`571-272-7822
`
`
`
`
`
`
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`
`
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`______
`
`CISCO SYSTEMS, INC.,
`
`Petitioner,
`
`v.
`
`CENTRIPETAL NETWORKS, INC.,
`
`Patent Owner.
`
`___________
`
`Case IPR2018-01454
`Patent 9,674,148 B2
`___________
`
`Record of Oral Hearing
`Held: December 5, 2019
`
`____________
`
`
`
`
`Paper No. 31
`
`
`
`
`
`
`
`Before KEVIN F. TURNER, BRIAN J. MCNAMARA, and
`STACEY G. WHITE, Administrative Patent Judges.
`
`
`
`
`
`
`
`
`
`
`
`Case IPR2018-01454
`Patent 9,674,148 B2
`
`APPEARANCES:
`
`ON BEHALF OF THE PETITIONR:
`
`GREGORY P. HUH, ESQUIRE
`DAVID MCCOMBS, ESQUIRE
`Haynes and Boone, LLP
`2505 North Plano Road
`Suite 400
`Richardson, TX 75082
`
`ON BEHALF OF THE PATENT OWNER:
`
`JAMES HANNAH, ESQUIRE
`Kramer Levin Naftalis Frankel
`990 Marsh Road
`Menlo Park, CA 94025
`
`The above-entitled matter came on for hearing on Thursday, December
`5, 2019, commencing at 1:00 p.m., at the U.S. Patent and Trademark Office,
`600 Dulany Street, Alexandria, Virginia, before Donna Jenkins, Notary
`Public.
`
`
`
`2
`
`
`
`Case IPR2018-01454
`Patent 9,674,148 B2
`
`P R O C E E D I N G S
` - - - - -
`
`JUDGE MCNAMARA: Good afternoon everybody.
`This is the formal hearing in IPR 2018-01454. I'm Judge
`McNamara. Judge Turner and Judge White are participating
`remotely. Could we have counsel introduce themselves please.
`Let's start with the Petitioner.
`MR. MCCOMBS: Good afternoon, Your Honors. I'm
`David McCombs here for Petitioner Cisco Systems and with me
`are Theo Foster and Gregory Huh. Gregory Huh will be making
`our presentation today.
`JUDGE MCNAMARA: Thank you very much. Patent
`
`Owner.
`
`MR. HANNAH: Good afternoon, Judge McNamara.
`Nice to see you again Judge White. I'll try to look that way at
`the camera instead of the screen. Nice to meet you Judge Turner.
`James Hannah on behalf of Centripetal Networks, and with me is
`Jeffrey Price.
`JUDGE MCNAMARA: Would you be doing the
`
`arguing?
`
`MR. HANNAH: Yes, I will be doing the argument
`today. Thanks.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`
`3
`
`
`
`Case IPR2018-01454
`Patent 9,674,148 B2
`
`
`JUDGE MCNAMARA: Okay. Thank you so much.
`
`
`Okay. Let me see. Each side has a 60 minutes maximum time.
`You don't have to use it all but you're certainly welcome to.
`We'll start with the Petitioner who'll present its case with regard
`to the challenged claims and any outstanding motions. Patent
`Owner will then present its opposition to Petitioner's case.
`Petitioner can use any time it reserved for rebuttal and then the
`Patent Owner will get a surrebuttal directed to issues that were
`raised in the Petitioner's rebuttal using whatever time he has
`reserved. I assume everyone is ready to proceed? All right. So
`let's begin with the Petitioner. Is there some amount of time
`you'd like me to alert you to?
`
`
`MR. HUH: Good afternoon, Your Honors. So we'll
`be reserving 15 minutes for rebuttal. Maybe a five minute
`warning would be great.
`
`
`JUDGE MCNAMARA: Okay. So why don't we let
`you know when you've used up 40 minutes.
`
`
`MR. HUH: Yes, please.
`
`
`JUDGE MCNAMARA: Okay. That sounds terrific.
`
`
`MR. HUH: Thank you.
`
`
`JUDGE MCNAMARA: All right. Please proceed.
`Oh, and you have handouts? Do you have hard copies of the
`demonstratives?
`
`
`MR. HUH: I do, Your Honor.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`
`
` 4
`
`
`
`Case IPR2018-01454
`Patent 9,674,148 B2
`
`
`JUDGE MCNAMARA: Could I have a set, please?
`
`
`MR. HUH: Certainly.
`
`
`JUDGE MCNAMARA: And does the court reporter
`
`
`have a set?
`
`
`MR. HUH: Yes.
`
`
`JUDGE MCNAMARA: Okay, great. Thanks very
`much.
`MR. HUH: Okay. So as I was saying, this is Gregory
`
`
`Huh on the record on behalf of Petitioner Cisco Systems. During
`today's discussion I will refer to the slides for everybody's
`benefit as well as the record. I just want to briefly note that the
`slides have not been filed per the Board's instructions in the
`email.
`So I'll begin by looking at slide 2 and if we look at
`
`
`slide 2 there are five issues for discussion here today, and the
`first one is the construction for the preprocessing limitation and
`I'll address that in detail as well as Patent Owner's and
`Petitioner's position on that construction.
`
`
`The other issues for discussion are the fact that the
`prior art teaches a preprocessing limitation under any
`construction, that the prior art teaches a caching limitation as
`well as the claim 3 limitation of dynamically adjusting and lastly
`I will address the secondary considerations. Very likely I'll
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`
`
`
`
`
` 5
`
`
`
`Case IPR2018-01454
`Patent 9,674,148 B2
`
`
`address that as part of my rebuttal given that it's Patent Owner's
`burden.
`Turning to slide 3 for context provided is claim 1 and
`
`
`for the first issue the preprocessing limitation is the first
`limitation in claim 1. If we look at claim 1 it recites
`preprocessing by a network device a first rule set and a second
`rule set. Now that limitation was not construed in the petition.
`Petitioner applied the plain and ordinary meaning and the
`Board's decision on Institution also did not construe that
`limitation. However, after Institution Patent Owner in Patent
`Owner's response proposed a construction and the petition's reply
`proposed an alternative construction. We also explained and
`showed that the prior art teaches this limitation under any
`construction.
`
`
`Now turning to slide 4. At the bottom of the slide we
`have patent Owner's response with the proposed construction for
`the preprocessing limitation and in Patent Owner's response they
`construe this term as, and I'll quote it in the record,
`
`
`"Performing operations on a first rule set and a second
`rule set prior to their respective rule set being applied to
`packets."
`We would urge that the Board not adopt this
`
`
`construction for a number of reasons including that it's
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`
`
`
`
`
` 6
`
`
`
`Case IPR2018-01454
`Patent 9,674,148 B2
`
`
`ambiguous as well as the fact that this construction limits
`preprocessing beyond the 148 specification.
`
`
`We turn to slide 5. Dr. Orso, who is Centripetal's
`expert witness, on deposition testified that well, an operation
`under its construction would be anything or nothing at all. So
`for instance, here we have a no op operation which according to
`Dr. Orso would nonetheless be an operation and no op can still
`be considered an operation, it wouldn't do anything. Given the
`fact that the operation under Centripetal's construction includes
`something that does nothing at all, the operation under its
`construction really has no meaning and therefore renders the
`construction ambiguous.
`
`
`I would further note, if we turn to slide 6, that the
`actual construction as argued by Patent Owner is not supported
`and is not limited as such in the 148 patent specification. If we
`look at Patent Owner's response according to Patent Owner on
`slide 6, Patent Owner asserts that this requires performing
`operations on the rule sets themselves. Notably, however,
`nowhere in the specification is there such a requirement.
`Absolutely nowhere in the 148 patent does it require performing
`operations on the rule sets themselves. If we look in the middle
`of slide 6 we have the 148 patent specification --
`
`
`JUDGE MCNAMARA: Let me interrupt you for just a
`quick second.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`
`
` 7
`
`
`
`Case IPR2018-01454
`Patent 9,674,148 B2
`
`
`MR. HUH: Sure.
`
`
`JUDGE MCNAMARA: When you said doesn’t
`
`
`require it, what does it say about performing operations on the
`rule sets themselves? I mean if it's not a requirement is it at
`least disclosed that you can perform those operations on the rule
`sets and that sort of thing?
`
`
`MR. HUH: So here we have on slide 6 in the middle
`the 148 patent which describes what the preprocessing would be.
`So we don't have to guess, we can look and see exactly what it
`says and I'll quote it in the record. This is the 148 patent at
`column 4, lines 38 to 42 and I'll quote,
`
`
`"In some embodiments preprocessing policies 130 and
`132 as rule sets may include merging two or more rules within
`the rule sets into one rule separating one or more rule sets within
`the rule sets into two or more rules or reordering one or more
`rule set within the rule sets."
`
`
`This is the extent of examples of preprocessing in the
`148 patent. What I would note here this is permissive language.
`Nowhere are these requirements in some embodiments may
`include this is permissive language, nowhere does it require
`doing any of these things. These are exemplary embodiments
`that may include these operations. I further note if you look at
`this --
`
`
`
`JUDGE MCNAMARA: Claim 1 --
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`
`
` 8
`
`
`
`Case IPR2018-01454
`Patent 9,674,148 B2
`
`
`MR. HUH: -- operation, go ahead Your Honor.
`
`
`JUDGE MCNAMARA: -- claim 1 though, it's not
`
`
`permissive right? Claim 1 says preprocessing, the very first
`thing is a method comprising and the first step is preprocessing
`by a network device a first rule set and a second rule set. So it's
`not permissive in the claim, it's a requirement of the claim,
`right?
`MR. HUH: The question is what does preprocessing
`
`
`include and what does it require.
`
`
`JUDGE MCNAMARA: Right.
`
`
`MR HUH: So these are examples, merging, separating
`and reordering are examples of preprocessing but they're not
`limiting examples. In fact if we look at the bottom of slide 6 Dr.
`Orso, Centripetal's expert, testifies during deposition that
`preprocessing includes other kinds of optimizations that would
`be considered preprocessing. Nowhere in the specification is it
`required that the rule sets themselves be modified or changed or
`preprocessed by some type of modification and that's exactly
`what Patent Owner is arguing and that's not supported.
`
`
`In fact, during the prosecution history of the parent
`patent in the 148, turning to slide 7, the examiner was headed on
`with the exact same question and the Office in that situation held
`that preprocessing doesn't have to modify the rule set. Here we
`have --
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`
`
` 9
`
`
`
`Case IPR2018-01454
`Patent 9,674,148 B2
`
`
`JUDGE MCNAMARA: What would --
`
`
`MR. HUH: -- patent, go ahead Your Honor.
`
`
`JUDGE MCNAMARA: -- what would preprocessing
`
`
`be if it doesn't modify the rule set? What would that consist of?
`
`
`MR. HUH: So here we have an example during the
`prosecution history of the parent which was the same limitation
`was at issue and the examiner considered a reference that taught
`associating a threshold with a rule set. The rule set itself was
`not modified but it was associated with a threshold. The action
`of associating the threshold with the rule set according to the
`Office was the preprocessing of the rule set and Patent Owner
`now is arguing that somehow that was distinguished but simply it
`wasn't. We see at the bottom of slide 7 in the Notice of
`Allowance the examiner maintains, and I'll quote in the record,
`
`
`"The prior art teaches rule preprocessing."
`
`
`So here we have the Office in the parent application
`giving the Patentee notice that preprocessing does not have to be
`done on the rule sets themselves and I would note this. In the
`148 patent examination once they had filed that child from the
`parent which is this from the parent on slide 7, Patent Owner was
`on notice of the examiner's position, okay? Had they chosen to
`limit preprocessing to perform operations on the rule sets
`themselves they could have included that language in the claim.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`
`
`
`
`
` 10
`
`
`
`Case IPR2018-01454
`Patent 9,674,148 B2
`
`
`They did not. That language is not included in the actual claim
`itself. Go ahead, Your Honors.
`
`
`JUDGE WHITE: Counsel, why wouldn't modifying
`the threshold that applies to the rule set be an operation
`performed on the rule set? It doesn't necessarily have to change
`the rules within the rule set but you are modifying how that rule
`set would be applied by performing that particular operation.
`
`
`MR. HUH: So to be clear, our position is that
`modifying the rule sets is an example of preprocessing. For
`example, if you modify the rule set by merging, separating or
`reordering, those are certainly examples of preprocessing. Those
`are non- limiting examples and even for example here associating
`a threshold with a rule set is as well preprocessing. But this
`association does not modify the rule set as the reorder or do
`anything operation on the rule set within it, on the rules within
`the rule set but there is a preprocessing of the rule set. So our
`position is that the examples in the 148 patent are examples of
`preprocessing but those examples are not limiting.
`
`
`JUDGE WHITE: Are you saying that preprocessing
`can be completely divorced from the rule set? So preprocessing
`includes other forms of processing that you do before you get
`into performing operations on the rule set themselves?
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`
`
`
`
`
` 11
`
`
`
`Case IPR2018-01454
`Patent 9,674,148 B2
`
`
`MR. HUH: Our position is that, for example if you do
`
`
`something with a rule set by associating something else with it,
`that is a preprocessing of the rule set.
`
`
`JUDGE MCNAMARA: And your position then is that
`because that's already been disclosed by the reference that was
`cited in the other case, it's in the prior art?
`
`
`MR .HUH: Our position is is that this issue on
`construction was already considered by the Office. But I will
`note this that regardless, and I will walk through this for the
`Board's benefit, but regardless -- and our briefings are clear on
`this point -- regardless of the construction the prior art teaches
`preprocessing clearly. But again --
`
`
`JUDGE TURNER: Counsel, l et me give you a
`hypothetical. So if you might say -- and I'm going to be
`speaking orally so I know you're not looking at this and writing
`so bear with me -- so if we adopted a construction of optimizing
`by a network device a first rule set and a second rule set that
`modifies those rule sets before the network device processes
`packets, would that be acceptable to Petitioner or do you think
`that the fact that if you have optimizing through modification
`that somehow that's too narrow?
`
`
`MR. HUH: Our position is that that would be too
`narrow based on the testimony of both experts because --
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`
`
`
`
`
` 12
`
`
`
`Case IPR2018-01454
`Patent 9,674,148 B2
`
`
`JUDGE TURNER: That would be, even though that
`
`
`really sort of comports with all the examples provided in the
`specification?
`
`
`MR. HUH: The examples are not limiting, Your
`Honor, Judge Turner.
`
`
`JUDGE TURNER: No, I understand that. I
`(indiscernible.) They have to mean something.
`
`
`MR. HUH: Certainly.
`
`
`JUDGE TURNER: If I divorce myself from the
`examples the nice people at the Federal Circuit are going to
`reverse them and --
`
`
`MR. HUH: we definitely don't want that.
`
`
`JUDGE TURNER: -- well, we are together I think in
`that review.
`
`
`MR. HUH: Certainly.
`
`
`JUDGE TURNER: But somehow if we're -- again my
`concern is if we're talking about untried nine (phonetic) if you
`want something everybody can see -- if you're just talking about
`optimizing, that can be sort of nebulous. That can mean I'm
`looking at the rules and I'm jotting down notes, I'm determining
`triggers, I'm doing something -- I'm not really doing anything
`with the rule sets which seems to be Patent Owner's concern in
`saying well, look at all my examples, all my examples are doing
`something. They're either reordering or they're merging these or
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`
`
` 13
`
`
`
`Case IPR2018-01454
`Patent 9,674,148 B2
`
`
`they're doing things to the set and that's what I'm really trying to
`get at here, and you say well, you should have claimed it better
`so this is an opportunity. If we adopted the claim construction
`that maybe makes that clear.
`
`
`MR. HUH: So --
`
`
`JUDGE TURNER: I think that --
`
`
`MR. HUH: Go ahead, I'm sorry.
`
`
`JUDGE TURNER: Go ahead.
`
`
`MR. HUH: So from our perspective, Your Honor, the
`Federal Circuit has made clear that it's not for the Board or for
`anyone else to fix claims that were not properly claimed. We're
`not here to salvage a claim when in fact it wasn't properly
`written. There's no such requirement in the claim that requires
`performing operations on the rule sets themselves and nowhere is
`that in the specification. It's also improper to incorporate and
`bind (phonetic) from the specification the claims themselves
`(indiscernible) forget about the Federal Circuit.
`
`
`So from our perspective when we look at slide 9,
`which Your Honor referenced, optimizing by a network device a
`first rule set, a second rule set, before the network device
`processes packets is a proper construction at least because both
`experts agree that there's two kinds of optimizations that occur.
`
`
`Specifically if we look at slide 10 Dr. Orso,
`Centripetal's own expert, testifies that the preprocessing allows
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`
`
` 14
`
`
`
`Case IPR2018-01454
`Patent 9,674,148 B2
`
`
`for fast swapping. Now if you simply modify a rule set, that's
`not going to provide fast swapping or if you reorder the rule set
`that's not going to (indiscernible) fast swapping, but both experts
`testify in the context of the 148 patent that preprocessing
`facilitates fast swapping. So if you have an optimization that
`facilitates fast rule swapping that wouldn't be a preprocessing in
`the context of the 148 patent. Both --
`
`
`JUDGE TURNER: (Indiscernible.)
`
`
`MR. HUH: Go ahead.
`
`
`JUDGE TURNER: Sure. And I assume that if that
`preprocessing is merging, you'd have to agree that that would
`allow for fast swapping?
`
`
`MR. HUH: That would allow for fast processing but
`not for fast swapping necessarily.
`
`
`JUDGE TURNER: Why?
`
`
`MR. HUH: Because if your --
`
`
`JUDGE TURNER: Like have fewer rules in a rule set
`it seems like it's easier to swap them out. I have less operations
`I would think but maybe I'm missing something.
`
`
`MR. HUH: Right. So the swapping in the context of
`the 148 patent is swapping the rule sets, not the rules in the rule
`sets. So if you have rule set one may have ten rules and then you
`merge them into five, that allows you to better and more
`efficiently apply the rule set to the process of the packets
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`
`
` 15
`
`
`
`Case IPR2018-01454
`Patent 9,674,148 B2
`
`
`because you have less rules to apply to packets and it gives you
`fast processing as far as it relates to packet processing but the
`swapping event by merging doesn't facilitate fast swapping.
`
`
`JUDGE TURNER: Okay. Maybe I'm missing
`something, I don't know. So if I have ten rules and then I'm
`preprocessing for five and then later on I'm swapping that out for
`something has three rules, it seems like there has to be fewer
`operations for doing that swapping because it's smaller. Maybe
`I'm missing something fundamentally here.
`
`
`MR. HUH: So the way that, from an operational
`perspective when you're swapping out the rules, you're just
`identifying a new rule set to be applied. You're really changing
`a pointer and pointing to a new rule set to be applied to packets.
`That's the effect of merging or separating when you're swapping
`or changing the pointer to a different rule set.
`
`
`JUDGE TURNER: Now is that true even if you're
`dynamically reallocating memory as sort of, and we haven't
`maybe gotten to those claims yet, but it seems like that's a
`different operation, isn't it?
`
`
`MR. HUH: I mean I think that's a little bit of
`different context. I don't want to muddy the record with talking
`about a limitation that's not part of -- that's part of claim 3.
`
`
`JUDGE TURNER: No, I understand it’s my job to
`muddy the record so.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`
`
` 16
`
`
`
`Case IPR2018-01454
`Patent 9,674,148 B2
`
`
`MR. HUH: I'm not trying to do that but I just want to
`
`
`be clear that that's a little different discussion that isn't tied to
`the construction issue.
`
`
`JUDGE TURNER: No, I understand that.
`Construction's on my mind because I'm going to get to it, so I'm
`just trying to understand. You're sort of saying well nothing you
`do is going to allow for a faster swapping I guess, is that what
`you're saying?
`
`
`MR. HUH: No. What we're saying is the context of
`the 148 patent is broad enough to facilitate and to permit
`preprocessing such that that preprocessing, if that allows for fast
`swapping, then that is preprocessing in the context of the 148
`patent.
`JUDGE TURNER: So again --
`
`
`MR. HUH: Again this is consistent with the
`
`
`examiner's position during prosecution where the examiner had
`the same reasoning.
`
`
`JUDGE WHITE: I think this is coming down to how
`this is coded. If you're coding this in a manner where you're
`swapping pointers and maybe this is not a big difference -- if
`you're coding in the manner where you're swapping out data
`structures, then there could be some difference . I think if you're
`really digging down to the detail level of how something's
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`
`
`
`
`
` 17
`
`
`
`Case IPR2018-01454
`Patent 9,674,148 B2
`
`
`specifically coded as to whether you want that faster and in some
`circumstances it would be another set, wouldn't it be?
`
`
`MR. HUH: I believe that's correct, Judge White, and
`this is the whole point is that the 148 patent never provides such
`nuanced detail as to how to code it, how to somehow optimize
`these differently, right? So the disclosure of the 148 patent is
`broad to cover both kind of circumstances, whether it is or it
`isn't faster swapping and this is exactly why both experts testify
`that the preprocessing can allow for fast swapping or will it
`allow for more efficient packet processing. You know, again --
`
`
`JUDGE WHITE: I certainly agree in certain
`circumstances it would allow for faster swapping depending on
`how it's coded of course.
`
`
`MR. HUH: Right. But my point is that going back to
`slide 6, the 148 patent has very limited disclosure as to what
`preprocessing may include. They don't describe somehow how to
`limit the preprocessing to anything, right? Some embodiments
`just may include merging, separating or reordering. I would note
`that nowhere did they tell you how to perform the merging step
`or reordering, right? That disclosure is nowhere present. It's
`allowing the POSITA leeway into implementing yet how the
`person of ordinary skill decides how to do so.
`
`
`But nonetheless, I would note that even if the Board
`decides that Petitioner's construction is not appropriate and the
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`
`
` 18
`
`
`
`Case IPR2018-01454
`Patent 9,674,148 B2
`
`
`Board adopts Patent Owner's construction which I believe it
`should not, the prior art nonetheless teaches preprocessing
`multiple ways.
`
`
`JUDGE TURNER: Just one more question before you
`move on.
`MR. HUH: Certainly.
`
`
`JUDGE TURNER: I understand that there seems to
`
`
`be, you know, you talked about the Patent Owner's expert saying
`that you can have no operation and that would be an operation.
`But couldn't the same thing be said about optimizing? Let's say
`that my rule sets are pristine, they're perfect, so that if I'm
`optimizing by the network device these rule sets it may actually
`do nothing. That's also a possibility even under Petitioner's
`construction, yes?
`
`
`MR. HUH: I don't believe there's expert testimony on
`that point but I will say --
`
`
`JUDGE TURNER: I don't know that there is either.
`I'm going to ask you and you can certainly not give me an
`answer.
`MR. HUH: No, it's not that. But what I will say is
`
`
`that the prior art -- under our construction, we identified what we
`believe is the optimization and there's no dispute that that
`optimization in fact does optimize the fast rule swapping and
`therefore would meet our construction. So I believe even beyond
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`
`
` 19
`
`
`
`Case IPR2018-01454
`Patent 9,674,148 B2
`
`
`what was necessary to apply the art, so the prior art meets the
`construction we propose and to the extent there's additional
`nuances neither expert pondered upon those.
`
`
`JUDGE TURNER: Okay. Thank you.
`
`
`JUDGE MCNAMARA: One other thing I wanted to
`ask you because I'm looking at the patent itself, and it talks
`about -- this is the 148 patent and this is in column 4 looking
`down around line 25, 26 -- and it talks about a rule set contained
`within a policy may be preprocessed and then says for example,
`recent advances in packet filtering technology have reduced the
`time required to apply large rule sets. So is that another example
`of preprocessing, some kind of filtering and in that context is
`filtering changing the rule set or is it not changing the rule set,
`just filtering out certain things? It's just above where they talk
`about large rule sets.
`
`
`MR. HUH: I have it, I see it, Your Honor. I don't
`believe that that's an example. Clearly the examples follow at
`line 38.
`JUDGE MCNAMARA: Right, that's clear. It says in
`
`
`some embodiments preprocessing may be, but this talks about
`advances in packet filtering that reduce the time and it's in the
`context of preprocessing and I'm just trying to determine if
`filtering, and I don't even know whether or not filtering would or
`would not be modifying the rule set, so that's another whole
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`
`
` 20
`
`
`
`Case IPR2018-01454
`Patent 9,674,148 B2
`
`
`question. Are they filtering packets, are they filtering the rule
`set, I don't now what they're filtering.
`
`
`MR. HUH: Right. I think that portion there is a little
`bit unclear and I don't want to speculate here as to what that
`means.
`JUDGE MCNAMARA: Sorry. Thanks very much. I
`
`
`just wanted to know if you had any thoughts on that.
`
`
`MR. HUH: Now if we turn to slide 11 I'd like to now
`address Centripetal's first argument which is that the prior art
`does teach preprocessing. That's simply incorrect. It teaches it
`multiple ways.
`
`
`Turning to slide 12, Your Honors, the petition
`explained that Roese teaches preprocessing in a plain and
`ordinary meaning and now this same teaching also teaches
`preprocessing on the Petitioner's construction submitted with the
`reply. Roese teaches associating mitigating policies as well as
`associated rapid response identifiers with the PDR sets. The
`PDR sets are rule sets in the context of Roese and it's clear that
`this association of the rapid response identifier with the rule set
`facilitates quick rule swapping and there's no dispute about that.
`In fact Patent Owner in their response recognizes this, toward the
`bottom of slide 12, that Roese's invention is intended to provide
`an immediate response to a proceed trigger condition in order to
`protect the network. That is, they respond immediately by
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`
`
` 21
`
`
`
`Case IPR2018-01454
`Patent 9,674,148 B2
`
`
`swapping the rule sets. So there's no dispute that in fact Roese
`does optimize the swapping of the rule sets by associating rapid
`response identifiers with the rule sets. What I would note is if
`the Board adopts Petitioner's construction then there's no dispute
`that the prior art does teach a preprocessing limitation.
`Moreover, the prior art teaches preprocessing multiple ways
`under Centripetal's construction. Turning to slide 13. We give
`the Board additional context for the operations in Centripetal's
`construction.
`
`
`JUDGE TURNER: Counsel, just before you move on.
`In Roese, would Petitioner acknowledge that no PER sets are
`modified? You're associating but you're not actually modifying
`a PER set, is that a correct statement on my part or would
`Petitioner quibble with that?
`
`
`MR. HUH: Well I think Roese would say otherwise.
`If you look to slide 14, our position initially is that the
`association is under Petitioner's construction of preprocessing
`but Roese also teaches changing, modifying, updating the rule
`set themselves. So I'll get to that but the answer is no. Roese
`also teaches changing the rule sets.
`
`
`JUDGE TURNER: Okay.
`
`
`MR. HUH: But just to give the Board context what
`operations are under Centripetal's construction, slide 13 here we
`have Patent Owner's response where they assert that an operation
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`
`
` 22
`
`
`
`Case IPR2018-01454
`Patent 9,674,148 B2
`
`
`on the rule sets is changes to the rule set. Dr. Orso said that if
`you modify the rule set you're making an operation, right? So if
`you change the rule set you're performing the operations under
`Centripetal's construction.
`
`
`Turning to slide 14. This is exactly what Roese
`teaches. At the top of slide 14 we have Roese's first (phonetic)
`portion where first I would note that they teach updating,
`revising, changing, adjusting at the rule sets. Those are
`operations on the rule sets themselves, right? That's clearly
`lacking on white paper, here it's yellow. I would also note that
`this could be done at any time when and as desired. So Roese
`itself teaches that you can change the rule sets at any time
`whenever you desire and as Dr. Reddy explains in his declaration
`at the bottom of slide 14 because these operations can be
`performed at any time and when and as desired, Roese
`contemplates performing the operations prior to processing
`packets. So here we have one way that Roese meets even
`Centripetal's narrow construction which should not be adopted.
`Moreover, Roese teaches preprocessing.
`
`
`Go back to slide 14 just briefly. The first underlined
`portion in red is preinstalled. Roese preinstalls the rule sets,
`right? This is also the abstract of Roese. So Roese gets rule sets
`and preinstalls them.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`
`
`
`
`
` 23
`
`
`
`Case IPR2018-01454
`Patent 9,674,148 B2
`
`
`Then go on to slide 15. Both experts agree that this
`
`
`would be an operation under Centripetal's construction. Here we
`have Dr. Reddy at the very top testifying that the preinstalling of
`an already received rule set is an operation performed on the rule
`sets themselves and therefore this also meets Centripetal's
`construction. Moreover, Dr. Orso, Centripetal's own expert
`testifies that writing the rules is an operation. So there's no
`dispute then that when you are installing, preinstalling the rule
`sets in Roese or making the right operation and that is an
`operation on the rule set itself. This is the second way that
`Roese teaches that preprocessing limitation even under
`Centripetal's construction which should not be adopted because
`it's not supported. Additionally, the petition explained that
`Roese in combination with Golnabi teaches a preprocessing
`limitation.
`
`
`Going to slide 16. This is from the petition at the
`very top citing to Golnabi reference which describes combining,
`reordering and splitting the rule set to generate updated rule sets.
`This is exactly what Golnabi teaches and this is exactly
`consistent with the merging, reordering or separating portion of
`the 148 patent. So this is exactly on top of the 148