`
`
`
`571-272-7822
`
`IPR2015-00314, Paper No. 38
`IPR2015-00315, Paper No. 40
`IPR2015-00316, Paper No. 38
`IPR2015-00531, Paper No. 38
`June 16, 2016
`
`
`
`
`
`
`RECORD OF ORAL HEARING
`UNITED STATES PATENT AND TRADEMARK OFFICE
`- - - - - -
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`- - - - - -
`QUALCOMM, INC.,
`Petitioner,
`vs.
`BANDSPEED, INC.,
`Patent Owner.
`- - - - - -
`Case IPR2015-00314, 00315, 00316
`Patent 7,447,624 B2
`Case IPR2015-00531
`Patent 8,542,643 B2
`Technology Center 2600
`Oral Hearing Held: Thursday, May 26, 2016
`
`
`Before: BART A. GERSTENBLITH; DAVID C. McKONE
`(via video link); and PATRICK M. BOUCHER (via video link),
`Administrative Patent Judges.
`
`The above-entitled matter came on for hearing on Thursday,
`May 26, 2016, at 1:07 p.m., Hearing Room A, taken at the U.S. Patent and
`Trademark Office, 600 Dulany Street, Alexandria, Virginia.
`
`REPORTED BY: RAYMOND G. BRYNTESON, RMR,
`
`CRR, RDR
`
`
`
`APPEARANCES:
`
`ON BEHALF OF THE PETITIONER:
`
`
`
`
`
`
`
`
`
`
`
`ON BEHALF OF THE PATENT OWNER:
`
`
`
`
`
`
`
`
`
`
`
`GREGORY DONAHUE, ESQ.
`ADAM G. PRICE, ESQ.
`DiNovo Price Ellwanger & Hardy LLP
`7000 North MoPac Expressway
`Suite 350
`Austin, Texas 78731
`512-709-9243
`
`2
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`NATHAN REES, ESQ.
`Norton Rose Fulbright US LLP
`2200 Ross Avenue
`Suite 3600
`Dallas, Texas 75201-7932
`214-855-8000
`
`RICHARD S. ZEMBEK, ESQ.
`DANIEL SCOTT LEVENTHAL, ESQ.
`Norton Rose Fulbright US LLP
`Fulbright Tower
`1301 McKinney
`Suite 5100
`Houston, Texas 77010-3095
`713-651-5151
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`3
`
`
`
`Case IPR2015-00314, 00315, 00316 (Patent 7,447,624 B2)
`Case IPR2015-00531 (Patent No. 8,542,643 B2)
`
`
`P R O C E E D I N G S
`
`(1:07 p.m.)
`JUDGE GERSTENBLITH: Please be seated. Good
`afternoon, everyone. We are here for an oral hearing in four
`related cases, IPR2015-00314, IPR2015-00315,
`IPR2015- 00316 and IPR2015-00531, Qualcomm, Incorporated
`vs. Bandspeed, Incorporated.
`I'm Judge Gerstenblith. To my left on the screen
`are Judges Boucher on the far left and McKone on the right.
`Judges McKone and Boucher are appearing remotely from
`Detroit and Denver, respectively.
`Because of limitations on the capabilities of our
`microphones, unless you are standing and speaking from the
`center podium, the Judges that are appearing remotely here
`will not be able to hear you very clearly. So we ask that when
`you are speaking and addressing the Board to please come up
`to the podium.
`Let's have each side enter their appearance. We
`will start with Petitioner. Please step up. Thank you. Sorry.
`I know it's a bit of an inconvenience. I'll remind you along
`the way.
`
`MR. REES: I am Nate Rees, representative for
`Petitioner, Qualcomm, Incorporated.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`
`
`
`4
`
`
`
`Case IPR2015-00314, 00315, 00316 (Patent 7,447,624 B2)
`Case IPR2015-00531 (Patent No. 8,542,643 B2)
`
`
`JUDGE GERSTENBLITH: Thank you. And do
`you have anybody else with you? Feel free to introduce
`anybody with you.
`MR. REES: With me is Richard Zembek, backup
`counsel, and Daniel Leventhal who is also backup counsel.
`And for reference, this is Hao Wu. He is a summer associate
`with our firm.
`JUDGE GERSTENBLITH: Excellent. Thank you
`very much. And for Patent Owner?
`MR. DONAHUE: Good afternoon, Your Honors.
`My name is Greg Donahue. I am the lead counsel for
`Bandspeed in this matter. And with me is Adam Price, who is
`backup counsel.
`JUDGE GERSTENBLITH: Excellent. Thank you
`very much.
`Welcome to the PTAB. It is good to have you
`here. If anybody brought along paper copies of their
`demonstratives, you can feel free to hand me a copy. I have
`your electronic versions so it is not required, and Judges
`McKone and Boucher have your electronic versions as well.
`Does anyone want to hand one up? Thank you
`very much.
`We set forth the procedure for today's hearing in
`the trial order, but just to remind everyone how it is going to
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`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
`
`Case IPR2015-00314, 00315, 00316 (Patent 7,447,624 B2)
`Case IPR2015-00531 (Patent No. 8,542,643 B2)
`
`work, each side will have 75 minutes to present their
`argument.
`We will first hear from Petitioner with respect to
`their case. Petitioner may reserve time for rebuttal. Then we
`will hear from Patent Owner. And then, to the extent
`Petitioner has reserved time for rebuttal, we will hear
`Petitioner's rebuttal.
`Because there are no motions in this case, Patent
`Owner does not have a rebuttal. They just have one time
`period to present their arguments.
`Additionally, what I see and what you all present
`on the screen here is not visible to Judges McKone and
`Boucher, but because they have your demonstratives, we ask
`that you please call out the slide numbers, both for their
`benefit and for the trial transcript.
`It is hard to remember that, too. I will try to
`remind you along the way, but I'm sure Judges Boucher and
`McKone will also chime in if it seems a little bit confusing
`where you are.
`But please try to say, you know, as we turn to
`slide 3, or something to that effect, or, for example, if you are
`looking at a figure from one of the patents, just call out the
`figure and the patent or reference name.
`I anticipate given the amount of time that we have
`for argument that we will be taking about a 10- minute recess.
`
`
`
`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
`
`Case IPR2015-00314, 00315, 00316 (Patent 7,447,624 B2)
`Case IPR2015-00531 (Patent No. 8,542,643 B2)
`
`Depending on how much time Petitioner decides to use in the
`opening arguments, my expectation is that we will take that
`recess after Petitioner finishes their opening and before Patent
`Owner begins its rebuttal case.
`Are there any questions on how we will proceed?
`Okay. With that, Mr. Rees, you have the floor.
`MR. REES: Thank you, Your Honor, and I would
`like to reserve 25 minutes for rebuttal.
`JUDGE GERSTENBLITH: Okay.
`MR. REES: May it please the Court. Again, my
`name is Nate Rees. I represent Qualcomm, Petitioner in this
`case.
`
`Before us are four IPRs. The 531 IPR is directed
`to the '643 patent. The 314, 15 and 16 IPRs are directed to
`the '624 patent.
`Now, while their records are relatively
`voluminous, the issues are actually fairly narrow for this case.
`For example, in the 531 IPR there are no claim construction
`issues that are presently contested. There are really only two
`independent claim limitations to discuss and two dependent
`claim limitations.
`For the 314 through 16 IPRs, four out of the seven
`claims have been admitted as being anticipated by the Gerten
`reference. There is one small claim construction issue, one
`independent claim limitation, and all of the claims have been
`
`
`
`7
`
`
`
`Case IPR2015-00314, 00315, 00316 (Patent 7,447,624 B2)
`Case IPR2015-00531 (Patent No. 8,542,643 B2)
`
`argued together, all of the independent claims have been
`argued together across the cases.
`There is two dependent claim limitations on one
`ground, and only one on another ground.
`MR. ZEMBEK: Your Honor, would you like to
`start the clock or can we have the clock started, please? Sorry
`to interrupt.
`JUDGE GERSTENBLITH: No problem. I can
`certainly try to do this. I will warn you, I haven't done this
`before, but I will give a reminder about 5 or 10 minutes
`before the end of the 50 minutes.
`MR. ZEMBEK: That's fine. Thank you. I just
`wanted to be sure we were on the clock as I am sure Patent
`Owners does.
`JUDGE GERSTENBLITH: No problem. Thank
`you for asking.
`MR. REES: Turning to slide 3, there are four
`grounds of rejection in the 531 case. Grounds 1 and 2 utilize
`Haartsen as a primary reference. Grounds 3 and 4 utilize
`Gerten as a primary reference. And grounds 2 and 4 include
`Dicker to fill in for some of the dependent claim limitations.
`Again, independent claim 1 is the only
`independent claim we will address, as Patent Owner has done
`as well, the arguments for claim 1 apply across the claim set.
`
`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 IPR2015-00314, 00315, 00316 (Patent 7,447,624 B2)
`Case IPR2015-00531 (Patent No. 8,542,643 B2)
`
`
`Turning to slide 4, as I stated, there are no claim
`construction terms being argued after the preliminary
`Institution Decision. This has been shown in the papers.
`Now, slides 5 through 10 -- I'm not going to read
`over these, we'll just skip over them -- but slides 5 through
`10, we recognize that we have the burden of proof. And while
`we are only discussing a few limitations at this hearing, I
`have provided for the PTAB's reference where in each of the
`papers each of the claim limitations have been discussed.
`If we turn to slide 11, please. By way of
`background, before we get into the grounds, I think a quick
`intro on what the '643 patent is about is useful.
`The '643 patent, the intent of the patent is to
`improve on existing Bluetooth communications. Bluetooth
`utilizes hopping sequences where a master and a slave will
`hop along a sequence of channels in order to communicate.
`Adaptive hopping, or sometimes we will refer to it
`as AFH, an adaptive hopping sequence, utilizes what
`Bluetooth does. As it hops through the channels it will check
`to see if there are channels with errors.
`If there are channels with excessive noise or any
`other reason that the channel should not be used, an adaptive
`hopping sequence can take out those channels and define the
`channel set to use.
`
`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
`
`
`
`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
`
`Case IPR2015-00314, 00315, 00316 (Patent 7,447,624 B2)
`Case IPR2015-00531 (Patent No. 8,542,643 B2)
`
`
`Like the '643 patent, the Haartsen reference
`teaches an improvement of Bluetooth by improving, by
`utilizing an adaptive hopping sequence. Shown here on the
`slide is claim 1. There are two issues to discuss. Claim 1
`requires loading a set of default channels into a default
`register and loading a set of good channels into a good
`channel register. And then the claim provides for replacing a
`bad channel stored in the default register with a good channel
`stored in the good channel register.
`If we turn to slide 12, like I said before, the
`Haartsen reference is an improvement on Bluetooth frequency
`hopping that provides for adaptive hopping sequence.
`Haartsen teaches a post-processing operation. And the
`post-processing operation is the mechanism that goes and
`measures the channels and makes the replacement.
`Haartsen teaches that channels are set in default
`registers as noted in the Institution Decision. The N1 register
`of Haartsen is a default channel register and the N2 register is
`the good channel register. And Haartsen discloses storing
`these channel sets in computer addressable medium.
`The question is, and actually the experts agree,
`would one of ordinary skill in the art find it obvious to utilize
`registers instead of -- as a computer addressable medium. Let
`me clarify what the experts agree on. The experts agree that
`registers were known to be quicker. And we say that, I mean,
`
`
`
`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
`
`Case IPR2015-00314, 00315, 00316 (Patent 7,447,624 B2)
`Case IPR2015-00531 (Patent No. 8,542,643 B2)
`
`Dr. Ding has stated that the reason why one of skill in the art
`would find it obvious is because registers are well known.
`They provide well known benefits. Dr. Melendez has also
`been on record saying registers are quicker.
`So the question of would one of ordinary skill in
`the art find it obvious to utilize registers in view of the
`Haartsen reference is before us.
`If we turn to slide 13, what I have here
`side-by-side is the Haartsen figure with the standard
`Bluetooth specification. This Haartsen figure notes that the
`dotted line is the original hopping scheme. Now, the original
`hopping scheme refers to the default Bluetooth. All right.
`If we look at the Bluetooth specification, the
`Bluetooth specification states that a kernel addresses a
`register containing hop frequencies. So --
`JUDGE McKONE: Before you go any further here,
`before you go any further, is it your argument now that
`Haartsen because it discusses Bluetooth is actually disclosing
`registers as the Bluetooth specification discloses? Is that
`what your argument is?
`MR. REES: No, no, no. What our argument was
`and still is today, in the petition we stated that the use of
`registers, the Haartsen teaching of addressable
`computer-readable storage medium, would render the use of
`registers obvious. In the response, Bandspeed argued that,
`
`
`
`11
`
`
`
`Case IPR2015-00314, 00315, 00316 (Patent 7,447,624 B2)
`Case IPR2015-00531 (Patent No. 8,542,643 B2)
`
`well, registers aren't obvious. Registers give you certain
`advantages.
`And in reply we stated, well, registers, not only
`are they obvious, they are the thing that everybody uses. So
`now the rejection is computer -- or, sorry, the ground, the
`asserted ground is whether computer-addressable medium
`renders the use of registers obvious.
`JUDGE McKONE: But it looked like the evidence
`that you were using to show that registers were well known is
`the Bluetooth specification. Is that right?
`MR. REES: Among other things. I mean, the
`testimony of Dr. Ding originally utilized -- Bluetooth has
`always been of evidence. We are not using it as a ground of
`rejection, though.
`JUDGE McKONE: But where was the Bluetooth
`specification referenced in the petition?
`MR. REES: You know, where we stated that
`registers were used is the petition at 16, but I believe that it
`was Dr. Ding that discussed the Bluetooth specification. I
`don't know that that was part of the IPR petition.
`JUDGE McKONE: It seems like the reply relied
`pretty heavily on the disclosure of the Bluetooth specification.
`And I'm concerned that that is a new argument that would be
`improper in the reply.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`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
`
`Case IPR2015-00314, 00315, 00316 (Patent 7,447,624 B2)
`Case IPR2015-00531 (Patent No. 8,542,643 B2)
`
`
`Why should we even consider the arguments as to
`the Bluetooth specification?
`MR. REES: Well, I don't think it is necessary to
`consider it. First of all, you know, obviously they come back
`and say one of ordinary skill in the art would not understand
`that registers were an obvious choice. I think it is proper in
`the scope of a reply to state, well, yes, as we stated, it would
`be an obvious choice. Computer-addressable medium --
`registers are computer-addressable medium and, oh, by the
`way, this is what we're dealing with. We're dealing with
`Bluetooth.
`I think that's within a proper scope of a reply. It
`is not a new argument asserting a new ground of rejection.
`But, like I said, I don't believe it is necessary to consider the
`Bluetooth reference at all.
`Does that answer your question?
`JUDGE McKONE: I understand your position.
`Thank you.
`MR. REES: Thank you. Now, turning to slide 15,
`the second issue is the claims require that when a bad channel
`is addressed in a default register that we replace the bad
`channel with a good channel that's chosen from a good
`channel register.
`What we have here is a teaching of Haartsen that
`was relied on in the petition that states: In one embodiment,
`
`
`
`13
`
`
`
`Case IPR2015-00314, 00315, 00316 (Patent 7,447,624 B2)
`Case IPR2015-00531 (Patent No. 8,542,643 B2)
`
`the set of allowable hop channels -- right, so these are the
`good channels -- are stored in computer-addressable medium.
`Whenever a preset phase -- so this is the phase in a
`hopping sequence going through the default register --
`whenever a preset phase is associated with a forbidden hop --
`so a bad channel -- an index to the table of allowable hop
`channels is generated.
`If we skip to the bottom highlight: The generated
`index is used to address an entry in the table 801, the good
`channel register, and retrieve a substitution hop channel.
`So this right here is a teaching of a good channel
`register. When we address a bad channel we are substituting
`the bad channel with the good channel from the good channel
`register.
`
`Turning to slide 16, in ground 2 there are a couple
`dependent claims at issue. The first one --
`JUDGE McKONE: Before we move on to that,
`what construction of register are you operating under? I know
`you proposed one construction in your petition. We gave a
`preliminary construction in our Institution Decision. And
`Patent Owner proposed something different. Yet you say
`there is no dispute as to the meaning of register.
`Can you tell me what construction you are
`operating under right now?
`
`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
`
`
`
`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
`
`Case IPR2015-00314, 00315, 00316 (Patent 7,447,624 B2)
`Case IPR2015-00531 (Patent No. 8,542,643 B2)
`
`
`MR. REES: We are operating under the PTAB's
`construction. I don't believe, I mean, there was no response
`in the Patent Owner's response that had a specific new
`construction that was different than the PTAB's construction
`of register. We didn't feel the need to reply to any
`construction. What we were saying is that, I mean, a register
`is a well known memory in computer architecture. It is a
`fundamental building block in computer architecture.
`JUDGE McKONE: Now, you also mentioned in
`your reply that the parties had reached an agreement as to the
`construction of register in District Court. Is that the case?
`MR. REES: I am not in the District Court
`litigation.
`MR. ZEMBEK: Your Honor -- Richard Zembek --
`I don't believe it is a disputed term. I don't -- I apologize, I
`don't recall the construction that was entered in the District
`Court and I don't believe that that construction would result in
`a different analysis of the issues being presented.
`We are happy to supplement the record with an
`e-mail to the Board if you would like that presents the
`agreed-to constructions from the District Court litigation
`immediately after the hearing, if that would be of use.
`JUDGE McKONE: We have Exhibit 1013 that has
`been introduced into the record. And that looks to be,
`although we don't have the cover sheet for that, it is Exhibit
`
`
`
`15
`
`
`
`Case IPR2015-00314, 00315, 00316 (Patent 7,447,624 B2)
`Case IPR2015-00531 (Patent No. 8,542,643 B2)
`
`A, which I believe the reply says it is the agreed construction.
`And that has the construction of register to be an area of
`memory having a set of addressable slots.
`Is that the parties' agreed construction in District
`Court or not?
`MR. ZEMBEK: Yes, Your Honor. It is from our
`joint submission.
`MR. PRICE: Good afternoon. Adam Price for the
`Patent Owner. It was agreed but it was borne out of an earlier
`claim construction that the court assembled in a prior case
`that Bandspeed had against another Defendant. So when we
`filed this litigation and went to Markm an, we agreed to adopt
`what the court had previously done.
`So, yes, we did agree to it in this round but it
`seemed like it would be futile to dispute it a second time to
`the court.
`
`JUDGE McKONE: Okay. I will address that with
`you more completely during your time.
`Mr. Rees, is that construction, the agreed
`construction in the District Court, incorrect? It seems broader
`than the construction that we gave and that the parties
`proposed in this matter.
`Was the District Court construction unreasonably
`
`broad?
`
`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
`
`
`
`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
`
`Case IPR2015-00314, 00315, 00316 (Patent 7,447,624 B2)
`Case IPR2015-00531 (Patent No. 8,542,643 B2)
`
`
`MR. REES: Certainly I don't think it was
`unreasonably broad. I mean, they are actually construing
`under a narrower standard in the District Court. Really when
`it comes down to it, I don't think that the construction of
`register is really dispositive in this case. I mean, a register,
`as I said, is a very well known component in computer
`architecture.
`And whether -- and a register, as Dr. Ding states, a
`computer-addressable media is a register, right, includes a
`register. A register is, in fact, a computer-addressable media.
`So one of skill in the art reading the disclosure of Haartsen
`would understand that that could be a register and would find
`that to be an obvious design choice.
`I would also note in our -- in the alternative
`grounds it explicitly teaches registers, so in the grounds
`relying on Gerten as the primary reference, in grounds 3 and
`4.
`
`JUDGE McKONE: Okay.
`MR. REES: All right. I will continue. So slide
`16, I said there were two dependent claims at issue in this
`ground. Claim 4 requires removing a particularly good
`channel from a good channel register based at least in part on
`a good channel usage timeout. Claim 5 is rescanning default
`channels based at least in part on the number of good
`channels.
`
`
`
`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
`
`Case IPR2015-00314, 00315, 00316 (Patent 7,447,624 B2)
`Case IPR2015-00531 (Patent No. 8,542,643 B2)
`
`
`Turning to slide 17, there is very little teaching in
`the '643 patent about what a good channel usage timeout is.
`But one example states that in some embodiments a good
`channel usage timeout may be helpful. Because interference
`changes over time, communication systems stop and start and
`changes in location occur, it is useful to periodically change
`the set of channels being used.
`If we go to slide 18, the Dicker reference has a
`very similar teaching. The Dicker reference utilizes two
`different error counters. There is a short-term error counter,
`that if a certain number of errors happens on a channel, it will
`swap a channel. Then there is the long-term error counter.
`And what Dicker does is they have a five-second
`slot, that once we go over this five-second slot, which is the
`good channel usage time slot, it will go into a, I believe it was
`called a rescanning phase, and where a blocked subset may be
`substituted for an unblocked subset based on the long-term
`error counter. So this is a substitution of a channel that is
`done, based at least in part, on the good channel usage
`timeout.
`
`JUDGE McKONE: Okay. For my own
`clarification here, since there is a one-second timer and a
`five-second timer, the good channel usage timeout value is the
`five-second timer?
`MR. REES: Yes, that's what we have asserted.
`
`
`
`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
`
`Case IPR2015-00314, 00315, 00316 (Patent 7,447,624 B2)
`Case IPR2015-00531 (Patent No. 8,542,643 B2)
`
`
`JUDGE McKONE: Okay. Thank you.
`MR. REES: And then this is consistent, if we look
`at slide 19, it is also consistent with what the priority
`document, the '594 provisional stated. The '594 provisional
`just talks about a clear channel usage timeout. And then the
`clear channel, after a timeout, the system scans a frequency
`band and can replace channels.
`JUDGE McKONE: Is this an argument you made
`in the petition?
`MR. REES: This was in the reply. The reference
`to the priority document, as they came back and argued that a
`good channel usage timeout is not necessarily a periodic
`timeout, we pointed out that, I think, based on the
`specification and its priority document, it looks like a
`periodic timeout should be able to apply to this claim
`limitation.
`Turning to slide 20, we are going to claim 5.
`Remember, claim 5 rescans based on the number of good
`channels. What I would like to do here -- and this is in our
`reply, too -- is focus on what the actual grounds of rejection
`are because Bandspeed has been trying to argue -- has been
`arguing the references separately on this.
`Haartsen is relied on as teaching rescanning sets
`of channels. The Haartsen reference teaches that forbidden
`channels and allowed channels can be changed dynamically
`
`
`
`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
`
`Case IPR2015-00314, 00315, 00316 (Patent 7,447,624 B2)
`Case IPR2015-00531 (Patent No. 8,542,643 B2)
`
`and there is going to be reasons why you need to rescan the
`channels.
`
`Now what is utilized on slide 21, in the Dicker
`reference, Dicker teaches the specific trigger, rescanning
`based at least in part on the good -- on the number of good
`channels. And what we noted in the petition is that Gerten
`utilizes an ISM band. There is a requirement that Gerten
`separates out its channels into 12 segments. And 10 segments
`must always be used based on the requirements of the ISM
`band.
`
`So if the number of good channels in a use
`segment starts falling below, the Dicker reference can't just
`stop using those channels. They have to, based on the number
`of good channels, substitute in a segment in order to keep to
`that 10. Sorry, I skipped into slide 22 without announcing it.
`Again, as Dicker stated, if the error rate is high we
`rescan the channels. The rescan is based on the number of
`good channels because we have to have that. And as Dr. Ding
`pointed out, Bluetooth communications have a similar
`requirement, right, there always has to be 75 of the 79
`channels working. So if we have channels drop below, one of
`skill in the art would find it obvious in view of the teachings
`of Dicker.
`
`JUDGE McKONE: I'm sorry, which counter was
`triggering Dicker's rescan, to make sure I understand you?
`
`
`
`20
`
`
`
`Case IPR2015-00314, 00315, 00316 (Patent 7,447,624 B2)
`Case IPR2015-00531 (Patent No. 8,542,643 B2)
`
`
`MR. REES: Sorry, on this one it would be the
`short- term error counter. If the short-term error counter notes
`that -- well, actually it probably could be both. But we relied
`on the short-term error counter. If the short -term error
`counter shows that there is too many errors and a segment of
`good channels needs to be substituted out, because the number
`of good channels has to be a specific number, it will
`substitute the segment based on the number of good channels,
`as a short-term.
`JUDGE McKONE: Are you saying that Dicker
`does a rescan as part of that process?
`MR. REES: Sorry, sorry, no, Dicker rescans on
`the long-term error counter. On the short-term error counter,
`that's a constant scan of the active channels. The long-term
`error counter does a constant scan on all of the channels.
`And, in fact, Bandspeed has argued that only the
`subset of active channels are scanned, but there is -- the
`long- term error counter does scan constantly both the active
`and the inactive. That is one reason why they are able to
`substitute out for the unused subset and without having to
`rescan first.
`JUDGE McKONE: Well, I thought what Dicker
`disclosed as far as the channels that weren't being used, is it
`just essentially deprecates the number of error count that it is
`
`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 IPR2015-00314, 00315, 00316 (Patent 7,447,624 B2)
`Case IPR2015-00531 (Patent No. 8,542,643 B2)
`
`assumed to have rather than rescanning those channels
`themselves. Am I incorrect?
`MR. REES: Dicker teaches that the error counter
`is -- the error counter is tracking the unused channels, to my
`knowledge. I would have to refresh my memory.
`JUDGE McKONE: Could you show me where that
`is disclosed?
`MR. REES: We are talking about Dicker column
`
`5.
`
`JUDGE McKONE: Because what I was referring
`to is in column 5 starting at about line 22 or 23: In this
`scheme, every five seconds the error count for each of the
`blocked subsets is reduced by 2.5 percent.
`And I took this to mean that this was, instead of
`doing a rescan, is if we assume that the error goes down a
`little bit on the blocked channels, rather than having to rescan
`that.
`
`But if you can show me where it shows actually
`rescanning the blocked channels, that would be helpful.
`MR. REES: What I was looking at is in column 5
`Dicker states: Or, in a five second -- so this is starting at line
`20 -- a five second period of any long- term error counter for
`an active subset has a count greater than the count of one of
`the blocked subsets, we're changing out.
`
`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
`
`
`
`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
`
`Case IPR2015-00314, 00315, 00316 (Patent 7,447,624 B2)
`Case IPR2015-00531 (Patent No. 8,542,643 B2)
`
`
`Yeah, okay. Now, that's where you are quoting:
`Every five seconds the error count in the blocked subsets is
`reduced by 2.5. This reduction is repeated up to six times
`until the error count is equal to 85 percent.
`I guess, can you give me your question again? I
`just want to make sure I'm addressing your question.
`JUDGE McKONE: The way that the Patent Owner
`reads it, and I think it is a reasonable reading, is that the
`blocked subset, they are not rescanned. Rather, we assume
`that whatever the last scan of that blocked subset the error
`count was, we're just going to reduce that by 2.5 percent, so
`I'm going to get better without having to rescan it. Why is
`that an improper reading?
`MR. REES: I think that, well, maybe not
`necessarily an improper reading of the reference, it is, first of
`all, it is Haartsen that's relied on for rescanning the channels.
`Haartsen does rescan all of the channels to change out what is
`in the forbidden channel or the allowed channel registers.
`The Gerten reference is the one that gives us that
`trigger based on the number of timeouts. Further -- sorry,
`Dicker says that.
`JUDGE McKONE: The reason why I was asking
`you to show me where the rescan was, and I was looking at
`your slide 22, and here you are saying that Dicker rescans the
`channels.
`
`
`
`23
`
`
`
`Case IPR2015-00314, 00315, 00316 (Patent 7,447,624 B2)
`Case IPR2015-00531 (Patent No. 8,542,643 B2)
`
`
`MR. REES: Right.
`JUDGE McKONE: So I'm confused. Now you're
`saying it's Haartsen actually instead of Dicker?
`MR. REES: Sorry, I apologize, I meant to say
`Dicker gives us the trigger based on a good channel timeout.
`That's probably just a miswording on the slide. We're
`utilizing Dicker as the trigger of changing out a set based on
`the number of good channels. The channel rescan, in the
`original grounds of rejection we've always utilized Haartsen
`as teaching the rescan.
`I would further note, though, that nothing in the
`claim requires that the rescan has to be done all at once.
`Nothing requires that, you know, so as Dicker is swapping out
`channels when it hits an inactive channel into an active
`subset, those channels will then be scanned. So as the process
`of Dicker goes forward, all of the channels are, in fact,
`scanned.
`
`JUDGE McKONE: Is that an argument you made
`in the petition, that a rescan can span multiple type and events
`--
`
`MR. REES: No, no, we utilized Haartsen -- I
`apologize. We utilized Haartsen for the rescan. We did not
`utilize Dicker.
`JUDGE McKONE: 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
`
`
`
`24
`
`
`
`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
`
`Case IPR2015-00314, 00315, 00316 (Patent 7,447,624 B2)
`Case IPR2015-00531 (Patent No. 8,542,643 B2)
`
`
`MR. REES: Moving on, ground 3, ground 3
`utilizes -- oh, sorry, if we look at slide 23 real quick, we have
`provided Dr. Ding saying why one of ordinary skill in the art
`would combine Haartsen in view of Dicker.
`In Patent Owner's response they stated that that
`would change