`
`
`
`
`DECLARATION OF DR. SAYFE KIAEI
`IN SUPPORT OF PETITIONER’S REPLY
`
`Case Nos. IPR2016-01006, 1007, 1008, 1009
`
`
`1
`
`CSCO-1100
`
`
`
`
`
`
`
`Cisco v. TQ Delta, IPR2016-01007
`Page 1 of 46
`
`
`
`Declaration of Dr. Kiaei In Support of Petitioner’s Reply
`
`
`TABLE OF CONTENTS
`CONTENTS
`
`A.
`
`B.
`
`Introduction ........................................................................................... 4
`
`Construction .......................................................................................... 5
`
`1.
`
`2.
`
`“subchannel” .............................................................................. 5
`
`“during Showtime” ..................................................................... 8
`
`C. Milbrandt teaches a “subchannel ” ....................................................... 8
`
`D. Milbrandt teaches “power level per subchannel” ............................... 16
`
`E. Milbrandt in combination with ANSI T1.413 teaches that “the
`power level per subchannel information is based on a Reverb
`signal”.................................................................................................. 18
`
`F. Milbrandt in combination with ANSI T1.413 teaches “an array
`representing Signal to Noise Ratio per subchannel during
`Showtime information” ....................................................................... 23
`
`1.
`
`2.
`
`3.
`
`4.
`
`5.
`
`Both Milbrandt and ANSI T1.413 teach a “subchannel” ........ 23
`
`Both Milbrandt and ANSI T1.413 teach a “Showtime” .......... 24
`
`ANSI T1.413’s “SNR” discloses “Signal to Noise
`Ratio…information” ................................................................ 27
`
`ANSI T1.413’s “SNR margin test parameters” discloses
`“Signal to Noise Ratio…information” ..................................... 29
`
`There are numerous and distinct reasons to combine the
`teachings of Milbrandt and ANSI T1.413 to measure and
`transmit SNR during Showtime ............................................... 30
`
`G. Milbrandt in combination with Chang teaches “idle channel
`noise information” ............................................................................... 34
`
`1.
`
`2.
`
`Milbrandt does not “teach away” from Chang ........................ 35
`
`Chang’s teachings are compatible with Milbrandt’s
`principle of operation ............................................................... 38
`
`
`
`2
`
`CSCO-1100
`
`Cisco v. TQ Delta, IPR2016-01007
`Page 2 of 46
`
`
`
`Declaration of Dr. Kiaei In Support of Petitioner’s Reply
`
`
`3.
`
`There are numerous distinct reasons to combine the
`teachings of Milbrandt and Chang ........................................... 40
`
`H.
`
`Conclusion ........................................................................................... 45
`
`3
`
`CSCO-1100
`
`
`
`
`
`
`
`Cisco v. TQ Delta, IPR2016-01007
`Page 3 of 46
`
`
`
`Declaration of Dr. Kiaei In Support of Petitioner’s Reply
`
`
`A.
`
`Introduction
`
`I, Sayfe Kiaei, do hereby declare as follows:
`
`1.
`
`I previously submitted different Declarations as Exhibit 1009 in each
`
`of IPR2016-01006, IPR2016-01007, IPR2016-01008, IPR2016-01009, setting
`
`forth my background and credentials and my curriculum vitae which provides
`
`further details.
`
`2.
`
`I submit this Declaration in reply to TQ Delta’s arguments and the
`
`Declaration of Douglas Chrissan, PhD, filed as Ex-2001.
`
`3.
`
`In preparing this declaration I have reviewed the following
`
`publications in addition to those identified in my first Declarations:
`
`a) George Abe, Residential Broadband (Cisco Press, Second
`
`Edition 2000) (selected pages), Ex-1101;
`
`b) Martin Rowe, ADSL Testing Moves Out of the Lab (April 1,
`
`1999), Ex-1102;
`
`c) Declaration of Robert Short, Ex-1103;
`
`d) U.S. 6,625,219, Ex-1104;
`
`e) U.S. 7,292,627, Ex-1105;
`
`f) Douglas Chrissan, Uni-DSL: One DSL for Universal Service,
`
`White Paper (June 2004), Ex-1106;
`
`g) U.S. 6,374,288, Ex-1107;
`
`
`
`4
`
`CSCO-1100
`
`Cisco v. TQ Delta, IPR2016-01007
`Page 4 of 46
`
`
`
`Declaration of Dr. Kiaei In Support of Petitioner’s Reply
`
`
`h) Ata Elahi, Network Communications Technology (Delmar
`
`Thomson Learning 2001) (selected pages), Ex-1108;
`
`i)
`
`FCC Filing for Alcatel Model 1000 ADSL Modem, 1999, Ex-
`
`1109; and
`
`j)
`
`Deposition Transcript of Douglas Chrissan, Ex-1110.
`
`4.
`
`I have also reviewed TQ Delta’s Patent Owner’s Responses and other
`
`documents cited in the below analysis.
`
`B. Construction
`
`1.
`
`“subchannel”
`
`5.
`
`I understand Dr. Chrissan to have concluded that a “subchannel,” as
`
`defined by U.S. Patent Nos. 7,835,430 (the “‘430 patent’), 8,432,956 (the “‘956
`
`patent’), and 8,238,412 (the “‘412 patent’) (collectively “the patents at issue”), is
`
`“a carrier of a multicarrier communication channel.” Ex-2001, ¶32. I disagree with
`
`this proposed interpretation.
`
`6.
`
`In my opinion, Dr. Chrissan’s proposed interpretation, which limits
`
`the term “subchannel” to a “carrier,” improperly excludes other terms that a person
`
`of ordinary skill in the art (“POSITA”) at the time would have understood as
`
`equivalent and interchangeable. In the field of multicarrier communications, a
`
`POSITA as of 1999, would have been familiar with the term “subchannel” and
`
`would have understood that it is equivalent and interchangeable with the terms
`
`
`
`5
`
`CSCO-1100
`
`Cisco v. TQ Delta, IPR2016-01007
`Page 5 of 46
`
`
`
`Declaration of Dr. Kiaei In Support of Petitioner’s Reply
`
`“tone,” “carrier,” “subcarrier,” “channel,” “band,” “sub-band.” Also, term “sub-
`
`frequency,” in the ADSL context, would also be understood to be equivalent and
`
`interchangeable with the term “subchannel.”
`
`7.
`
`This is confirmed by the patents at issue, which interchangeably refer
`
`to “subchannels,” “tones,” and “carriers.” ‘956 patent, 1:42-48, 4:38-40.1 Dr.
`
`Chrissan also agreed during deposition that the terms “channel,” “carrier,”
`
`“subcarrier,” and “subband” can also be used interchangeably to refer to a
`
`“subchannel.” Ex-1109, 43:13-49:15; 53:20-54:1. Further, Dr. Robert Short, TQ
`
`Delta’s expert in a related proceeding, filed a declaration testifying that “adding to
`
`potential confusion is that the terms ‘carrier,’ ‘subcarrier,’ ‘band,’ ‘sub-band,’
`
`‘bin,’ ‘channel,’ and ‘tone’ are often used interchangeably.” Ex-1103, ¶ 36.
`
`8.
`
`I also note that other references evidence that the terms tone, carrier,
`
`subcarrier, band, sub-band, channel, sub-frequency are used interchangeably to
`
`describe the same concept referred to as a “subchannel” in the patents at issue. See
`
`e.g., Ex-1101, 69 (“there are 256 subbands.”); Ex-1102, p. 3 (“channel” “a tone”);
`
`Ex-1104, Abstract (“frequency (tones or sub-channels)”), 1:41 (“tones or bands”);
`
`Ex-1105, 1:36 (“Carrier signals (carriers) or sub-channels”); Ex-1106, 13
`
`(“channel” “tone” “carrier”); Ex-1014, 46 (“carriers” “sub-carriers.”); Ex-1107,
`
`1 Because the ‘956 patent shares a common specification with the ‘430 patent and
`
`the ‘412 patent, for brevity, citations in this document cite only to one patent.
`
`
`
`6
`
`CSCO-1100
`
`Cisco v. TQ Delta, IPR2016-01007
`Page 6 of 46
`
`
`
`Declaration of Dr. Kiaei In Support of Petitioner’s Reply
`
`1:29-33 (“sub-frequency” “channel”); Ex-1108, 108-109, (“subchannels”
`
`“subfrequency” “channel”); Ex-1011, 11:2-4. (“sub-frequency” “channel.”).
`
`9.
`
`Accordingly, it is my opinion that Dr. Chrissan’s proposed
`
`interpretation of the term “subchannel” is not the broadest reasonable interpretation
`
`in light of the specification since it is limited to a “carrier” and excludes other
`
`equivalent and interchangeable terms.
`
`10. Also, Dr. Chrissan’s proposed construction is unhelpful since it is
`
`circular. His proposed construction includes the term “channel,” which is
`
`equivalent and interchangeable with the term “carrier” that is already present in his
`
`interpretation. Ex-1109, 53:20-54:1 (“I personally in appropriate contexts equate,
`
`channel, subchannel, carrier, and subcarrier. I believe that other people in
`
`appropriate contexts would equate any of those four terms as well.”). As such,
`
`including the term “channel” in the interpretation adds to the confusion in
`
`terminology. For this additional reason, Dr. Chrissan’s circular and confusing
`
`interpretation fails to be of any use in helping the Board understand how the term
`
`“subchannel” applies to the prior art in this proceeding.
`
`11. Therefore, it is my opinion that a POSITA, after reviewing the patents
`
`at issue, would have understood that the term “subchannel” is equivalent and
`
`interchangeable with a tone, carrier, subcarrier, band, sub-band, sub-frequency, or
`
`channel, of a multicarrier frequency spectrum. Consistent with this understanding,
`
`
`
`7
`
`CSCO-1100
`
`Cisco v. TQ Delta, IPR2016-01007
`Page 7 of 46
`
`
`
`Declaration of Dr. Kiaei In Support of Petitioner’s Reply
`
`a “subchannel” would be understood to be “a portion of a frequency spectrum used
`
`for communication.”
`
`2.
`
`“during Showtime”
`
`12.
`
`I understand Dr. Chrissan to have concluded that the term “during
`
`Showtime,” as used in the patents at issue includes other DSL standards, beside the
`
`ANSI T1.413 standard I discussed in my first declaration. Specifically, Dr.
`
`Chrissan contends that “’Showtime’ was a concept that was also used in
`
`connection with the ITU-T G.992.1 and G.992.2 DSL communications standards.”
`
`Ex-2001, ¶31. Further, during deposition, Dr. Chrissan also added that the term
`
`also applied to “ADSL2 and VDSL2” communication standards. Ex-1110, 80:2.
`
`Dr. Chrissan provides no evidence that the term of art Showtime was used in
`
`connection with any other DSL standards. Therefore, since Showtime was a term
`
`of art in DSL standards, it is my opinion that a POSITA would have understood the
`
`term of art “during Showtime” to be “during normal communications of a device
`
`compliant with the ANSI T1.413, ITU-T G.992.1, G.992.2, ADSL2, or VDSL2
`
`communication standards.”
`
`C. Milbrandt teaches a “subchannel ”
`
`13.
`
`I understand that Dr. Chrissan, concludes that “Milbrandt’s sub-
`
`frequency does not correspond to the claimed ‘subchannel.” Ex-2001, ¶¶ 36-37.
`
`
`
`8
`
`CSCO-1100
`
`Cisco v. TQ Delta, IPR2016-01007
`Page 8 of 46
`
`
`
`Declaration of Dr. Kiaei In Support of Petitioner’s Reply
`
`According to Dr. Chrissan, Milbrandt’s sub-frequency “refers to the upstream or
`
`downstream frequency bands in ADSL.” Id., ¶40. I disagree.
`
`14. As I explain below, there is no dispute that Milbrandt’s “channels”
`
`teach the claimed “subchannels.” Ex-1011, 10:15-65. And Milbrandt equates these
`
`“channels” (which are “subchannels”) with “sub-frequencies.” Ex-1011, 11:2-4.
`
`Thus, Milbrandt’s “sub-frequencies” are “subchannels,” as claimed.
`
`1. Milbrandt’s “channel” is a “subchannel”
`
`15. Milbrandt explains that “ADSL modems 60…use[] DMT technology
`
`to divide the bandwidth of a subscriber line 16…into many individual …
`
`channels.” Ex-1011, 10:58-63. Because Milbrandt uses discrete multitone
`
`technology to divide the ADSL frequency spectrum into channels, a POSITA
`
`would have understood that there would be, for example, 256 tones/channels. Ex-
`
`1013, 2:66-3:5. Dr. Chrissan agrees with this understanding, noting that ADSL
`
`modems have 256 channels. Ex-2001, ¶ 53.
`
`16. Milbrandt further explains that “[e]ach channel of a subscriber line 16
`
`uses a form of quadrature amplitude modulation (QAM) to transmit data in each
`
`channel simultaneously.” Ex-1011, 10:63-65. Based on this passage, a POSITA
`
`would understand that Milbrandt is describing how its ADSL modem performs
`
`QAM modulation on each of the individual 256 tones/channels to transmit the data
`
`simultaneously as a single signal. Ex-1013, 2:66-3:5. In ADSL, QAM does not
`
`
`
`9
`
`CSCO-1100
`
`Cisco v. TQ Delta, IPR2016-01007
`Page 9 of 46
`
`
`
`Declaration of Dr. Kiaei In Support of Petitioner’s Reply
`
`operate on the upstream or downstream frequency spectrum as a whole—it
`
`operates at the tone/channel level. Each channel is allocated certain number of
`
`bits, and these bits are modulated for that channel as a QAM. This was illustrated
`
`by TQ Delta’s expert, Dr. Short, in his declaration with a four channel/carrier
`
`example, which is reproduced below:
`
`
`
`17. Notably, Dr. Chrissan conceded that he “personally… equate[s]
`
`channel [and] subchannel” and further conceded that, in Milbrandt’s ANSI
`
`T1.413 standard compliant device, "the QAM modulation is performed on each
`
`individual subchannel.” Ex-1110, 53:20-54:1, 65:6-12; id. 62:2-3; Resp., 28 n.2
`
`(TQ Delta agreeing that Milbrandt “complie[s] with the ANSI T1.413 standard.”)
`
`18. Therefore, it is my opinion that it would have been understood to a
`
`POSITA that Milbrandt’s “channel” is the “smallest division of the data
`
`
`
`10
`
`CSCO-1100
`
`Cisco v. TQ Delta, IPR2016-01007
`Page 10 of 46
`
`
`
`Declaration of Dr. Kiaei In Support of Petitioner’s Reply
`
`transmission in a multicarrier communication system that uses DMT modulation,”
`
`and therefore discloses a “subchannel,” as argued by TQ Delta. IPR2016-01007,
`
`Resp., 14.
`
`2. Milbrandt’s “channel” (which is a “subchannel”) is a “sub-
`frequency”
`
`19. Milbrandt continues and explains that the “frequency range from 25
`
`kHz to 1.1. MHz …is divided into sub-frequencies. Each sub-frequency is an
`
`independent channel and supports transmission of its own stream of data signals.”
`
`Ex-1011, 11:2-4. A POSITA would have understood that here Milbrandt equates
`
`the earlier discussed “channel” of ADSL modem (which is a “subchannel”) with a
`
`“sub-frequency” and that each independent sub-frequency/channel is a discrete
`
`non-overlapping portion of a multicarrier frequency spectrum from 25 kHz to 1.1
`
`MHz. Milbrandt’s description was a common way of explaining how to divide the
`
`frequency range. See e.g., Ex-1107, 1:29-33 (“Discrete MultiTone (DMT) is a
`
`modulation technique used in xDSL technologies, that divides the frequency range
`
`into 256 sub-frequencies… Each sub-frequency is an independent channel which
`
`has its own stream of signals.”); Ex-1108, 108-109 (“The frequency spectrum
`
`above 26 kHz is divided into 249 independent subchannels, each containing 4.3
`
`kHz bandwidth. Each subfrequency is an independent channel and has its own
`
`stream of signals.”) Therefore, based on the express teachings, it would have been
`
`understood that Milbrandt’s “sub-frequency” is a “subchannel.”
`
`
`
`11
`
`CSCO-1100
`
`Cisco v. TQ Delta, IPR2016-01007
`Page 11 of 46
`
`
`
`Declaration of Dr. Kiaei In Support of Petitioner’s Reply
`
`
`20.
`
`I disagree with Dr. Chrissan’s statement that “[b]ased on Milbrandt’s
`
`use of the term ‘sub-frequency,’ one of ordinary skill in the art would conclude that
`
`the term refers to the upstream or downstream frequency bands in ADSL.” Ex-
`
`2001, ¶40. Dr. Chrissan’s statement amounts to an assertion that Milbrandt’s
`
`ADSL modem 60 has only two sub-frequencies—one for upstream and another for
`
`downstream. Putting it plainly, such an assertion is unreasonable and does not flow
`
`from Milbrandt’s disclosure.
`
`21. Milbrandt expressly states that “[e]ach sub-frequency is an
`
`independent channel” and that “one group of channels [plural] is allocated for the
`
`uplink transmission of data and the other for the downlink transmission of data.”
`
`Ex-1011, 11:2-10. In other words, Milbrandt provides a “group of channels” for
`
`uplink and another “group of channels” for downlink transmission. This further
`
`demonstrates that Milbrandt’s sub-frequency/channel is not the whole upstream
`
`and downstream frequency spectrum. Therefore, based on Milbrandt’s plain
`
`teachings, Dr. Chrissan’s statement is wholly unreasonable.
`
`22. Accordingly, it is my opinion that a POSITA would have understood
`
`that Milbrandt’s disclosure of its ADSL modem 60 (spanning 10:58-11:10) uses
`
`the terms “sub-bands,” “channels,” “sub-channels,” and “sub-frequencies”
`
`interchangeably to describe discrete non-overlapping portions (e.g., 256 carriers)
`
`
`
`12
`
`CSCO-1100
`
`Cisco v. TQ Delta, IPR2016-01007
`Page 12 of 46
`
`
`
`Declaration of Dr. Kiaei In Support of Petitioner’s Reply
`
`of a frequency spectrum from 25 kHz to 1.1 MHz that uses DMT/QAM
`
`modulation for communication.
`
`23. For explanation purposes, I have provided below an illustration of
`
`ADSL frequencies with annotations from Milbrandt’s above cited portions.
`
`“DMT technology to divide the bandwidth . . . into many
`individual … channels.” Ex-1011, 10:58-63
`
`
`
`“subchannels are divided into
`groups and one group of channels is
`
`allocated for the uplink” Ex-1011,
`11:6-10
`
`“subchannels are divided into groups
`and . . . other [group of channels is
`allocated] for the downlink.” Ex-1011,
`11:6-10
`
`
`“frequency range from 25 kHz to 1.1. MHz . . . divided into sub-frequencies.
`Each sub-frequency is an independent channel.” Ex-1011, 11:2-5
`
`
`
`24. As can be seen from the above illustration and corresponding text,
`
`when Milbrandt describes the ADSL frequency spectrum from 25 kHz to 1.1 MHz,
`
`the terms “sub-bands,” “channels,” “sub-channels,” and “sub-frequencies” are used
`
`interchangeably to refer to a discrete non-overlapping portion of a multicarrier
`
`
`
`13
`
`CSCO-1100
`
`Cisco v. TQ Delta, IPR2016-01007
`Page 13 of 46
`
`
`
`Declaration of Dr. Kiaei In Support of Petitioner’s Reply
`
`frequency spectrum. The interchangeability of these terms is consistent with my
`
`conclusion that a POSITA would have understood the term “sub-frequency” in
`
`Milbrandt to disclose the claimed “subchannel,” even under TQ Delta’s narrow
`
`construction.
`
`3.
`
`V.90 protocol is an alternative to ADSL protocol
`
`25. Further, I disagree with Dr. Chrissan that somehow Milbrandt’s
`
`disclosure of V90 protocol supports his position that a “sub-frequency” is not a
`
`“subchannel.” Ex-2011, ¶¶40-41. The V.90 protocol is a MODEM standard for
`
`data communication over the telephone network. The V.90 protocol uses the voice
`
`frequency spectrum, which is in fact a “sub-frequency” of the overall frequency
`
`spectrum. And to be clear, the V.90 protocol is an alternative to the ADSL
`
`protocol (Milbrandt’s primary embodiment), and uses the POTS frequency range
`
`of 0-4kHz (which is approximately the size of one channel). That these are
`
`alternatives was confirmed by Dr. Chrissan during deposition. Ex-1110, 142:2-5
`
`(“Milbrandt does describe in his patent in general as of xDSL and V.90 being
`
`alternative protocols.”) A POSITA would have understood that since the V.90
`
`protocol is an alternative to the ADSL protocol, Milbrandt’s teaching or use of the
`
`term “sub-frequency” in the context of V.90 protocol is not inconsistent with the
`
`use of that term in the context of the ADSL protocol.
`
`
`
`14
`
`CSCO-1100
`
`Cisco v. TQ Delta, IPR2016-01007
`Page 14 of 46
`
`
`
`Declaration of Dr. Kiaei In Support of Petitioner’s Reply
`
`
`4. Milbrandt’s Figure 3 example is not limiting
`
`26.
`
`I also disagree with Dr. Chrissan’s assertion that Figure 3 somehow
`
`supports his position. Ex-2001, ¶44. Dr. Chrissan states regarding Figure 3 that
`
`“six columns are shown, and although a person of skill in the art would not
`
`interpret this to mean exactly six columns, that person would also recognize that
`
`Milbrandt is not describing hundreds of columns.” Id. In other words, he admits
`
`that Figure 3 is not limited to six sub-frequencies as illustrated, but then concludes
`
`that there is an upper limit. The basis for this conclusion is lacking—he cites
`
`nothing for that proposition. Since ADSL was known to use 256 tones/channels, it
`
`would have been understood that Milbrandt’s ADSL modem in fact used 256
`
`tones/channels rather than only six as illustrated. This is confirmed by Milbrandt
`
`itself which states that there are “many individual . . . channels” and that “[e]ach
`
`sub-frequency is an independent channel,” without providing an upper limit of how
`
`many sub-frequencies/channels are used. Ex-1011, 10:58-11: 4.
`
`27. Therefore, it is my opinion that a POSITA would have understood that
`
`Milbrandt is not limited to the example illustrated in Figure 3 and that in fact it
`
`includes many individual sub-channels/channels/sub-frequencies (e.g., 256
`
`normally used ADSL communication).
`
`
`
`15
`
`CSCO-1100
`
`Cisco v. TQ Delta, IPR2016-01007
`Page 15 of 46
`
`
`
`Declaration of Dr. Kiaei In Support of Petitioner’s Reply
`
`
`D. Milbrandt teaches “power level per subchannel”
`
`28.
`
`I understand that Dr. Chrissan, concludes that Milbrandt’s power
`
`spectral density (PSD) per sub-frequency does not correspond to the claimed
`
`“power level per subchannel” because Milbrandt’s “sub-frequency” “may
`
`comprise hundreds of subchannels.” Ex-2001, ¶49. I disagree.
`
`29. As I already explained above, Milbrandt’s “sub-frequency” would be
`
`understood to be a discrete non-overlapping portion (e.g., one of 256 carriers) of a
`
`frequency spectrum from 25 kHz to 1.1 MHz that uses DMT/QAM modulation for
`
`communication. And as I explained in my first declaration, Milbrandt’s PSD per
`
`sub-frequency is representative of the power within that sub-frequency. It was
`
`well-known that PSD is simply the derivative, with respect to frequency, of power,
`
`which means that PSD is the power level (watts) in 1 Hz bandwidth. A POSITA
`
`would have understood, based on this background knowledge, that the power level
`
`in Milbrandt’s sub-frequency is represented by the integral of measured PSD
`
`across that sub-frequency. This relationship is textbook knowledge and given by
`
`the following equation:
`
`where:
`Power Level between ω1 and ω2 is ΔPg;
`PSD is Sg;
`
`
`
`
`
`16
`
`CSCO-1100
`
`Cisco v. TQ Delta, IPR2016-01007
`Page 16 of 46
`
`
`
`Declaration of Dr. Kiaei In Support of Petitioner’s Reply
`
`
`ω1 is the lower frequency bound; and
`ω2 is upper frequency bound.
`
`Ex-1021, 126-127.
`
`30. During deposition, Dr. Chrissan did not dispute that PSD represents
`
`“power level,” per se. To the contrary, Dr. Chrissan, confirmed that it was well-
`
`known that the integral of PSD over a frequency range represents the power level.
`
`Ex-1110, 104:2-15. (“Q. And so if you integrated the power spectral density
`
`function from Frequency 1 to Frequency 2 what you would end up with is a
`
`measure of power expressed in units of power such as watts or dBm, right? A. If
`
`you did that, you would have a measure of power of a signal from Frequency F1
`
`to Frequency F2 as you described.”) Since it was known that ADSL sub-
`
`frequencies have a frequency range of 4.3125 kHz, when Milbrandt’s PSD is
`
`integrated for each sub-frequency across its respective range, the power level for
`
`that sub-frequency is obtained. Ex-1013, 3:3-5
`
`31. Therefore, it is my opinion that a POSITA would have understood that
`
`Milbrandt’s PSD sub-frequency does “represent[] power level per subchannel
`
`information.”
`
`
`
`17
`
`CSCO-1100
`
`Cisco v. TQ Delta, IPR2016-01007
`Page 17 of 46
`
`
`
`Declaration of Dr. Kiaei In Support of Petitioner’s Reply
`
`
`E. Milbrandt in combination with ANSI T1.413 teaches that “the
`power level per subchannel information is based on a Reverb
`signal”
`
`32. Claims 2, 4, 6, and 8 in the ‘956 patent and claims and 12, 4, 6, 8, 10,
`
`and 12, in the ‘412 patent recite that the “the power level per subchannel
`
`information is based on a Reverb signal.” I have explained in my first declarations
`
`(Ex-1009, IPR2016-01007, IPR2016-01008, IPR2016-01009) how the prior art
`
`teaches this limitation.
`
`33. As I explained above, Milbrandt measures PSD per sub-frequency,
`
`which represents “power level per subchannel information.”
`
`34.
`
`I understand that Dr. Chrissan, argues that “Cisco is incorrect that
`
`ANSI T1.413 discloses measuring power level per subchannel values based on a
`
`Reverb signal.” Ex-2001, ¶52. Dr. Chrissan, in making this argument regarding the
`
`“per subchannel” portion of the claim analyzes the references separately (rather
`
`than the proposed combination). Nevertheless, I disagree with him on this point.
`
`ANSI T1.413 teaches that “During CQUIET3, or QPILOT1 as appropriate, the
`
`ATU-C shall measure the aggregate received upstream power on subcarriers 7–
`
`18 of R-REVERB1, and thereby calculate a downstream PSD.” “Ex-1014, 94. A
`
`POSITA would have understood that the aggregate includes individual values for
`
`each of sub-carriers. This is in fact how the patent at issue discloses that it
`
`measures power level per subchannel based on “Average Reverb Signal contains
`
`
`
`18
`
`CSCO-1100
`
`Cisco v. TQ Delta, IPR2016-01007
`Page 18 of 46
`
`
`
`Declaration of Dr. Kiaei In Support of Petitioner’s Reply
`
`the power levels per tone.” Ex-1001, 4:34-35. Dr. Chrissan’s testimony that the
`
`“aggregate received upstream power of ANSI T1.413 is more accurately an
`
`average of individual power levels,” is not inconsistent with my position. Ex-
`
`2001, ¶53. The aggregate includes for each tone/subchannel an average power
`
`level value, which means that for subcarriers 7–18, there would be 11 aggregate
`
`average power levels—just like the Average Reverb of the patents at issue.
`
`35. Dr. Chrissan further asserts that “adjusting gain and equalization
`
`would not have been a reason to transmit power level per subchannel
`
`information—rather, it would have been a reason to transmit the Reverb signals.”
`
`Ex-2001, ¶55. It appears that Dr. Chrissan does not understand that Milbrandt
`
`already determines PSD per sub-frequency and transmits the determined PSD per
`
`sub-frequency to the central office modem. Ex-1011, 11:11-24. The combination
`
`relied on ANSI T1.413 to measure PSD “based on a Reverb signal,” not for
`
`transmitting.
`
`36. A POSITA would have understood that when ANSI T1.413 states that
`
`the REVERB “allows the ATU-R receiver to . . . train any receiver equalizer” it
`
`would have been understood that this includes determining PSD based on the
`
`REVERB signal and using that PSD to adjust the equalizer. When confronted with
`
`a question specifically on this point, Dr. Chrissan admitted that he did not dispute
`
`my understanding. Ex-1110, 100:6-15. (“Q. Could the notion of power [for
`
`
`
`19
`
`CSCO-1100
`
`Cisco v. TQ Delta, IPR2016-01007
`Page 19 of 46
`
`
`
`Declaration of Dr. Kiaei In Support of Petitioner’s Reply
`
`training the equalizer] be a notion of power derived from a reverb signal? A. I
`
`haven’t thought about that and I don’t have an opinion on it.”). Therefore, it is my
`
`opinion that a POSITA would have understood that ANSI T1.413’s teaching of
`
`measuring PSD based on Reverb would be beneficial to Milbrandt since it would
`
`permit for adjusting the equalizer.
`
`37. Likewise, regarding adjusting the automatic gain. It is my opinion that
`
`a POSITA would have understood that when ANSI T1.413 states that the
`
`REVERB “allows the ATU-R receiver to adjust its automatic gain control (ACG)
`
`to an appropriate level” it would have been understood that this includes
`
`determining PSD based on a REVERB signal and using that PSD to adjust the
`
`AGC. This is because the gain is based on power, which is represented by PSD.
`
`Again, Dr. Chrissan does not dispute this understanding. Ex-2001, ¶55.
`
`38. Another motivation was to make Milbrandt’s system compliant with
`
`the ANSI T1.413 standard. IPR2016-01007, Ex-1009, ¶86-87; Ex-1011, 9:31-34.
`
`Dr. Chrissan argues that “even if ANSI T1.413’s PSD based on Reverb did
`
`represent the claimed power level per subchannel based on a Reverb signal, ANSI
`
`T1.413 does not disclose or require transmitting it—it only discloses calculating
`
`it.” Ex-2001, ¶56. The fact that ANSI does not disclose transmitting the PSD is
`
`ancillary since the combination does not rely on ANSI T1.413 for that teaching.
`
`As already noted above, Milbrandt already teaches transmitting PSD to the central
`
`
`
`20
`
`CSCO-1100
`
`Cisco v. TQ Delta, IPR2016-01007
`Page 20 of 46
`
`
`
`Declaration of Dr. Kiaei In Support of Petitioner’s Reply
`
`office modem. Ex-1011, 11:11-24. And, Dr. Chrissan readily admits that ANSI
`
`T1.413 “discloses calculating” “PSD based on Reverb.” Thus, Dr. Chrissan does
`
`not dispute complying with ANSI T1.413 standard in fact requires calculating PSD
`
`based on Reverb signal.
`
`39. Dr. Chrissan’s argument also ignores that it is important for an ADSL
`
`modem to measure PSD based on REVERB to demonstrate compliance with the
`
`ANSI T1.413 standard, e.g., power limits under FCC rules. For example, provided
`
`is a copy of an Alcatel FCC filing, which recognizes the importance of
`
`“demonstrate[ing] compliance with the requisite ANSI T1.413 . . . standard.” Ex-
`
`1109, 3. To demonstrate compliance, Alcatel measures PSD based on Reverb.
`
`
`
`21
`
`CSCO-1100
`
`Cisco v. TQ Delta, IPR2016-01007
`Page 21 of 46
`
`
`
`Declaration of Dr. Kiaei In Support of Petitioner’s Reply
`
`
`
`
`Ex-1109, FIG. 5.6.
`
`40. As shown above, a POSITA would have measured the PSD based on
`
`REVERB to understand the changes over the frequency spectrum (i.e., per
`
`subchannel) and demonstrate compliance with ANSI T1.413. Therefore, and as I
`
`stated in my first declaration, it is my opinion that a POSITA would have measured
`
`PSD per sub-frequency in Milbrandt’s ADSL system based on Reverb to also
`
`comply with ANSI T1.413, which is a goal of Milbrandt. See e.g., Ex-1011, 9:31-
`
`34 (“support communication using ADSL techniques that comply with ANSI
`
`Standard T1.413.”).
`
`
`
`22
`
`CSCO-1100
`
`Cisco v. TQ Delta, IPR2016-01007
`Page 22 of 46
`
`
`
`Declaration of Dr. Kiaei In Support of Petitioner’s Reply
`
`
`F. Milbrandt in combination with ANSI T1.413 teaches “an array
`representing Signal to Noise Ratio per subchannel during
`Showtime information”
`
`41. Claim 9 in the ‘956 patent and claims 13-15 in the ‘412 patent recite
`
`the transmission or reception of a test or diagnostic message comprising “an array
`
`representing Signal to Noise Ratio per subchannel during Showtime information.”
`
`To address this limitation, I explained in my first declaration (Ex-1009, IPR2016-
`
`01007, IPR2016-01008, IPR2016-01009) how the combination of Milbrandt and
`
`ANSI T1.413 teaches this limitation and provided motivations to combine.
`
`42.
`
`I understand Dr. Chrissan to have concluded that the combination of
`
`Milbrandt and ANSI T1.413 does not “satisfy the claim limitation [“Signal to
`
`Noise Ratio per subchannel during Showtime”], nor would it have been obvious
`
`that they satisfy the claim limitation.” Ex-2001, ¶ 59. I disagree.
`
`1.
`
`Both Milbrandt and ANSI T1.413 teach a “subchannel”
`
`43. Dr. Chrissan’s conclusion is primarily based on the faulty premise that
`
`“Milbrandt does not disclose measuring or determining any test or diagnostic
`
`parameters ‘per subchannel.’” Ex-2001, ¶ 60.
`
`44. As I already explained above, Milbrandt’s “sub-frequency” would be
`
`understood to be a discrete non-overlapping portion (e.g., one of 256 carriers) of a
`
`frequency spectrum from 25 kHz to 1.1 MHz that uses DMT/QAM modulation for
`
`communication, and therefore discloses a “subchannel,” even under TQ Delta’s
`
`
`
`23
`
`CSCO-1100
`
`Cisco v. TQ Delta, IPR2016-01007
`Page 23 of 46
`
`
`
`Declaration of Dr. Kiaei In Support of Petitioner’s Reply
`
`narrow interpretation. And as I explained in my first declaration, Milbrandt
`
`measures noise information per sub-frequency. IPR2016-01007, Ex-1009, p. 83-
`
`84. Therefore, it is my opinion that Milbrandt does in fact teach measuring noise
`
`information per “subchannel.”
`
`45. Dr. Chrissan further contends that this limitation is not taught because
`
`“ANSI T1.413 only discloses measuring ‘signal to noise ratio margin’ as a single
`
`value for the entire communications channel—not ‘per subchannel.’” Ex-2001, ¶
`
`60. I disagree. As I explained in my declaration, both of ANSI T1.413’s “signal-
`
`to-noise ratio (SNR) margin test parameters” and “SNR, as measured,” teach SNR
`
`per subchannel. IPR2016-01007, Ex-1009, p. 87. Dr. Chrissan completely ignores
`
`and does not address in his declaration the measured SNR relied upon. However,
`
`during deposition, Dr. Chrissan conceded that ANSI T1.413 measures “SNR for
`
`each tone” and that a “tone [is] the same as [a] subchannel.” Ex-1110, 88:5-7,
`
`125:23-126:12.
`
`46. Therefore, consistent with my first declaration, both Milbrandt and
`
`ANSI T1.413 teach measuring noise information per “subchannel.”
`
`2.
`
`Both Milbrandt and ANSI T1.413 teach a “Showtime”
`
`47.
`
`I also disagree with Dr. Chrissan’s position that Milbrandt and ANSI
`
`T1.413 do not teach “noise or signal to noise parameters ‘during Showtime.’” Ex-
`
`2001, ¶ 60. In my declaration I relied on Milbrandt at 12:58-63 statement that
`
`
`
`24
`
`CSCO-1100
`
`Cisco v. TQ Delta, IPR2016-01007
`Page 24 of 46
`
`
`
`Declaration of D