`
`
`
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`____________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`____________
`
`TCT MOBILE, INC. AND TCT MOBILE (US) INC.,
`Petitioner,
`
`v.
`
`WIRELESS PROTOCOL INNOVATIONS, INC.,
`Patent Owner.
`____________
`
`Cases IPR2016-01494, 01704, 01861 and 01865
`Patents 8,274,991 B2; 8,565,256 B2; and 9,125,051 B2
`____________
`
`Held: December 7, 2017
`____________
`
`
`
`BEFORE: KEVIN T. TURNER, MITCHELL G. WEATHERLY,
`and KAMRAN JIVANI, Administrative Patent Judges.
`
`
`
`APPEARANCES:
`
`ON BEHALF OF THE PETITIONER:
`JEREMY D. PETERSON, ESQUIRE
`JOHN D. ZELE, ESQUIRE
`Morgan, Lewis & Bockius, LLP
`1111 Pennsylvania Avenue, N.W.
`Washington, D.C. 20004-2541
`
`ON BEHALF OF PATENT OWNER:
`
`1
`
`
`
`Cases IPR2016-01494, 01704, 01861 and 01865
`Patents 8,274,991 B2; 8,565,256 B2; and 9,125,051 B2
`
`
`RICHARD E. CAMPBELL, ESQUIRE
`MICHAEL C. JONES, ESQUIRE
`Procopio
`525 B Street
`Suite 2200
`San Diego, California 92101
`
`
`
`The above-entitled matter came on for hearing on
`
`
`Thursday, December 7, 2017, commencing at 1:00 p.m., at the
`U.S. Patent and Trademark Office, 600 Dulany Street,
`Alexandria, Virginia.
`
`
`
`
`
`P R O C E E D I N G S
`- - - - -
`JUDGE WEATHERLY: Good afternoon, everyone
`here and good morning to Judge Turner. This is a hearing for
`four inter partes reviews collectively addressing claims in three
`related patents as follows: IPR2016-01494 relates to U.S. patent
`number 8,274,991; IPR2016-01704 relates to U.S. patent number
`8,565,256; and IPR2016-01861 and 01865 both relate to U.S.
`patent number 9,125,051.
`I'm Judge Weatherly, and I'm joined remotely by Judge
`Turner from our regional office in San Jose and Judge Jivani from
`Georgia. The camera position above the screen to my left on
`which Judges Turner and Jivani appear captures their view of the
`proceedings. So if at any time you wish to address them, it would
`
`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
`27
`28
`29
`30
`
`
`
`
`
` 2
`
`
`
`Cases IPR2016-01494, 01704, 01861 and 01865
`Patents 8,274,991 B2; 8,565,256 B2; and 9,125,051 B2
`
`be helpful probably for them and make it a little more natural for
`them if you look over there.
`Please also be sure to describe any slides that you are
`discussing by number so that Judges Turner and Jivani can follow
`along with the presentation a little more easily and also to
`improve the usefulness of the transcript for the three of us after
`the hearing.
`Pursuant to our hearing order, each party has a total of
`90 minutes to present their argument. Petitioner will proceed first
`because it bears the burden of proof, followed by the patent
`owner. Petitioner may reserve time solely to rebut arguments that
`patent owner advances in their presentation.
`Because the hearing addresses four proceedings, I
`wanted to get the parties' input on how, if at all, they wanted to
`divide time among the different proceedings. I know that patent
`owner has presented not only separate slide decks for individual
`proceedings, but it has also provided a unified slide deck.
`MR. CAMPBELL: Your Honor, that's the one we
`intend to proceed with today.
`JUDGE WEATHERLY: Petitioner, do you have any
`preferences about how we --
`MR. PETERSON: Your Honor, I just submitted four
`because at the time I didn't know how that time was going to be
`allocated. I was just going to go through them right in order, if
`that's okay.
`
`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
`
`
`
`
`
` 3
`
`
`
`Cases IPR2016-01494, 01704, 01861 and 01865
`Patents 8,274,991 B2; 8,565,256 B2; and 9,125,051 B2
`
`
`JUDGE WEATHERLY: So in your view, it's fine for
`you to do a single presentation on all four proceedings in one
`shot, so to speak, and patent owner will have its time to oppose
`and then whatever time you've reserved?
`MR. PETERSON: That's correct.
`JUDGE WEATHERLY: Fantastic. Before we actually
`really begin the arguments, I would like everybody in the room to
`introduce themselves and who they brought with them, beginning
`with counsel for petitioner.
`MR. PETERSON: My name Jeremy Peterson
`representing petitioner, and with me is John Zele, also
`representing petitioner.
`MR. CAMPBELL: I'm Richard Campbell representing
`the patent owner, Wireless Protocol Innovations. With me today
`is my colleague, Michael Jones. We also have representatives
`from the patent owner, Nicholas Wilson and Adrian Smith.
`JUDGE WEATHERLY: Thanks very much. Also,
`before we begin the substantive argument, I have one request of
`patent owner. I know petitioner, at least in the 1494 IPR and the
`1704 IPRs, filed objections to your demonstrative exhibits. And I
`wanted to know from you whether there are any factual
`inaccuracies in those objections, in those two proceedings.
`MR. CAMPBELL: What they are, Your Honor, is they
`are what we call adjacent quotes from quotes recited before. And
`they also arose --
`
`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
`
`
`
`
`
` 4
`
`
`
`Cases IPR2016-01494, 01704, 01861 and 01865
`Patents 8,274,991 B2; 8,565,256 B2; and 9,125,051 B2
`
`
`JUDGE WEATHERLY: I'm familiar with what the
`objections are and I didn't really want to address the objections
`unless there were factual inaccuracies.
`MR. CAMPBELL: We are happy with that, that you
`can consider them after the hearing.
`JUDGE WEATHERLY: Are there any factual
`inaccuracies of which you are aware in those objections?
`MR. CAMPBELL: No, not that we are aware.
`JUDGE WEATHERLY: You understand what I mean
`by factual inaccuracies?
`MR. CAMPBELL: Yes.
`JUDGE WEATHERLY: So with that out of the way,
`petitioner, how much time would you like to reserve for rebuttal?
`MR. PETERSON: Probably about a half an hour.
`JUDGE WEATHERLY: I made the right guess then.
`So you'll have 60 minutes in your opening presentation and you
`may begin whenever you are ready.
`MR. PETERSON: Good afternoon, Your Honor. Good
`morning, Your Honor. My name is Jeremy Peterson. I'll be
`representing the petitioners in the instituted IPRs. I'll start with
`the demonstratives for the '991 patent. Patent owner has five
`different arguments relating to the '991 patent. First patent owner
`argues that the descriptive words or figures -- I'm sorry, this is
`demonstrative 5. The descriptive words or figures used in the
`prior art are not the same ones used in the patent. For instance,
`
`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
`
`
`
`
`
` 5
`
`
`
`Cases IPR2016-01494, 01704, 01861 and 01865
`Patents 8,274,991 B2; 8,565,256 B2; and 9,125,051 B2
`
`patent owner argues state boundaries are different even though
`underlying operations are the same as the prior art.
`Second, patent owner attempts to distinguish a claim
`element from a prior art feature that is identical to the feature in
`the patent specification. For example, the prior art in the '991
`patent both use unicast polling for noncontention access from the
`grant pending absent state.
`Third, patent owner ignores relevant passages of the
`prior art while focusing on irrelevant ones. For example, patent
`owner repeatedly focuses on Figure K1 instead of focusing on
`where noncontention access happens.
`Fourth, patent owner reads the prior art in a vacuum and
`makes no attempt to understand it as a person of ordinary skill in
`the art would. For example, patent owner just keeps looking to
`Figure K1 just because it has a state called idle.
`Fifth, patent owner redefines simple words to have
`restrictive meanings unsupported by the patent specification. For
`instance, patent owner refuses to understand that a BSC sending
`empty packets means that it has no data to send.
`Turning to demonstrative 10, there are three grounds
`related to the '991 patent. Ground 1 is a combination of
`Abi-Nassif and the DOCSIS 1.1 and ground 2 is a combination of
`DOCSIS and Abi-Nassif.
`JUDGE WEATHERLY: Why do you present them as
`different grounds?
`
`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
`
`
`
`
`
` 6
`
`
`
`Cases IPR2016-01494, 01704, 01861 and 01865
`Patents 8,274,991 B2; 8,565,256 B2; and 9,125,051 B2
`
`
`MR. PETERSON: Well, because we look to
`different -- we look to different evidence to support the
`obviousness combinations on both of them. It's not just that for
`Abi-Nassif we look only to Abi-Nassif but to DOCSIS for just
`one element and for DOCSIS we look to only DOCSIS and
`Abi-Nassif for just one element. They both pretty much have all
`or arguably they have all of the claim elements except for the
`time now in Abi-Nassif, but they are different grounds.
`Abi-Nassif teaches the second-to-last limitation of
`claim 1. Abi-Nassif teaches starting in the inactive state after
`receiving a contention-free opportunity to transmit a request. The
`user will request bandwidth, receive a grant and transition from
`the inactive to the active state. Abi-Nassif also teaches that the
`MAC user can contend for bandwidth via the contention state.
`This route also meets the claim element.
`Turning to demonstrative 17, patent owner, however,
`asserts that Abi-Nassif does not disclose the grant pending absent
`state in the three elements shown. Turning to demonstrative 18,
`the grant pending absent state is characterized by transmitting a
`noncontention request which unquestionably exists in Abi-Nassif.
`Both the patent and Abi-Nassif teach unicast polling.
`Turning to demonstrative 19, Abi-Nassif also teaches
`the transition from the grant pending absent state to the grant
`pending state. The inactive state of Abi-Nassif means all the
`requirements of the grant pending absent state. Annotated
`
`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
`
`
`
`
`
` 7
`
`
`
`Cases IPR2016-01494, 01704, 01861 and 01865
`Patents 8,274,991 B2; 8,565,256 B2; and 9,125,051 B2
`
`Figure 9 also shows a noncontention request prior to the
`transition to the active state or the grant pending state. This
`figure shows that a noncontention request is met, i.e., that a grant
`transmission occurs by the words "request satisfied."
`Turning to demonstrative 20, patent owner argues their
`unsound argument number 1 that this claim element is not met
`because the noncontention request is sent in an inactive state and
`the grant is received in an active state. As shown, patent owner's
`own expert undercut their position. Whether the transitions are
`part of one state or another is purely a conceptual exercise. Not a
`patentable distinction.
`Turning to demonstrative 21, Abi-Nassif teaches the last
`element of claim 1. Abi-Nassif teaches transitioning into the
`inactive state after transmitting all its data. Abi-Nassif also
`teaches remaining in the inactive state so long as it has no data to
`transmit. Abi-Nassif does not teach a timeout, but DOCSIS
`teaches the time out. Figure 9 of Abi-Nassif shows transitioning
`from the inactive state to the contention state. This is the same
`transition described by the '991 patent and going from idle to a
`contention request. Therefore, Abi-Nassif's inactive state teaches
`the claimed grant pending absent and idle states. As stated above,
`the only thing missing is the timeout.
`If one were to put a timeout between the lines on the
`inactive state between the noncontention request and the
`contention request of Figure 9, then we would have the exact
`
`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
`
`
`
`
`
` 8
`
`
`
`Cases IPR2016-01494, 01704, 01861 and 01865
`Patents 8,274,991 B2; 8,565,256 B2; and 9,125,051 B2
`
`disclosure in Figure 9 of the patentable claims. We could be in
`an active state which has piggy-backing going to the inactive
`state which has unicast requests and then waiting for data, and
`when the data is received, performing a noncontention request
`and going back to idle. Or if no data is received, the inactive
`state becomes idle and you have to take a contention request to
`get out of the idle state.
`Turning to demonstrative 22, there are a couple of
`reasons to combine Abi-Nassif with DOCSIS 1.1. First,
`Abi-Nassif is based on DOCSIS 1.0 as its preferred embodiment
`which is incorporated by reference in its entirety in Abi-Nassif.
`Because Abi-Nassif is based on DOCSIS 1.0, a person of
`ordinary skill in the art would have recognized that its method
`was ready for improvement by adding additional features from a
`newer degeneration of DOCSIS 1.1. The improvement would
`have been predictable because DOCSIS 1.1 is backward
`compatible. This combination would also preserve resources
`because it would provide a cutoff for when the CPE in Abi-Nassif
`no longer needs to respond to polling requests thereby, saving
`bandwidth resources.
`Turning to demonstrative 31, DOCSIS 1.1 teaches the
`highlighted limitation. The limitation reads on the nrtPS service
`flow. Here the CMTS must provide the CM with timely unicast
`opportunities whether or not the CMS data transmits. Again,
`DOCSIS 1.1 uses unicast opportunities just like Abi-Nassif and
`
`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
`
`
`
`
`
` 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
`
`Cases IPR2016-01494, 01704, 01861 and 01865
`Patents 8,274,991 B2; 8,565,256 B2; and 9,125,051 B2
`
`just like the patent. Logically, if the CM has no data to transmit,
`the CM will not request a grant of upstream bandwidth.
`Turning to demonstrative 32, first patent owner claims a
`service flow running out of data to send as an imagined scenario.
`In unicast polling, the BSC asks the CPE if it needs to transmit
`data. It would not ask if the possibility were not there that it had
`no data to send. DOCSIS 1.1 also describes a timeout for service
`flow which is one ahead of data. The patent owner is also
`consistently pointing to the idle state in Figure K1 which
`undercuts its own argument. It would not need an idle state if
`there was no data to transmit.
`Turning to demonstrative 33, patent owner goes to great
`lengths to rely on Figure K1 which is related to contention access.
`However, the patent owner ignores the unicast polling in
`DOCSIS 1.1.
`JUDGE WEATHERLY: Let me back up a little bit.
`You say on your slide 32 there's a statement that unicast polling is
`well known and that it is the very grant pending absent state
`disclosed in the '991 patent specification. By "it," I presume you
`mean to refer to unicast polling?
`MR. PETERSON: Yes.
`JUDGE WEATHERLY: But unicast polling, as I
`understand it, isn't really a state. It's more an operation.
`MR. PETERSON: Well, the state is just waiting for
`
`data.
`
`
`
`
`
` 10
`
`
`
`Cases IPR2016-01494, 01704, 01861 and 01865
`Patents 8,274,991 B2; 8,565,256 B2; and 9,125,051 B2
`
`
`JUDGE WEATHERLY: And unicast polling occurs
`during a certain state?
`MR. PETERSON: The grant pending absent state is
`really defined by just waiting for data and then sending that data
`through the unicast request.
`JUDGE WEATHERLY: So is your argument that
`unicast polling, when unicast polling is occurring, the hardware is
`in a grant pending absent state as the claims use that term?
`MR. PETERSON: When there's no data to send.
`Demonstrative 33, patent owner goes to great lengths to
`rely on Figure K1, which is related to contention access. The
`grant pending absent state exists in DOCSIS between when the
`CM last required bandwidth when it had data and the maximum
`duration of time that the resources remain unused on an active
`service flow such as an nrtPS. The CM continues to receive
`unicast polling opportunities and can still request bandwidth in
`the unicast request log.
`Turning to demonstrative 34, DOCSIS 1.1 also teaches
`the highlighted limitation. Upon receipt of data, DOCSIS teaches
`the CM will request bandwidth in a unicast slot identified by the
`polling CMTS.
`Turning to demonstrative 35, DOCSIS 1.1 also teaches
`the second-to-last limitation of claim 1. After receiving a
`bandwidth grant from the CMTS in response to the unicast
`request, the CM can again transmit data to the CMTS.
`
`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
`
`
`
`
`
` 11
`
`
`
`Cases IPR2016-01494, 01704, 01861 and 01865
`Patents 8,274,991 B2; 8,565,256 B2; and 9,125,051 B2
`
`Accordingly, the nrtPS service flow resumption of upstream
`transmission to the CMTS corresponds to the claimed transition
`back to the grant pending absent state after a subsequent
`bandwidth grant is received at the CPE.
`Turning to demonstrative 36, patent owner once again
`would like the Board to ignore the evidence supplied in the
`petition and focus on the irrelevant K1. As discussed previously,
`during the grant pending absent state, the CM is still being polled
`with the unicast request log. If the CM receives data, it can
`request additional noncontention bandwidth and enter the grant
`pending state, thus meeting this claim limitation.
`Turning to demonstrative 37, DOCSIS 1.1 teaches the
`final limitation of claim 1. DOCSIS 1.1 teaches a timeout to
`determine whether the nrtPS service flow has become inactive.
`Upon expiration of this timeout, the nrtPS service flow is deemed
`inactive and will no longer be able to send packets, thereby
`teaching transitioning to an idle state. The CM is waiting for data
`that has no data to send. That's why it went into the inactive state
`in the first place.
`As CM's failure to request bandwidth prior to expiration
`of the timeout teaches reaching the claim -- or teaches the
`claimed if the CM does not transmit any first-time bandwidth
`request to the BSC during a timeout period.
`Turning to demonstrative 38, the petition shows that
`DOCSIS' polling service discloses the grant pending absent state.
`
`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
`
`
`
`
`
` 12
`
`
`
`Cases IPR2016-01494, 01704, 01861 and 01865
`Patents 8,274,991 B2; 8,565,256 B2; and 9,125,051 B2
`
`However, nearly the entirety of patent owner's response focuses
`on the alleged deficiencies of K1. Patent owner also claims that
`since the idle state of DOCSIS is similar to the idle state of the
`'991 is patent, it would not have been obvious to add the idle state
`to the '991 patent.
`Turning to demonstrative 39, Figure K1 is irrelevant to
`this claim element. It's related to contention resolution. Not
`noncontention requests. The proper place to find noncontention
`access is a service flow with noncontention access such as the
`nrtPS. The idle state of claim 1 recites nothing more than a
`nonoperative state which is met by an inactive service flow.
`Patent owner is attempting to narrow the idle state definition
`beyond its broadest reasonable interpretation.
`JUDGE WEATHERLY: Which aspect of that
`narrowing is the improper narrowing, in your view?
`MR. PETERSON: Well, they are trying to define it by
`just a very specific preferred embodiment in the patent
`specification which was not included in the actual claim
`elements. They are attempting to limit it to just waiting, that an
`idle state is a state in which you are waiting for data. An idle
`state could mean something much broader than that. It could
`mean it's inactive or it's not working.
`JUDGE WEATHERLY: Well, okay. The interesting
`part about an idle state could mean something is that part of our
`job is to determine what an idle state does mean. And so you've
`
`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
`
`
`
`
`
` 13
`
`
`
`Cases IPR2016-01494, 01704, 01861 and 01865
`Patents 8,274,991 B2; 8,565,256 B2; and 9,125,051 B2
`
`said that patent owner has improperly tried to limit idle state to a
`specific embodiment in its specification. Are there other
`embodiments disclosed in the specification that comport with
`your more broad, your broader definition of idle state?
`MR. PETERSON: The patent specification is almost
`completely silent on the idle state. It basically just says it's
`waiting for data in, I think, one spot.
`JUDGE WEATHERLY: You say it's almost silent, but
`you have also said that patent owner identifies a single
`embodiment. So that's at least one place in which the
`specification isn't silent about idle state, correct?
`MR. PETERSON: Correct. They are pointing to a spot
`in the specification that describes an idle state as a state waiting
`for data. But we think that's -- it's improper to limit it from the
`one embodiment in the specification.
`JUDGE WEATHERLY: Back to my other question,
`are there other embodiments disclosed and that use the
`concept -- discuss the concept of idle state in a broader fashion?
`MR. PETERSON: No, Your Honor.
`JUDGE WEATHERLY: So they have some
`ammunition. Patent owner has some ammunition for their
`interpretation in their specification, and you can't identify other
`parts of the specification that are consistent with your position?
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`
`
`
`
`
` 14
`
`
`
`Cases IPR2016-01494, 01704, 01861 and 01865
`Patents 8,274,991 B2; 8,565,256 B2; and 9,125,051 B2
`
`
`MR. PETERSON: Not in the specification. But we
`also believe the claims are met even if the construction for idle
`state is waiting for data.
`JUDGE WEATHERLY: So focus on that for a little bit
`for me, then.
`MR. PETERSON: Right now?
`JUDGE WEATHERLY: Yeah.
`MR. PETERSON: I mean, for instance, in the inactive
`state in Abi, it is waiting for data because when it gets a data, it
`responds --
`JUDGE WEATHERLY: By "it," you mean the cable
`modem?
`MR. PETERSON: Sorry, yes. The cable modem, when
`it gets data, responds to a polling request and moves back into the
`active state. In DOCSIS, DOCSIS has a service flow which is
`either sending data upstream through piggy-back requests or it's
`capable of not having any data which would be the grant pending
`absent state, because that's why there's a timeout. It doesn't want
`it to stay in that state where there's no data for a long time. So it
`has a timeout.
`JUDGE WEATHERLY: Let me slow you down a little
`bit. One of the challenges for this case for me, I'm not going to
`speak for my co-panelists, but for me is that a lot of terminology,
`a lot of acronyms, I, for example, don't like calling a cable
`modem CM. It makes it a lot more difficult for my mechanical
`
`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
`
`
`
`
`
` 15
`
`
`
`Cases IPR2016-01494, 01704, 01861 and 01865
`Patents 8,274,991 B2; 8,565,256 B2; and 9,125,051 B2
`
`brain to sort of visualize. A lot of use of pronouns in that stream
`of argument that you just gave me. But I want to try to
`understand more fully. As I understand the claims, the claims are
`methods but they are methods that are performed by hardware,
`pretty well known hardware, frankly, in a system that delivers
`data to a cable modem from a head end.
`And I realize I don't really have a question pending, but
`I stopped you because your argument was too hard for me to
`follow. Too many pronouns, too many acronyms, and if you
`could keep that in mind, it will help me a lot to understand what
`you are saying, which is my primary goal.
`MR. PETERSON: I will try to keep that in mind. And
`if I relapse back into it, please let me know.
`JUDGE WEATHERLY: Okay, I will let you know.
`JUDGE TURNER: Counsel, before you move on, can I
`ask a quick question, from an electrical background, state
`machines aren't new. An idle state is a known concept. Doesn't
`the IEEE have a definition for idle state?
`MR. PETERSON: They may. I'm not aware if they do
`
`or not.
`
`JUDGE TURNER: It seems like that might be
`important to your case. Let me be so bold as to suggest if you are
`saying that idle state isn't really disclosed in the specification
`other than one embodiment so you are on the horns of a dilemma
`because they are calling the tune, but if you say idle state was
`
`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
`
`
`
`
`
` 16
`
`
`
`Cases IPR2016-01494, 01704, 01861 and 01865
`Patents 8,274,991 B2; 8,565,256 B2; and 9,125,051 B2
`
`understood by one of ordinary skill in the art to mean this, and it's
`a nice broad definition, that's something that might be persuasive.
`So perhaps I'm slightly perplexed why you wouldn't refer to a
`broad definition, because it seems like you need to.
`MR. PETERSON: I'm unaware if there is one. I was
`not able to find one.
`JUDGE WEATHERLY: So what is the evidence that
`you're using? I know you cite evidence in your papers, but tell
`me now what is the evidence that you adduce for your definition
`of idle state?
`MR. PETERSON: We cite, I believe we cite to our
`expert's declaration. We cite to the very dictionaries that patent
`owner cited to for their support for their definition.
`JUDGE WEATHERLY: Any deposition testimony
`that's relevant on that point?
`MR. PETERSON: I believe there is. I think I have it in
`a later slide.
`JUDGE WEATHERLY: Okay.
`MR. PETERSON: The '991 patent describes idle such
`that the CPE would have to use contention that would be allowed
`resources in order to transmit the data. After the nrtPS timeout,
`it's a nonrealtime polling service, and DOCSIS, the service flow
`ends and begins a new service flow that needs to be used in
`resources allocated the same as the '991 patent.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`
`
` 17
`
`
`
`Cases IPR2016-01494, 01704, 01861 and 01865
`Patents 8,274,991 B2; 8,565,256 B2; and 9,125,051 B2
`
`
`JUDGE WEATHERLY: Also I'm having a little
`trouble hearing you. I don't know if your mic is active or not, but
`it's a small room and I'm having trouble hearing you.
`MR. PETERSON: I'll try to speak up. Turning to
`demonstrative 41, patent owner also argues that the timeout in
`DOCSIS is an operation of the cable modem. Not -- sorry, the
`cable modem --
`JUDGE WEATHERLY: You can call the head end the
`
`CMTS.
`
`MR. PETERSON: Not the cable modem. The result is
`the same. The timeout still terminates the service flow for both
`the cable modem and the head end, i.e., there's a transition. The
`transition is the CM to an idle state, the cable modem, because
`the cable modem is notified of the change. Moreover, the claims
`do not recite where the timeout is located, only that the state
`transitioned if no data is received at a cable modem before a
`timeout. Patent owner's expert concedes this.
`Turning to demonstrative 46, we'll now explore
`ground 3, Sen, patent owner's admitted prior art and Rydnell.
`Turning to demonstrative 52, Sen teaches the limitation of
`transitioning from the grant pending state to the grant pending
`absent state. Sen's transition from the packet transfer state to the
`packet standby state teaches this limitation. Data is being
`transmitted upstream in the packet transfer state. In the packet
`
`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
`
`
`
`Cases IPR2016-01494, 01704, 01861 and 01865
`Patents 8,274,991 B2; 8,565,256 B2; and 9,125,051 B2
`
`standby state, the MS transmits an arbitrary number of empty
`packets which denotes that the MS is inactive.
`Turning to demonstrative 53, patent owner asserts that
`since packet transfer state is not the claimed grant pending absent
`state because Sen's MS allegedly receives a packet immediate
`assignment message in the Mac contention state and not the
`Mac -- or the packet transfer state. As previously discussed, one,
`subsequent bandwidth requests, the ones after the first; and two,
`piggy-backing requests satisfy the transmit the further bandwidth
`request limitation. Subsequent bandwidth requests and
`piggy-backing requests happen in the packet transfer state.
`Turning to demonstrative 54, the grant pending state,
`has three features, waiting for a grant, transmitting data using the
`granted bandwidth and sending further requests. Sen describes
`each of these.
`Turning to demonstrative 55, Sen describes (a) at
`Exhibit 1023, page 4 or column 4, lines 20 to 25, a person of
`ordinary skill in the art would understand teaches receiving the
`packet immediate assignment message at the start of its packet
`transfer state during the transition or as the final step of the prior
`state. Sen describes (b) at Figure 4 which shows the packet
`transfer state. Sen describes (c) sending further requests using the
`granted bandwidth at Exhibit 1023 at 4, 25 to 27. Even if it did
`not disclose (c) piggy-backing, it was known as prior art.
`
`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
`
`
`
`Cases IPR2016-01494, 01704, 01861 and 01865
`Patents 8,274,991 B2; 8,565,256 B2; and 9,125,051 B2
`
`
`Turning to demonstrative 56, patent owner again
`attempts to say that a claim element is not met because a
`bandwidth grant happens in one state versus another. This is just
`impermissible line drawing.
`Turning to slide 57, patent owner claims to be mystified
`as to the purpose of the empty packets. This is an example of
`unsound argument number 4. Sen teaches that the MS transitions
`from the packet transfer state to this packet standby state after the
`network receives an arbitrary number of empty packets. The
`rational understanding is it's sending empty packets because it has
`no data to send.
`JUDGE WEATHERLY: Empty packets are those
`packets that have header information but no data?
`MR. PETERSON: Right. Sen teaches, turning to
`demonstrative 58, Sen teaches the highlighted limitation. Sen
`teaches sending a control packet in Sen's packet standby state.
`The packet standby state permits the MS to stay connected to the
`network even though it does not need to send any data, thereby
`facilitating a fast switch from the inactive state to being able to
`transmit data immediately when the MS becomes active. When
`data again needs to be transmitted, the MS can transmit a control
`packet in packet standby state to request more bandwidth.
`Patent owner also claims that Sen's control packets use
`contention. However, path F in Figure 4 shows that this occurs
`without contention. Turning to demonstrative 60, as is shown,
`
`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
`
`
`
`
`
` 20
`
`
`
`Cases IPR2016-01494, 01704, 01861 and 01865
`Patents 8,274,991 B2; 8,565,256 B2; and 9,125,051 B2
`
`Sen teaches the second-to-last limitation of claim 1. Sen teaches
`that in response to a control packet requesting more bandwidth,
`the base station reallocates bandwidth to the MS. The MS can
`use this grant of extra bandwidth to transmit its data upstream.
`Turning to demonstrative 62, the petition also identifies
`the claim states and transitions. The petition references packet
`standby state as the GPAS or the grant pending absent state and
`Sen's packet transfer state as the grant pending state. Sen also
`teaches the last limitation of claim 1. Sen teaches a transition
`from the packet standby state into an idle state if a release timer
`expires on demonstrative 64.
`Turning to demonstrative 65, as shown, Rydnell
`discloses repeated uplink opportunities to the mobile station in
`the FPM as long as the FPM inactivity timer does not expire. If it
`expires, the mobile station enters a sleep mode, which is the idle
`state. A person of ordinary skill in the art would have been
`motivated to use Rydnell's timeout instead of transmitting empty
`packets.
`Turning to demonstrative 69, patent owner faults the
`petition for not correlating Rydnell's FPM with the grant pending
`state. However, petitioner did not rely on Rydnell for this
`purpose. The grant pending absent state was already disclosed by
`Sen. Petitioner relies on Rydnell for disclosing an inactivity
`timer in a sleep mode.
`
`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
`
`
`
`Cases IPR2016-01494, 01704, 01861 and 01865
`Patents 8,274,991 B2; 8,565,256 B2; and 9,125,051 B2
`
`
`Turning to demonstrative 74, as shown, there are
`numerous reasons to combine the p