`
`·2· · · · · · · ·----------------------------
`
`·3· · · · BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`·4· · · · · · · ·----------------------------
`
`·5· · · · · · · · · · · · APPLE INC.
`
`·6· · · · · · · · · · · · Petitioner
`
`·7· · · · · · · · · · · · · · v.
`
`·8· · · · · · · ANDREA ELECTRONICS CORPORATION
`
`·9· · · · · · · · · · · Patent Owner
`
`10· · · · · · · ·----------------------------
`
`11· · · · · · · · · Case No. IPR2017-00626
`
`12· · · · · · · · U.S. Patent No. 6,363,345
`
`13
`
`14
`
`15· · · · · DEPOSITION OF BERTRAND HOCHWALD, PH.D.
`
`16· · · · · · · · · · ·Washington, D.C.
`
`17· · · · · · · · ·Tuesday, March 13, 2018
`
`18· · · · · · · · · · · · 10:14 a.m.
`
`19
`
`20
`
`21
`
`22· ·Job No.:· BO-166003
`
`23· ·Pages 1 - 79
`
`24· ·Reported By:· Joan V. Cain
`
`IPR2017-00626
`Ex. 2007 1/93
`
`
`
`·1· · · · · Deposition of BERTRAND HOCHWALD, PH.D., held
`
`·2· ·at the law offices of:
`
`·3
`
`·4· · · · · · · ·SIDLEY AUSTIN, LLP
`
`·5· · · · · · · ·1501 K Street, Northwest
`
`·6· · · · · · · ·Washington, D.C. 20005
`
`·7· · · · · · · ·(202) 736-8000
`
`·8
`
`·9· · · · · Pursuant to Notice, before Joan V. Cain,
`
`10· ·Court Reporter and Notary Public in and for the
`
`11· ·District of Columbia.
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`IPR2017-00626
`Ex. 2007 2/93
`
`
`
`·1· · · · · · · ·A P P E A R A N C E S
`
`·2
`
`·3· ·ON BEHALF OF THE PETITIONER:
`
`·4· · · ·THOMAS A. BROUGHAN, III, ESQUIRE
`
`·5· · · ·SIDLEY AUSTIN LLP
`
`·6· · · ·1501 K Street, Northwest
`
`·7· · · ·Washington, D.C. 20005
`
`·8· · · ·Telephone: (202) 736-8000
`
`·9· · · ·E-mail:· tbroughan@sidley.com
`
`10
`
`11· ·ON BEHALF OF THE PATENT OWNER:
`
`12· · · ·BRADLEY T. LENNIE, ESQUIRE
`
`13· · · ·PEPPER HAMILTON LLP
`
`14· · · ·600 Fourteenth Street, Northwest
`
`15· · · ·Washington, D.C. 20005
`
`16· · · ·Telephone: (202) 220-1200
`
`17· · · ·E-mail:· lennieb@pepperlaw.com
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`IPR2017-00626
`Ex. 2007 3/93
`
`
`
`·1· · · · · · · · · · ·C O N T E N T S
`
`·2· ·EXAMINATION OF BERTRAND HOCHWALD, PH.D.· · · · PAGE
`
`·3· · By Mr. Lennie· · · · · · · · · · · · · · · · · ·5
`
`·4
`
`·5· · · · · · · · · · ·E X H I B I T S
`
`·6· · · · · · · (Attached to the Transcript.)
`
`·7· ·DEPOSITION EXHIBITS· · · · · · · · · · · · · · PAGE
`
`·8· · EXHIBIT 1· ·Reply Declaration of· · · · · · · · 5
`
`·9· · · · · · · · Bertrand Hochwald Regarding
`
`10· · · · · · · · U.S. Patent 6,363,345,
`
`11· · · · · · · · 2/7/18
`
`12· · EXHIBIT 2· ·An Efficient Algorithm to· · · · · 14
`
`13· · · · · · · · Estimate the Instantaneous
`
`14· · · · · · · · SNR of Speech Signals by
`
`15· · · · · · · · Rainier Martin
`
`16· · EXHIBIT 3· ·Figure 2 from An Efficient· · · · ·20
`
`17· · · · · · · · Algorithm to Estimate the
`
`18· · · · · · · · Instantaneous SNR of Speech
`
`19· · · · · · · · Signals by Rainier Martin
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`IPR2017-00626
`Ex. 2007 4/93
`
`
`
`·1· · · · · · · · · P R O C E E D I N G S
`
`·2· · · · · · · · ·BERTRAND HOCHWALD, PH.D.
`
`·3· ·having been duly sworn under penalties of perjury by
`
`·4· ·the Notary Public, was examined and did testify as
`
`·5· ·follows:
`
`·6· · · · EXAMINATION BY COUNSEL FOR THE PATENT OWNER
`
`·7· ·BY MR. LENNIE:
`
`·8· · · Q· · Good morning, Dr. Hochwald.
`
`·9· · · A· · Good morning.
`
`10· · · Q· · Let's start off by just introducing your
`
`11· ·declaration that you prepared in connection with the
`
`12· ·626 IPR.
`
`13· · · A· · I have a clean copy here if --
`
`14· · · Q· · I need to mark it, so I'll give you one
`
`15· ·that's marked, and we can use those.
`
`16· · · · · ·MR. LENNIE:· Here's one for Tom.
`
`17· · · · · ·MR. BROUGHAN:· Thank you.
`
`18· · · · · ·MR. LENNIE:· Mark this as Exhibit 1.
`
`19· · · · · · · · ·(Deposition Exhibit 1 was marked for
`
`20· ·identification.)
`
`21· ·BY MR. LENNIE:
`
`22· · · Q· · So can you identify Exhibit 1 for the
`
`23· ·record?
`
`24· · · A· · Yes.· It's my reply declaration.
`
`IPR2017-00626
`Ex. 2007 5/93
`
`
`
`·1· · · Q· · And can you explain the process that you
`
`·2· ·went through to prepare this declaration?
`
`·3· · · · · ·MR. BROUGHAN:· I'll object to the extent it
`
`·4· ·calls for privileged or work product protected
`
`·5· ·information.
`
`·6· · · · · ·I'll caution you not to reveal the
`
`·7· ·substance of our communications, but if you want to
`
`·8· ·discuss at a high-level processes used, you may.
`
`·9· · · · · ·THE WITNESS:· Yeah, sure.· I used, well, I
`
`10· ·guess the reply of the Andrea expert Douglas as a
`
`11· ·starting point to try to identify things that needed
`
`12· ·response.
`
`13· ·BY MR. LENNIE:
`
`14· · · Q· · So you started off by looking at
`
`15· ·Dr. Douglas's declaration that was submitted with
`
`16· ·the patent owner's response?
`
`17· · · A· · Yes.· That's correct.
`
`18· · · Q· · And did you look at Dr. Douglas's
`
`19· ·declaration that was submitted in the patent owner's
`
`20· ·response in connection with the 627 IPR?
`
`21· · · A· · I glanced at it, yes.
`
`22· · · Q· · You didn't read it thoroughly?
`
`23· · · A· · Well, I probably did, but that would have
`
`24· ·been I guess a couple months ago, but, yeah, I
`
`IPR2017-00626
`Ex. 2007 6/93
`
`
`
`·1· ·didn't -- didn't rely on that for this response.
`
`·2· · · Q· · Okay.· What else, if anything, did you
`
`·3· ·review in connection with your preparation of your
`
`·4· ·reply declaration that we've marked as Exhibit 1?
`
`·5· · · · · ·MR. BROUGHAN:· Just caution you again not
`
`·6· ·to reveal the substance of communications with
`
`·7· ·counsel, but subject to that, you may answer.
`
`·8· · · · · ·THE WITNESS:· Sure.· I -- the various
`
`·9· ·references that are contained in here, I reviewed
`
`10· ·those.· Those are basic ingredients for this reply.
`
`11· ·BY MR. LENNIE:
`
`12· · · Q· · And did you prepare the first draft of your
`
`13· ·declaration that we've marked as Exhibit 1?
`
`14· · · A· · Can't say for sure.· At least parts of it,
`
`15· ·yes.
`
`16· · · Q· · Which parts of Exhibit 1 did you prepare
`
`17· ·the first draft of?
`
`18· · · · · ·MR. BROUGHAN:· I'll caution you not to
`
`19· ·reveal the substance of communications with counsel
`
`20· ·as it's privilege or work product protected.· To the
`
`21· ·extent you can remember, you can answer.
`
`22· · · · · ·THE WITNESS:· Yeah, I probably don't
`
`23· ·remember exactly what or where, but yeah.
`
`24· ·BY MR. LENNIE:
`
`IPR2017-00626
`Ex. 2007 7/93
`
`
`
`·1· · · Q· · So you can't identify any sections of
`
`·2· ·Exhibit 1 that you prepared the first draft of
`
`·3· ·sitting here today?
`
`·4· · · A· · It's all blended together.· So, no, I can't
`
`·5· ·identify what sections, especially after revisions,
`
`·6· ·what was first and what survived.
`
`·7· · · Q· · So there were some sections of Exhibit 1
`
`·8· ·that you didn't prepare the first draft of; is that
`
`·9· ·correct?
`
`10· · · · · ·MR. BROUGHAN:· Object to form.· Object to
`
`11· ·the extent it calls for privilege or work product
`
`12· ·protected information.
`
`13· · · · · ·Caution you not to reveal the substance of
`
`14· ·communication with counsel.· Subject to that, you
`
`15· ·can answer.
`
`16· · · · · ·THE WITNESS:· Again, I don't have a good
`
`17· ·recollection of how the process of this was put
`
`18· ·together.· There was some interactions with counsel.
`
`19· ·BY MR. LENNIE:
`
`20· · · Q· · And just to be clear, I'm not asking for
`
`21· ·any discussions that you had with counsel.· I'm
`
`22· ·asking just whether you recall whether you prepared
`
`23· ·a first draft of sections or whether someone else
`
`24· ·prepared it, so just to be clear on that.
`
`IPR2017-00626
`Ex. 2007 8/93
`
`
`
`·1· · · · · ·You signed the copy of your declaration on
`
`·2· ·February 7th, correct?
`
`·3· · · A· · Yes.
`
`·4· · · Q· · And that's of this year?
`
`·5· · · A· · Yes.· That's correct.· Yeah.
`
`·6· · · Q· · So roughly five weeks ago?
`
`·7· · · A· · Mm-hmm.
`
`·8· · · Q· · When did you start preparing Exhibit 1?
`
`·9· · · A· · I've forgotten now when the original
`
`10· ·replies from the Andrea expert Douglas came, but
`
`11· ·probably roughly around that time.
`
`12· · · Q· · And do you know how many hours you spent
`
`13· ·working on your preparation for finalizing Exhibit
`
`14· ·1?
`
`15· · · A· · No.· I'd have to check my records for that.
`
`16· · · Q· · Do you know whether it was more than a
`
`17· ·hundred hours?
`
`18· · · A· · I think it was less.
`
`19· · · Q· · Do you have any more specific recollection
`
`20· ·of the amount of fees that you charged in connection
`
`21· ·with your work in preparation for Exhibit 1?
`
`22· · · A· · Just the customary fee that I mentioned in
`
`23· ·the earlier declaration that I provided.
`
`24· · · Q· · Do you know what the aggregate amount was
`
`IPR2017-00626
`Ex. 2007 9/93
`
`
`
`·1· ·that you worked on in connection with preparing
`
`·2· ·Exhibit 1?
`
`·3· · · A· · No, sorry.· I don't.
`
`·4· · · Q· · Other than the attorneys you're working
`
`·5· ·with at Sidley & Austin, did you speak with anybody
`
`·6· ·in connection with preparing Exhibit 1?
`
`·7· · · A· · No.
`
`·8· · · Q· · Other than the documents that have been
`
`·9· ·marked as exhibits in the IPR proceeding 2017-626,
`
`10· ·did you consider any other documents in connection
`
`11· ·with your preparation of Exhibit 1?
`
`12· · · A· · I considered the documents that's listed in
`
`13· ·my references.
`
`14· · · Q· · What did you do to prepare for today's
`
`15· ·deposition, if anything?
`
`16· · · · · ·MR. BROUGHAN:· Caution you not to reveal
`
`17· ·the substance of communications with counsel, but
`
`18· ·subject to that, you can answer.
`
`19· · · · · ·THE WITNESS:· We met yesterday to go over
`
`20· ·my declaration and the references.
`
`21· ·BY MR. LENNIE:
`
`22· · · Q· · And how long did you meet with counsel in
`
`23· ·connection with preparing for today's deposition?
`
`24· · · A· · Most of the day I guess.
`
`IPR2017-00626
`Ex. 2007 10/93
`
`
`
`·1· · · Q· · Did you do any other work besides your work
`
`·2· ·yesterday in connection with preparing for today's
`
`·3· ·deposition?
`
`·4· · · A· · Some brief review of the material before I
`
`·5· ·got here, yes.
`
`·6· · · Q· · What materials did you review in connection
`
`·7· ·with the preparation for today's deposition?
`
`·8· · · A· · Again, the declaration and the references
`
`·9· ·contained therein.
`
`10· · · Q· · Did you review any other documents besides
`
`11· ·the declaration and the prior art references?
`
`12· · · A· · Well, I might have glanced at some of the
`
`13· ·other material of the -- pertaining to the earlier
`
`14· ·declaration but --
`
`15· · · Q· · When you say the earlier declaration, which
`
`16· ·declaration are you referring to?
`
`17· · · A· · I guess it's related to the 626
`
`18· ·declaration, the one that was -- the one that this
`
`19· ·is ultimately the reply to.
`
`20· · · Q· · Are you referring to Dr. Douglas's
`
`21· ·declaration submitted in the 626 IPR proceeding?
`
`22· · · A· · Yes, and also the prior -- my prior
`
`23· ·declaration before Douglas.
`
`24· · · Q· · Okay.· Did you review the deposition
`
`IPR2017-00626
`Ex. 2007 11/93
`
`
`
`·1· ·transcript from Dr. Douglas in the 626 IPR
`
`·2· ·proceeding?
`
`·3· · · A· · Briefly, yes.
`
`·4· · · Q· · Which portions of the Douglas' deposition
`
`·5· ·transcript did you review, if you recall?
`
`·6· · · · · ·MR. BROUGHAN:· I'll object to the extent it
`
`·7· ·calls for privileged or work product protected
`
`·8· ·information.· To the extent you selected the
`
`·9· ·passages to review, you can answer.
`
`10· · · · · ·THE WITNESS:· No, I don't -- I don't
`
`11· ·recall.
`
`12· ·BY MR. LENNIE:
`
`13· · · Q· · Did you review the entirety of
`
`14· ·Dr. Douglas's deposition transcript in the 626
`
`15· ·proceeding?
`
`16· · · A· · I probably flipped through it, yes.
`
`17· · · Q· · But you didn't read the entire deposition
`
`18· ·transcript; is that correct?
`
`19· · · · · ·MR. BROUGHAN:· Objection to form.
`
`20· · · · · ·THE WITNESS:· I'm guessing I did, but,
`
`21· ·again, how deeply or how carefully I looked at
`
`22· ·various aspects would have been some deeper than
`
`23· ·others.
`
`24· ·BY MR. LENNIE:
`
`IPR2017-00626
`Ex. 2007 12/93
`
`
`
`·1· · · Q· · And when did you review Dr. Douglas's
`
`·2· ·deposition transcript in connection with your
`
`·3· ·preparation for today's deposition?
`
`·4· · · A· · Whenever it became available.· I'm not sure
`
`·5· ·the dates on that.
`
`·6· · · Q· · Did you review Dr. Douglas's deposition
`
`·7· ·transcript yesterday?
`
`·8· · · A· · Yes, briefly.
`
`·9· · · Q· · Do you know how much time you spent
`
`10· ·preparing for today's deposition in the entirety?
`
`11· · · A· · There was the day yesterday I spent here
`
`12· ·and some period of time in advance of coming here.
`
`13· ·I don't know how many hours.
`
`14· · · Q· · Can you make a ballpark estimate of the
`
`15· ·number of hours that you spent preparing for today's
`
`16· ·deposition?
`
`17· · · A· · Maybe on the order of six hours or so.
`
`18· · · Q· · How long did you meet with counsel
`
`19· ·yesterday?
`
`20· · · A· · Most of the day.
`
`21· · · Q· · I mean, the day is 24 hours.· Can you
`
`22· ·narrow it down a little bit more on the total number
`
`23· ·of hours?
`
`24· · · A· · Oh, sure.· Sure.· Probably around 10
`
`IPR2017-00626
`Ex. 2007 13/93
`
`
`
`·1· ·o'clock in the morning till about 4:00 in the
`
`·2· ·afternoon, somewhere around there.· I meant most of
`
`·3· ·the working day.
`
`·4· · · Q· · And did you review your signed declaration
`
`·5· ·that we've marked as Exhibit 1 in connection with
`
`·6· ·your preparation for today's deposition?
`
`·7· · · A· · Yes.
`
`·8· · · Q· · And in connection with your review of
`
`·9· ·Exhibit 1, did you identify any inaccuracies or
`
`10· ·typographical errors in your declaration?
`
`11· · · A· · No.· I remember seeing the name of Hirsch
`
`12· ·was misspelled once or twice, but nothing else.
`
`13· · · Q· · No substantive errors in the testimony that
`
`14· ·you provided in Exhibit 1?
`
`15· · · A· · That's correct.
`
`16· · · Q· · In Exhibit 1 you provide an opinion that
`
`17· ·the Martin 1993 prior art reference discloses that
`
`18· ·sub-windows are an optional feature of Martin's
`
`19· ·algorithm; is that correct?
`
`20· · · A· · Yes.
`
`21· · · · · ·MR. LENNIE:· Let me introduce as Exhibit 2
`
`22· ·a copy of the Martin 1993 reference.
`
`23· · · · · · · · ·(Deposition Exhibit 2 was marked for
`
`24· ·identification.)
`
`IPR2017-00626
`Ex. 2007 14/93
`
`
`
`·1· ·BY MR. LENNIE:
`
`·2· · · Q· · And you're familiar with what we've marked
`
`·3· ·as Exhibit 2, the Martin prior art reference
`
`·4· ·entitled "An Efficient Algorithm to Estimate the
`
`·5· ·Instantaneous SNR of Speech Signals," correct?
`
`·6· · · A· · Yes.
`
`·7· · · Q· · Where does the Martin reference state that
`
`·8· ·sub-windows are an optional feature of the algorithm
`
`·9· ·described in this reference?
`
`10· · · A· · I don't think he says it exactly in those
`
`11· ·words, but he provides the parameter W, which he
`
`12· ·calls windows, but for our purposes we're calling
`
`13· ·them sub-windows.· That's a variable at the user's
`
`14· ·discretion, and setting W to any value seems to be
`
`15· ·very much acceptable to Martin, especially, he
`
`16· ·doesn't exclude W equal to one, which would make it
`
`17· ·one sub-window.· So to that extent it's an optional
`
`18· ·feature.
`
`19· · · Q· · So Martin's algorithm describes the use of
`
`20· ·sub-windows, correct?
`
`21· · · A· · He calls them windows, W, but we're calling
`
`22· ·them sub-windows, yes.· I assume that's what you
`
`23· ·mean.
`
`24· · · Q· · Correct.· W is the number of sub-windows;
`
`IPR2017-00626
`Ex. 2007 15/93
`
`
`
`·1· ·whereas, L is the length of the larger full window.
`
`·2· ·Is that correct?
`
`·3· · · A· · Right.· Yes.
`
`·4· · · Q· · So when we talk about W windows, you
`
`·5· ·understand that to be sub-windows; whereas, L refers
`
`·6· ·to the broader window.· Is that correct?
`
`·7· · · A· · Yes.· That's correct.
`
`·8· · · Q· · The Martin reference that we've marked as
`
`·9· ·Exhibit 2 doesn't expressly state that sub-windows
`
`10· ·are an optional feature of his algorithm; is that
`
`11· ·correct?
`
`12· · · A· · He says that -- he makes comments on what
`
`13· ·the overall window length L should be, but makes --
`
`14· ·is pretty much silent on the choice of how W should
`
`15· ·be made.· So whether it's W equal one or any other
`
`16· ·value seems to be within the scope of what he
`
`17· ·intended.
`
`18· · · Q· · So just to be clear on the question, so the
`
`19· ·Martin reference that we've marked as Exhibit 2
`
`20· ·doesn't state that sub-windows are an optional
`
`21· ·feature of his algorithm; is that correct?
`
`22· · · · · ·MR. BROUGHAN:· Object to form.
`
`23· · · · · ·THE WITNESS:· He's silent on the matter.
`
`24· ·He doesn't say it's optional, but he also doesn't
`
`IPR2017-00626
`Ex. 2007 16/93
`
`
`
`·1· ·say it's mandatory.
`
`·2· ·BY MR. LENNIE:
`
`·3· · · Q· · Now, if W is set to equal one -- let me
`
`·4· ·strike that.· Let me back up.
`
`·5· · · · · ·W is a parameter that's utilized in
`
`·6· ·Martin's algorithm, correct?
`
`·7· · · A· · Yes.
`
`·8· · · Q· · And the value of W will correspond to the
`
`·9· ·number of W windows or what we've been referring to
`
`10· ·as sub-windows, correct?
`
`11· · · A· · Yes.
`
`12· · · Q· · Where does Martin state that W can be set
`
`13· ·equal to one?
`
`14· · · A· · Again, he's silent on the matter.· He
`
`15· ·doesn't say what constraints there are on W other
`
`16· ·than you would infer from his use that W is a
`
`17· ·positive integer whether one, two, three, four,
`
`18· ·five.· There's no real guidelines.
`
`19· · · Q· · Does Martin provide an example where W is
`
`20· ·set to a certain value in the algorithm -- I mean,
`
`21· ·in the algorithm in his paper?
`
`22· · · · · ·MR. BROUGHAN:· Object to form.
`
`23· · · · · ·THE WITNESS:· So his flowchart of the
`
`24· ·algorithm, for example, has no comment on W.· He
`
`IPR2017-00626
`Ex. 2007 17/93
`
`
`
`·1· ·does provide an example where the algorithm is used
`
`·2· ·where W is equal to four.
`
`·3· ·BY MR. LENNIE:
`
`·4· · · Q· · Is that the only example that the Martin
`
`·5· ·reference provides in terms of how the algorithm
`
`·6· ·works?
`
`·7· · · · · ·MR. BROUGHAN:· Object to form.
`
`·8· · · · · ·THE WITNESS:· He -- well, he has many
`
`·9· ·results in the -- in his paper.· What values of W he
`
`10· ·chose, again, he's silent on except for that one
`
`11· ·example where he mentions W equal to four.
`
`12· ·BY MR. LENNIE:
`
`13· · · Q· · So the paper that we've marked as Exhibit 2
`
`14· ·doesn't provide any express examples where W is set
`
`15· ·to equal one; is that correct?
`
`16· · · A· · I believe -- I believe, again, that he is
`
`17· ·silent on what values of W he chose for all these
`
`18· ·various examples except that one example he gives
`
`19· ·where W is equal to four.
`
`20· · · Q· · So I'm just looking for a yes-or-no answer.
`
`21· ·Does the Martin reference that we've marked as
`
`22· ·Exhibit 2 provide any examples where W is set equal
`
`23· ·to one?
`
`24· · · · · ·MR. BROUGHAN:· Objection, asked and
`
`IPR2017-00626
`Ex. 2007 18/93
`
`
`
`·1· ·answered.
`
`·2· · · · · ·THE WITNESS:· Again, I don't know what
`
`·3· ·values of W he chose for all his other examples.· He
`
`·4· ·does not specify W equal to one on those examples,
`
`·5· ·but we can't exclude that possibility.
`
`·6· ·BY MR. LENNIE:
`
`·7· · · Q· · So is it your testimony that W can be set
`
`·8· ·to any value?
`
`·9· · · · · ·MR. BROUGHAN:· Object to form.
`
`10· · · · · ·THE WITNESS:· Again, he doesn't constrain
`
`11· ·it, and, yes, any integer value seems to be within
`
`12· ·the scope of what he intended, positive integer.
`
`13· ·BY MR. LENNIE:
`
`14· · · Q· · So W has to be an integer.· Is that your
`
`15· ·testimony?
`
`16· · · A· · Positive integer, yes.
`
`17· · · Q· · And it has to be positive?
`
`18· · · A· · Yes.
`
`19· · · Q· · And why does it have to be positive?
`
`20· · · A· · Well, otherwise it doesn't make -- it
`
`21· ·doesn't fit into the algorithm to have M times W
`
`22· ·equal to L where we know that L is positive and M is
`
`23· ·positive.· W should be positive as well.
`
`24· · · Q· · Of course the Martin reference doesn't say
`
`IPR2017-00626
`Ex. 2007 19/93
`
`
`
`·1· ·that W needs to be positive, does it?
`
`·2· · · A· · By having that formula, he does, yes.
`
`·3· · · Q· · Why does W need to be set to an integer?
`
`·4· · · A· · Again, the way he uses W, it's used as an
`
`·5· ·index into an array for example.· It only makes
`
`·6· ·sense in the context of that that W should be a
`
`·7· ·positive integer.· It's in the way he's using it.
`
`·8· · · Q· · You're familiar with figure 2 from the
`
`·9· ·Martin reference that we've marked as Exhibit 2?
`
`10· · · A· · Yes.
`
`11· · · Q· · It's a little difficult to read in the
`
`12· ·actual Martin reference.· So let me mark as Exhibit
`
`13· ·3 just a blown-up copy of figure 2 from the Martin
`
`14· ·reference.
`
`15· · · · · · · · ·(Deposition Exhibit 3 was marked for
`
`16· ·identification.)
`
`17· ·BY MR. LENNIE:
`
`18· · · Q· · You want to just take a minute to confirm,
`
`19· ·if you can, whether what we've marked as Exhibit 3
`
`20· ·is an accurate depiction of figure 2 from the Martin
`
`21· ·reference that we've marked as Exhibit 2.
`
`22· · · A· · Yes, it appears to be.
`
`23· · · Q· · So looking at figure 2 from Martin that
`
`24· ·we've marked as Exhibit 3 to today's deposition, you
`
`IPR2017-00626
`Ex. 2007 20/93
`
`
`
`·1· ·see the third diamond-shaped box in figure 2.· Do
`
`·2· ·you see that, the third one down from the top?
`
`·3· · · A· · Yes.· Monotonically increasing power?
`
`·4· · · Q· · Yeah.· So it's the diamond-shape box that
`
`·5· ·indicates inside the box:· Monotonically increasing
`
`·6· ·power?
`
`·7· · · · · ·Do you see that?
`
`·8· · · A· · Yes, I do.
`
`·9· · · Q· · And you agree with respect to the Martin
`
`10· ·algorithm this box is referring to whether or not
`
`11· ·the signal power is monotonically increasing within
`
`12· ·a given sub-window; is that correct?
`
`13· · · A· · Signal power?· I guess I would need to
`
`14· ·understand what the definition of signal power is.
`
`15· · · Q· · Well, the box refers to monotonically
`
`16· ·increasing power, correct?
`
`17· · · A· · Right.
`
`18· · · Q· · Do you understand that to refer to whether
`
`19· ·or not the signal power is monotonically increasing?
`
`20· · · A· · I understand that to refer to these
`
`21· ·variables inside this vector, these various PMmin
`
`22· ·that he's storing in this vector to be increasing or
`
`23· ·not.
`
`24· · · Q· · The PMmin value that we're referring to,
`
`IPR2017-00626
`Ex. 2007 21/93
`
`
`
`·1· ·does that correspond to the power of a signal that's
`
`·2· ·being analyzed by the algorithm?
`
`·3· · · A· · Well, so there's verbiage, again, that
`
`·4· ·Martin uses for that PMmin.· I'm trying to recall
`
`·5· ·what the words are he uses for that.· The power
`
`·6· ·minimum of the last M samples.
`
`·7· · · Q· · So what is your understanding as to what
`
`·8· ·that diamond-shaped box, monotonically increasing
`
`·9· ·power, is referring to?
`
`10· · · A· · It appears to be examining the elements of
`
`11· ·that vector, min_vec, and deciding whether or not
`
`12· ·those values are increasing monotonically.
`
`13· · · Q· · Okay.· And are those min_vec values that
`
`14· ·are being compared to determine whether or not
`
`15· ·they're monotonically increasing, are those for a
`
`16· ·given sub-window in the Martin algorithm?
`
`17· · · · · ·MR. BROUGHAN:· Object to form.
`
`18· · · · · ·THE WITNESS:· So the min_vec is -- each
`
`19· ·value in min_vec is a Pmin -- or PMmin, and the number
`
`20· ·of elements in the min_vec is W.
`
`21· ·BY MR. LENNIE:
`
`22· · · Q· · And W is what we've been referring to as a
`
`23· ·sub-window?
`
`24· · · A· · Yes.· Number of sub-windows, yes.
`
`IPR2017-00626
`Ex. 2007 22/93
`
`
`
`·1· · · Q· · So the min_vec values that are being
`
`·2· ·compared to determine whether or not they're
`
`·3· ·monotonically increasing are the min_vec values for
`
`·4· ·W, correct?
`
`·5· · · · · ·MR. BROUGHAN:· Object to form.
`
`·6· · · · · ·THE WITNESS:· Let me try to clarify.· So
`
`·7· ·the -- again, the min_vec vector has W values in it,
`
`·8· ·each one corresponding to a PMmin value, and each
`
`·9· ·PMmin value corresponds to one sub-window.
`
`10· ·BY MR. LENNIE:
`
`11· · · Q· · So if W is equal to one, how many min_vec
`
`12· ·values are in that vector?
`
`13· · · A· · One.
`
`14· · · Q· · And if W is equal to one and you have one
`
`15· ·min_vec value, can the algorithm determine whether
`
`16· ·or not the min_vec values are monotonically
`
`17· ·increasing in power?
`
`18· · · A· · So, again, the whole notion of whether or
`
`19· ·not monotonically increasing in power is if you --
`
`20· ·it becomes a trivial case when you have a vector of
`
`21· ·just one value that you can assert that it is or it
`
`22· ·isn't monotonically increasing, but, again, it's not
`
`23· ·excluded from the algorithm.
`
`24· · · Q· · Where W is equal to one, is the algorithm
`
`IPR2017-00626
`Ex. 2007 23/93
`
`
`
`·1· ·able to definitively determine whether or not the
`
`·2· ·values of min_vec for that W window are
`
`·3· ·monotonically increasing in power?
`
`·4· · · · · ·MR. BROUGHAN:· Object to form.
`
`·5· · · · · ·THE WITNESS:· Again, if there's just one
`
`·6· ·element, the issue of monotonically increasing is
`
`·7· ·trivially answered either yes or no, and it doesn't
`
`·8· ·matter.
`
`·9· ·BY MR. LENNIE:
`
`10· · · Q· · What do you mean by trivially answered yes
`
`11· ·or no?
`
`12· · · A· · Well, when you have a minimum of one
`
`13· ·element, then the value of that minimum is the
`
`14· ·element itself.· So the question of the
`
`15· ·monotonically increasing in power can be answered
`
`16· ·yes or no.· The algorithm works.· Steps of the
`
`17· ·algorithm follow through just fine.
`
`18· · · Q· · So you have one value in W, is that
`
`19· ·correct, for min_vec when W is equal to one?
`
`20· · · A· · You have one min_vec -- so min_vec contains
`
`21· ·one value when W is equal to one, yes.
`
`22· · · Q· · And with that one value, the algorithm
`
`23· ·wouldn't be able to determine whether or not this
`
`24· ·value is monotonically increasing or not; is that
`
`IPR2017-00626
`Ex. 2007 24/93
`
`
`
`·1· ·correct?
`
`·2· · · · · ·MR. BROUGHAN:· Object to form.
`
`·3· · · · · ·THE WITNESS:· Again, there is no notion of
`
`·4· ·monotonically increasing when you have just one
`
`·5· ·value.
`
`·6· ·BY MR. LENNIE:
`
`·7· · · Q· · So if W is equal to one and you only have
`
`·8· ·one min_vec value for that W sub-window, what would
`
`·9· ·the algorithm do when it's faced with the decision
`
`10· ·block question monotonically increasing power?
`
`11· · · A· · It would send Pn(i) equal to PMmin.
`
`12· · · Q· · And that's following the yes decision path
`
`13· ·where Pn(i) is equal to PMmin; is that correct?
`
`14· · · A· · Yes.· If it wanted to follow no path, the
`
`15· ·same results hold, that Pn(i) is equal to PMmin.
`
`16· · · Q· · But isn't that the purpose of the diamond
`
`17· ·decision block, to determine whether the answer to
`
`18· ·that question monotonically increasing power is yes
`
`19· ·or no?
`
`20· · · · · ·MR. BROUGHAN:· Object to form.
`
`21· · · · · ·THE WITNESS:· Again, by having set W equal
`
`22· ·to one, you are -- the person using the algorithm
`
`23· ·has concluded that the issue of yes or no in that
`
`24· ·block is not material.· The result is the same in
`
`IPR2017-00626
`Ex. 2007 25/93
`
`
`
`·1· ·either case.
`
`·2· ·BY MR. LENNIE:
`
`·3· · · Q· · So when W is equal to one, it makes no
`
`·4· ·difference whether you go down the yes or no path in
`
`·5· ·connection with the decision tree in the
`
`·6· ·monotonically increasing power decision block; is
`
`·7· ·that correct?
`
`·8· · · A· · That's correct.
`
`·9· · · Q· · So is it your opinion that the
`
`10· ·determination in this decision block of
`
`11· ·monotonically increasing power that we've been
`
`12· ·discussing within a given sub-window is an optional
`
`13· ·step of Martin's algorithm?
`
`14· · · A· · No.· I didn't say that's an optional step.
`
`15· ·You leave it in.· It's just the result of it becomes
`
`16· ·immaterial to the performance of the algorithm when
`
`17· ·W is equal to one.· It behaves the same whether you
`
`18· ·choose yes or no in that particular block.
`
`19· · · Q· · When W is equal to one, the algorithm can't
`
`20· ·make a determination of whether or not the signal
`
`21· ·within the W window is monotonically increasing in
`
`22· ·power; is that correct?
`
`23· · · · · ·MR. BROUGHAN:· Object to form.
`
`24· · · · · ·THE WITNESS:· Again, by choosing W equal to
`
`IPR2017-00626
`Ex. 2007 26/93
`
`
`
`·1· ·one, the user has decided that the outcome of that
`
`·2· ·particular block is not material to the performance
`
`·3· ·of the algorithm.
`
`·4· ·BY MR. LENNIE:
`
`·5· · · Q· · And by making that decision that it's not
`
`·6· ·material, the user has effectively read out that
`
`·7· ·decision block from the algorithm; isn't that
`
`·8· ·correct?
`
`·9· · · · · ·MR. BROUGHAN:· Object to form.
`
`10· · · · · ·THE WITNESS:· No.· The result of whether
`
`11· ·you go yes or no is the same.· That's the only
`
`12· ·conclusion.
`
`13· ·BY MR. LENNIE:
`
`14· · · Q· · But when you have W equal to one, you can't
`
`15· ·make a determination of whether it goes yes or no;
`
`16· ·isn't that correct?
`
`17· · · A· · The point is that it doesn't matter.· So
`
`18· ·whether you can or can't is not material as much as
`
`19· ·it doesn't matter because the algorithm does the
`
`20· ·same thing either way.
`
`21· · · Q· · So setting aside the issue of whether it
`
`22· ·does or doesn't matter, you'll agree with me that
`
`23· ·when W is equal to one, the algorithm cannot
`
`24· ·determine whether or not the signal is monotonically
`
`IPR2017-00626
`Ex. 2007 27/93
`
`
`
`·1· ·increasing in power for that W window; is that
`
`·2· ·correct?
`
`·3· · · · · ·MR. BROUGHAN:· Object to form.
`
`·4· · · · · ·THE WITNESS:· Again, the question becomes
`
`·5· ·an empty question because if I have a list of one
`
`·6· ·item, the notion of whether that one item is
`
`·7· ·monotonically increasing or not becomes a trivial
`
`·8· ·question.· Then you can answer yes or no, and either
`
`·9· ·way you do the same thing in this algorithm.
`
`10· ·BY MR. LENNIE:
`
`11· · · Q· · Well, it's not a trivial question.· The
`
`12· ·issue is whether or not when you have one value in
`
`13· ·the W window you're unable to determine whether or
`
`14· ·not the signal for that window -- that's window W --
`
`15· ·is monotonically increasing in power or not; is that
`
`16· ·correct?
`
`17· · · · · ·MR. BROUGHAN:· Object to form.
`
`18· · · · · ·THE WITNESS:· Again, I don't think
`
`19· ·characterizing it as unable as much as you've
`
`20· ·decided that it's not material would be my
`
`21· ·characterization of the block.
`
`22· ·BY MR. LENNIE:
`
`23· · · Q· · And by setting W equal to one and making
`
`24· ·the determination that this decision block is not
`
`IPR2017-00626
`Ex. 2007 28/93
`
`
`
`·1· ·material, you're effectively reading this decision
`
`·2· ·block out of the algorithm; isn't that correct?
`
`·3· · · · · ·MR. BROUGHAN:· Object to form.
`
`·4· · · · · ·THE WITNESS:· Again, I don't know how to
`
`·5· ·just pull a decision block out.· I just can follow
`
`·6· ·the algorithm for this case of W equals to one,
`
`·7· ·which he does not exclude, and see what it does, and
`
`·8· ·I don't want to change the algorithm or affect it in
`
`·9· ·any material way.· I'm just applying particular
`
`10· ·values of W and seeing what it does.
`
`11· ·BY MR. LENNIE:
`
`12· · · Q· · All right.· W is equal to one in your
`
`13· ·hypothetical, right, in a hypothetical that you
`
`14· ·think fits within the Martin algorithm; is that
`
`15· ·correct?
`
`16· · · A· · That's one choice possible, yes.
`
`17· · · Q· · According to your testimony?
`
`18· · · A· · Right.
`
`19· · · Q· · And if W is equal to one, you have one
`
`20· ·min_vec value for that W window, correct?
`
`21· · · A· · Yes.
`
`22· · · Q· · When the algorithm comes down to the
`
`23· ·decision step of monotonically increasing in power
`
`24· ·or not, can the algorithm determine whether or not
`
`IPR2017-00626
`Ex. 2007 29/93
`
`
`
`·1· ·the output of that answer is yes or no?
`
`·2· · · A· · Again, that issue is not material in the
`
`·3· ·sense that if you have a list of one item, whether
`
`·4· ·that item has -- you can call that item increasing
`
`·5· ·or decreasing and it doesn't matter, and the
`
`·6· ·algorithm shows that it does the same thing in
`
`·7· ·either case too.· So whether or not you want to
`
`·8· ·answer that question yes or no, the result doesn't
`
`·9· ·matter.
`
`10· · · Q· · So you said when you have one value for
`
`11· ·min_vec for the window W, it doesn't matter whether
`
`12· ·it's increasing or decreasing.· Is that what you
`
`13· ·said?
`
`14· · · A· · Well, by choosing W equal to one, you've
`
`15· ·decided that the list you're looking at has only one
`
`16· ·item in it, and so, therefore, the issue of whether
`
`17· ·it's monotonically increasing in power or not --
`
`18· ·whether that item is monotonically increasing or
`
`19· ·not, that one item is, you've decided that that's
`
`20· ·not a material issue, and so the outcome won't be
`
`21· ·affected by your decision, whether it's yes or no.
`
`22· · · Q· · But with one value you can't make a
`
`23· ·determination whether it's increasing or decreasing;
`
`24· ·is that correct?
`
`IPR2017-00626
`Ex. 2007 30/93
`
`
`
`·1· · · · · ·MR. BROUGHAN:· Objection to form.
`
`·2· · · · · ·THE WITNESS:· Again, it doesn't matter.
`
`·3· ·Because whether you've decided that's yes or no, the
`
`·4· ·algorithm behaves the same w