`By: Peter J. McAndrews
`McAndrews, Held & Malloy, Ltd.
`500 W. Madison St., 34th Floor
`Chicago, IL 60661
`Tel: 312-775-8000
`Fax: 312-775-8100
`E-mail: pmcandrews@mcandrews-ip.com
`
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`_____________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`_____________
`
`
`CISCO SYSTEMS, INC., DISH NETWORK, LLC,
`COMCAST CABLE COMMUNICATIONS, LLC,
`COX COMMUNICATIONS, INC.,
`TIME WARNER CABLE ENTERPRISES LLC,
`VERIZON SERVICES CORP., and ARRIS GROUP, INC.,
`Petitioner,
`
`v.
`
`TQ DELTA, LLC
`Patent Owner
`_____________
`
`Case No. IPR2016-010211
`Patent No. 8,718,158
`_____________
`
`PATENT OWNER MOTION FOR OBSERVATION REGARDING
`CROSS-EXAMINATION OF DR. JOSE TELLADO
`
`1 DISH Network, L.L.C., who filed a Petition in IPR2017-00255, and Comcast
`Cable Communications, L.L.C., Cox Communications, Inc., Time Warner Cable
`Enterprises L.L.C., Verizon Services Corp., and ARRIS Group, Inc., who filed a
`Petition in IPR2017-00417, have been joined in this proceeding.
`
`
`
`Patent Owner Motion For Observation
`IPR2016-01021
`
`
`Patent Owner moves for observation regarding the cross-examination of Dr.
`
`Jose Tellado, the reply declarant of Petitioner (transcript filed as Ex. 2013).
`
`Observation #1
`
`In Ex. 2013, at 43:24–44:20, Dr. Tellado testified:
`
`Q. The noise profiles that you reference in your paragraphs 9 through 13,
`you didn't actually apply those in selecting your 182 random carriers and
`52 Shively carriers, did you?
`
`A. ADSL transceivers have to work over many combinations of loops,
`gauges, crosstalk attenuation. I didn't go through all the combinations. I just
`picked one combination to justify my simulation.
`
`I only need to find one example to justify a simulation. There is many
`other combinations.
`
`Q. And do you believe that the one 12,000-foot attenuation curve that you
`picked justifies your selection of those carriers?
`
`A. So my simulation shows that 182 QAM random carriers, 52 structured
`carriers has high PAR implementing Shively’s techniques.
`
`To come up with 182 and the 52, I need to justify where I get the 182
`and 52, and I could have done it through combinations of loss and
`crosstalks. There is infinite number of combination. I pick one to justify it,
`and then it applies more generally.
`
`That testimony is relevant to Dr. Tellado’s second declaration, Ex. CSCO-
`
`1026, at ¶¶ 7-14 (pp. 3-9), where he discusses a number of noise profiles and states
`
`that “Shively states that bit-spreading is a way to ‘compensate for high attenuation
`
`and/or high noise’” (emphasis in original) and “[s]ince noise can occur on a line
`
`of any length, a POSITA would not have considered Shively’s bit-spreading
`
`technique to be limited to being used on only long lines.” The testimony is
`
`
`
`1
`
`
`
`Patent Owner Motion For Observation
`IPR2016-01021
`
`
`relevant because it shows that Dr. Tellado “just picked one combination to justify
`
`[his] simulation,” and that simulation was not based the long lines with high
`
`attenuation and noise addressed by Shively, but was instead based on a line of
`
`12,000 feet that is not a long line and does not have high attenuation or high noise.
`
`Observation #2
`
`In Ex. 2013, at 45:23–46:5, Dr. Tellado testified:
`
`Q. Do you know which attenuation and noise characteristics Dr. Short
`relied on in choosing his random, Shively and unusable carriers?
`
`A. I believe Dr. Short used Figure 6 or variants of it, and he selected one
`of these high-attenuation loops, and I believe he used the thin-gauge, high-
`loss AWG26 that's marked 18,000 in this figure.
`
`That testimony is relevant to (1) Dr. Tellado’s statement, in Ex. CSCO-1026
`
`at ¶ 7 (p. 4), that Shively (Ex. CSCO-1011) “describes using its bit spreading
`
`technique ‘to compensate for high attenuation and/or high noise in those parts of
`
`the communication channel frequency band that would otherwise not be usable due
`
`to noise and attenuation effects’” and (2) to the teaching in Shively, Ex. CSCO-
`
`1011, at 9:65–10:1, that “[i]n long loop systems where cable 3 is of length of the
`
`order 18,000 feet or more, high signal attenuation at higher frequencies (greater
`
`than 500 kHz) is usually observed[]” and, at 11:11-12 that “[s]uch noisy and/or
`
`highly attenuated sub-bands can occur for example in long-run twisted pair
`
`conductors.” The testimony is relevant because it shows that the opinions of
`
`
`
`2
`
`
`
`Patent Owner Motion For Observation
`IPR2016-01021
`
`
`Patent Owner’s expert, Dr. Short, are based on the long loops with high attenuation
`
`and noise to which Shively’s teachings are directed.
`
`Observation #3
`
`In Ex. 2013, at 46:10–47:20, Dr. Tellado testified:
`
`Q. Did you run a simulation --
`
`A. Yes.
`
`Q. -- using an 18,000-foot loop with the attenuation characteristics shown
`in Figure 6 on page 18 of your declaration?
`
`A. So in this AWG26 loop of 18,000 feet, I did a quick estimate.
`
`Q. But you didn't run a full simulation on it?
`
`A. Not a full simulation.
`
`***
`
`Q. What did you determine from your quick estimate?
`
`A. That Dr. Short’s approximation of a Gaussian approximation was poor.
`It was worse than -- than Dr. Short said.
`
`Q. How much worse?
`
`A. I don't recall. It was significantly worse.
`
`Q. You don't recall. Did you run a simulation?
`
`A. I said I ran a quick estimate to see if the Gaussian approximation was
`good, and it was not.
`
`Q. How did you do that quick estimation?
`
`A. Using similar techniques to the ones that I provided.
`
`Q. Where is that simulation?
`
`
`
`3
`
`
`
`Patent Owner Motion For Observation
`IPR2016-01021
`
`A. So when I started working the declaration, I just did a quick estimate to
`see if the Gaussian approximation was correct, and I determined it was not.
`
`
`
`Q. Did you use MATLAB for that?
`
`A. Yes.
`
`Q. Where are the results of that MATLAB?
`
`A. I don't have them.
`
`That testimony is relevant to the credibility of, sufficiency of, and factual
`
`basis for Dr. Tellado’s statement, in Ex. CSCO-1026 at ¶ 29 (p. 17), that “Dr.
`
`Short’s analysis is flawed ….” The testimony is relevant because (1) it shows that,
`
`in evaluating Dr. Short’s analysis, Dr. Tellado did not run a “full simulation,” he
`
`does not “recall” the results, and he no longer has the results, and (2) that
`
`simulation would show whether Dr. Short’s analysis is flawed.
`
`Observation #4
`
`In Ex. 2013, at 50:6–56:17, Dr. Tellado testified:
`
`Q. Are you suggesting that Dr. Short’s -- if you had run a full simulation on
`Dr. Short’s 18,000-foot loop, assuming the 88 usable carriers and 16
`Shively carriers and the remainder unusable, are you telling me that that
`would be worse than your Scenario 1 here?
`
`A. I didn't say that. I just said it was diverging relative to a Gaussian
`process.
`
`***
`
`Q. If we were to run a Gaussian on the 104 carriers that you ran this quick
`simulation on, where would that line show up in graph 2 on your -- page 30
`of your declaration?
`
`
`
`4
`
`
`
`Patent Owner Motion For Observation
`IPR2016-01021
`
`A. The Gaussian process?
`
`Q. Mm-hmm.
`
`
`
`A. The Gaussian process would be proportional to the power of 104
`carriers.
`
`Q. Okay. So would it be to the left of the Gaussian process for the 250
`carriers that you used for –
`
`A. Yes.
`
`Q. -- graph 2?
`
`A. Uh-huh. Yes.
`
`Q. And the simulation -- so using your simulation and your MATLAB
`script on 88 random carriers and 16 Shively carriers, where would that
`show up on graph 2 on page 30 of your declaration?
`
`A. The Gaussian approximation or the actual simulation?
`
`Q. The actual simulation.
`
`A. The actual simulation, I don't remember where it crossed over.
`
`Q. Well, would it be to the left of the Scenario 1 line that you show here in
`graph 2?
`
`A. I don’t recall.
`
`***
`
`Q. Based on your expertise, do you believe it would have crossed the
`clipping threshold to the left of your Scenario 1 in graph 2?
`
`A. I don't want to guess.
`
`That testimony is relevant to (1) the credibility of, sufficiency of, and factual
`
`basis for Dr. Tellado’s statement, in Ex. CSCO-1026 at ¶ 29 (p. 17), that “Dr.
`
`Short’s analysis is flawed …” and (2) the accuracy of Dr. Short’s statement, in Ex.
`
`
`
`5
`
`
`
`Patent Owner Motion For Observation
`IPR2016-01021
`
`
`2003 at ¶ 63 (p. 32), that “[w]hile Shively’s ‘spreading’ idea will cause a small
`
`uptick in clipping probability, any increase is negated many times over by the
`
`enormous reduction in clipping achieved by reducing signal power by more than
`
`half. Based on worst-case assumptions regarding Shively’s spreading idea, the
`
`clipping probability for both normal and power-boost modes is virtually zero.”
`
`The testimony is relevant because it shows that Dr. Tellado does not know whether
`
`Dr. Short’s conclusion is flawed.
`
`Observation #5
`
`In Ex. 2013, at 58:10–60:24, Dr. Tellado testified:
`
`Q. So Dr. Tellado, did you save, in any form, your MATLAB simulation
`script for the 18,000-foot loop scenario?
`
`MR. McDOLE: Objection; asked and answered.
`
`THE WITNESS: I don't recall.
`
`***
`
`Q. Did you share with anyone, including your lawyers, a copy of your
`MATLAB simulation script for 88 usable carriers and 16 Shively carriers?
`
`A. I don't recall.
`
`***
`
`Q. Did you save, in any form, the output of your MATLAB simulation for
`the 88 random carriers and 16 Shively carriers?
`
`A. I don't recall.
`
`Q. Do you currently have in your possession, in any form, the output of
`your MATLAB simulation for the 88 usable carriers and 16 Shively
`carriers?
`
`
`
`6
`
`
`
`Patent Owner Motion For Observation
`IPR2016-01021
`
`A. I don't recall.
`
`
`
`That testimony is relevant to the credibility of, sufficiency of, and factual
`
`basis for Dr. Tellado’s statement, in Ex. CSCO-1026 at ¶ 29 (p. 17), that “Dr.
`
`Short’s analysis is flawed ….” The testimony is relevant because it shows that Dr.
`
`Tellado used a MATLAB simulation to evaluate the 18,000 foot loop scenario that
`
`Dr. Short relied on, but Dr. Tellado “does not recall” whether any record of that
`
`simulation still exists, and that simulation would show whether Dr. Short’s analysis
`
`is flawed.
`
`Observation #6
`
`In Ex. 2013, at 114:12-16 and 116:18-20, Dr. Tellado testified:
`
`Q. Do you know intuitively whether 88 random carriers and 16 Shively
`carriers would have a lower probability of clipping than Scenario 2?
`
`MR. McDOLE: Objection; form.
`
`THE WITNESS: No, I’m not going to guess.
`
`***
`
`Q. Do you know intuitively whether Scenario 5 presents a lower chance of
`clipping than Scenario 2?
`
`A. I don’t want to guess.
`
`That testimony is relevant to (1) the scenarios shown on the graph in Ex.
`
`2011 and (2) the credibility of Dr. Tellado’s statement, in Ex. CSCO-1026 at ¶ 52
`
`(p. 32), about “what a POSITA would have intuitively recognized without
`
`performing mathematical calculations.” The testimony is relevant because Ex.
`
`
`
`7
`
`
`
`Patent Owner Motion For Observation
`IPR2016-01021
`
`
`2011 shows a lower probability of clipping for Scenario 5 (a long loop scenario)
`
`than for Scenario 2 (the clipping rate allowed by the ADSL standard) but Dr.
`
`Tellado says he would not have known this intuitively.
`
`Observation #7
`
`In Ex. 2013, at 54:10-23, Dr. Tellado testified:
`
`Q. If we were to run a Gaussian on the 104 carriers that you ran this quick
`simulation on, where would that line show up in graph 2 on your -- page 30
`of your declaration?
`
`A. The Gaussian process?
`
`Q. Mm-hmm.
`
`A. The Gaussian process would be proportional to the power of 104
`carriers.
`
`Q. Okay. So would it be to the left of the Gaussian process for the 250
`carriers that you used for --
`
`A. Yes.
`
`Q. -- graph 2?
`
`A. Uh-huh. Yes.
`
`That testimony is relevant to Dr. Short’s statement, in Ex. 2003 at ¶ 58
`
`(pp. 26-27), that “[i]n Shively’s proposed system using normal mode for ADSL-
`
`1995 across 18,000 foot cables, … about 60% of the carriers are unusable. . . .
`
`Consequently, the power of a transmitted signal will be reduced by 60%, thereby
`
`resulting in power levels only 40% of maximum” and, at ¶ 63 (p. 32), that there is
`
`an “enormous reduction in clipping achieved by reducing signal power by more
`
`
`
`8
`
`
`
`Patent Owner Motion For Observation
`IPR2016-01021
`
`
`than half.” The testimony is relevant because it confirms that Dr. Tellado agrees
`
`that the proportional drop in signal power due to using only 104 carriers (as in the
`
`long loop scenario) has a lower chance of clipping than when all 250 carriers are
`
`used as in Scenario 2 (the clipping rate allowed by the ADSL standard) and Dr.
`
`Tellado knew this without running a simulation.
`
`Observation #8
`
`In Ex. 2013, at 70:4-21 and 92:24–93:9, Dr. Tellado testified:
`
`Q. So you used the 2-bit threshold that, in your annotated figure on page 21
`of your declaration, is shown in black; correct?
`
`***
`
`Q. So you used the 2-bit threshold that’s illustrated as a black dotted line;
`correct?
`
`A. At around minus 125 dBm?
`
`Q. Yes. Is that the line you used?
`
`A. That’s about -- I mean, just going by eye accuracy. Yes, that’s taken
`from Dr. Short’s 2-bit line.
`
`Q. Okay. But you understand that Dr. Short used a 1-bit threshold as the
`proper line. Do you know that?
`
`A. Okay. …
`
`***
`
`Q. So you didn’t do a simulation that would have shown the chance of
`clipping if a 1-bit threshold had been used -- [objection to form] -- and
`applied to the figure on page 21 of your declaration; right?
`
`MR. McDOLE: Objection; form.
`
`
`
`9
`
`
`
`Patent Owner Motion For Observation
`IPR2016-01021
`
`THE WITNESS: Can you ask it a different way? So I don’t recall running
`that combination.
`
`
`
`That testimony is relevant to (1) the graph, in Ex. CSCO-1026 at ¶ 33 (p.
`
`21), annotating a “1-bit (non-repeating) threshold,” (2) Dr. Tellado’s opinions, in
`
`Ex. CSCO-1026 at ¶ 33 (p. 21) that were based on a using a 2-bit theshold, and (3)
`
`Shivley’s teaching, in Ex. CSCO-1011 at 2:6-8, that “[t]he power of a signal in a
`
`given sub-band must be sufficiently high to carry a minimal (1-bit) QAM tone to
`
`obtain a predefined bit error rate.” The testimony is relevant because it shows that
`
`Dr. Short used a 1-bit threshold as described in Shively, Dr. Tellado instead used a
`
`2-bit threshold, and Dr. Tellado did not run a simulation to determine the chance of
`
`clipping using Shively and a 1-bit threshold.
`
`Observation #9
`
`In Ex. 2013, at 127:13-20, Dr. Tellado testified:
`
`Q. So let me read this to try to make sure its clear.
`
`So your testimony is that a system that rotates all of the QAM symbols
`from the same DMT symbol by the same amount, that rotation is applied to
`each of the QAM symbols in that DMT symbol that have been rotated?
`
`A. Yes.
`
`That testimony is relevant to (1) Patent Owner’s statements, in the Patent
`
`Owner Response (Paper 15) at p. 44, that “Stopler’s directive is to adjust the
`
`phases of all carriers within a single symbol by the same amount,” and, at p. 51,
`
`that “Stopler does not reduce PAR because phase scrambling is performed from
`
`
`
`10
`
`
`
`Patent Owner Motion For Observation
`IPR2016-01021
`
`
`symbol-to-symbol and not from carrier-to-carrier,” and (2) Petitioner’s argument,
`
`in its Reply (Paper 20) at pp. 16-17, that “Stopler’s phase scrambler reduces PAR
`
`because it scrambles phases of individual QAM symbols.” The testimony is
`
`relevant because it shows that Petitioner’s argument that Stopler scrambles QAM
`
`symbols is irrelevant to whether Stopler teaches scrambling QAM symbols over
`
`time or from carrier-to-carrier.
`
`Observation #10
`
`In Ex. 2013, at 142:9–144:23, Dr. Tellado testified:
`
`Q. At the time of the invention of the ‘158 and ‘243 patents, were there
`multiple ways of reducing PAR that were known to those of skill in the art?
`
`A. Yes.
`
`***
`
`Q. Exhibit 2012 has been placed in front of you, Dr. Tellado. Can you tell
`me what Exhibit 2012 is?
`
`A. This one says United States Patent by Tellado, et al, and it’s titled ‘Peak
`to Average Power Ratio Reduction,’ and its filed April 20th, 1998.
`
`Q. This is your patent?
`
`A. Yes
`
`Q. It says:
`
`***
`
`***
`
`“One method randomly shuffles the phase of the signals 10(1)-10(n) at
`each carrier frequency f(1)-f(n). Random shuffling does not completely
`eliminate the problem, although randomizing has been shown to somewhat
`
`
`
`11
`
`
`
`Patent Owner Motion For Observation
`IPR2016-01021
`
`reduce the peak to average power ratio to an extent. Random shuffling also
`requires performing an additional IFFT. In addition to not completely
`reducing the peak to average power ratio to a practical point, that particular
`method also requires that additional information, side information, be sent
`along with the transmitted signal. In order for the receiver to be able to
`decode the transmitted signal the receiver must also know how the signals
`10(1)-10(n) were randomized. Thus, the randomization scheme requires
`extra bandwidth to transmit the side information and does not effectively
`reduce the peak to average power ratio.”
`
`
`
`Do you see that?
`
`A. Yes.
`
`Q. Did I accurately read that?
`
`A. I believe so.
`
`That testimony is relevant to (1) Patent Owner’s statement, in the Patent
`
`Owner Response (Paper 15) at p. 45, that “Petitioners’ (and Petitioners’ expert’s)
`
`reasons to combine Shively and Stopler are without a rational basis, factually
`
`wrong, and suffer from hindsight bias,” (2) Petitioner’s statement, in its Reply
`
`(Paper 20) at p. 8, that “a POSITA would have found it obvious to incorporate a
`
`phase scrambler—like that in Stopler—into Shively’s system to counteract the
`
`increase in PAR caused by Shively’s bit spreading technique,” and (3) Dr.
`
`Tellado’s statement in Ex. CSCO-1026 at ¶ 54 (p. 33) that “the desire to reduce the
`
`cost of a transmitter employing Shively’s technique would have motivated a
`
`POSITA to incorporate a phase scrambler (like in Stopler) into Shively’s system.”
`
`The testimony is relevant because it shows that Petitioner’s own expert, Dr.
`
`
`
`12
`
`
`
`Patent Owner Motion For Observation
`IPR2016-01021
`
`
`Tellado, acknowledged that, at the time of the invention, there were multiple ways
`
`of reducing PAR and he criticized using phase randomization for this purpose.
`
`
`
`
`
`
`
`/Peter J. McAndrews/
`Peter J. McAndrews
`Registration No. 38,547
`McAndrews, Held, & Malloy, Ltd.
`500 West Madison St., Suite 3400
`Chicago, IL 60661
`Telephone: (312) 775-8000
`
`
`
`
`
`Dated: June 30, 2017
`
`
`
`
`CUSTOMER NUMBER: 23446
`
`
`
`13
`
`
`
`Patent Owner Motion For Observation
`IPR2016-01021
`
`CERTIFICATE OF SERVICE
`
`I hereby certify that a true and correct copy of the foregoing Patent Owner
`
`
`
`Motion for Observation Regarding Cross-Examination of Dr. Jose Tellado
`
`was served on June 30, 2017, via email to counsel for Petitioner at the following:
`
`
`
`Lead Counsel
`David L. McCombs
`HAYNES & BOONE, LLP
`2323 Victory Ave., Suite 700
`Dallas, TX 75219
`Tel. 214-651-5533
`Fax 214-200-0853
`david.mccombs.ipr@haynesboone.com
`
`Heidi L. Keefe
`COOLEY LLP
`ATTN: Patent Group
`1299 Pennsylvania Ave., Suite 700
`Washington, DC 20004
`Tel. 650-843-5001
`Fax 650-849-7400
`hkeefe@cooley.com
`Dish-TQDelta@cooley.com
`zpatdcdocketing@cooley.com
`
`John M. Baird
`Duane Morris LLP
`505 9th St. NW, Ste 1000
`
`Back-up Counsel
`Theodore M. Foster
`Tel. 972-739-8649
`Russell Emerson
`Tel. 214-651-5328
`Jamie H. McDole
`Tel. 972-651-5121
`HAYNES & BOONE, LLP
`2323 Victory Ave., Suite 700
`Dallas, TX 75219
`Fax 972-692-9156
`ipr.theo.foster@haynesboone.com
`russell.emerson.ipr@haynesboone.com
`jamie.mcdole@haynesboone.com
`
`Stephen McBride
`COOLEY LLP
`ATTN: Patent Group
`1299 Pennsylvania Ave., Suite 700
`Washington, DC 20004
`Tel. 650-843-5001
`Fax 650-849-7400
`smcbride@cooley.com
`
`
`Christopher Tyson
`Duane Morris LLP
`505 9th St. NW, Ste 1000
`
`
`
`
`
`
`
`Patent Owner Motion For Observation
`IPR2016-01021
`
`
`
`Washington, DC 20004
`Tel. 202-776-7819
`Fax 202-776-7801
`CJTyson@duanemorris.com
`
`Washington, DC 20004
`Tel. 202-776-7819
`Fax 202-776-7801
`JMBaird@duanemorris.com
`
`
`
`
`Date: June 30, 2017
`
`
`
`
`
`
`
`
`
`/Peter J. McAndrews/
`Peter J. McAndrews
`
`
`
`
`
`
`
`