply. There
`
`
`
`17
`
`

`

`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`
`
`
`
`IPR2021-00909
`Patent 8,243,593 B2
`
`was no mention of this evidence in Patent Owner's response. So, it had it,
`for months in advance.
`So, if you go to Slide 46, we added a couple of other things with our
`reply brief.
`JUDGE DIRBA: Counsel.
`JUDGE BAER: Mr. Day.
`MR. DAY: Yes.
`JUDGE BAER: Mr. Day, can you hear me?
`MR. DAY: Yes.
`JUDGE BAER: This is Judge Baer. A question about your
`statement that all of this was timely. Is this an evidentiary issue, or is this an
`issue that you needed to establish as part of your principal case?
`MR. DAY: Yes, I would -- well, the answer to that is the Federal
`Circuit has told us that when the Patent Owner challenges public
`acceptability, it is appropriate and timely for the Petitioner to add evidence
`of public accessibility with their reply. And I, so here I would refer you to
`Slide 52. First of all, we made, I think a wholesome presentation of this in
`our reply brief. And there's no response to that in the sur-reply.
`But Slide 52, cites two Federal Circuit cases that are the dead-on
`point. The Valve v. Ironburg case says, you know, Petitioner may provide
`evidence of public accessibility of a reference after the petition stage. And
`there it was put in with the reply. And the Federal Circuit was okay in that.
`Then cites to the VidStream case, where the exact same thing
`happened. A bunch of evidence comes in with the reply, and the Federal
`Circuit says, yes, that's timely. So, the Federal Circuit has told us, that this
`
`
`
`18
`
`

`

`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`
`
`
`
`IPR2021-00909
`Patent 8,243,593 B2
`
`is an appropriate way for us to bring in evidence, establishing public
`accessibility.
`JUDGE BAER: And are those cases, do those cases stand for the
`proposition that it is permissible that we allow it, or that it is mandatory that
`we allow it? In other words, is it error for us to hold otherwise, or is it just
`permissive? We could go either way?
`MR. DAY: I don't believe that the decisions draw a distinction.
`They, in both cases, they reject the argument that it's untimely. I suspect if
`the question was presented, it could be an error, because the Federal Circuit
`has rejected the argument, that the Board might be relying on. But that, I
`didn't read that distinction from these cases, Your Honor.
`JUDGE BAER: Thank you, that's helpful.
`MR. DAY: So, let me talk about just the last two pieces of evidence
`that came in with our reply. We do that by going back to Slide 46. One of
`the things that the Patent Owner did, is put in a declaration from Patent
`Owner's Counsel, saying, if you look at the Internet Archives, using the
`Wayback machine, looking for archived versions of the Four-Steps
`Whitepaper, after March 17th, 2003, she got some error messages. We think
`what happened is if you, that Packeteer changed to a different version of this
`document. So, that was leading to some error messages.
`What we did, in response, is instead of looking after March 17th,
`2003, we looked from the period beginning in September 2002, which is the
`date on the face of the document, through early 2003. And my colleague,
`Mr. Callaway, went to the -- he repeated the process that Dr. Jeffay had
`employed in his supplemental declaration. Start with an archive Home Page,
`from Packeteer's Home Page and use that to click to the Product Page, the
`
`
`
`19
`
`

`

`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`
`
`
`
`IPR2021-00909
`Patent 8,243,593 B2
`
`PacketShap