`May 8, 2018
`Apple, Inc. v. California Institute of Technology.
`Case No. IPR2017-00728
`
`CALTECH - EXHIBIT 2040
`Apple Inc. v. California Institute of Technology
`IPR2017-00728
`
`
`
`Instituted Grounds: Pat. No. 7,421,032
`IPR2017-00700
`Ground
`Claims
`Basis
`Prior Art
`1
`11, 12, 14-16
`103
`Ping, MacKay, Divsalar
`IPR2017-00701
`2
`13
`103
`Ping, MacKay, Divsalar, Luby97
`Ground
`Claims
`Basis
`Prior Art
`IPR2017-00728
`1
`1, 4-10
`103
`Ping, MacKay, Divsalar, Luby97
`Ground
`Claims
`Basis
`Prior Art
`18-23
`1
`103
`Ping, MacKay, Divsalar, Luby97
`Patent Owner Caltech’s Oral Argument
`2
`IPR2017-00728
`
`
`
`Caltech’s Groundbreaking Invention
`
`3
`
`
`
`IRA codes solved a decades-old problem
`POR1, 55-63
`(cid:1)Gallagerintroduced LDPCcodes in 1963, but they required
`complex quadratic encoding that prevented their adoption
`for decades. (MM ¶¶181-182, 194.)
`Richardson 2001 (EX 2011) at 2 (emphasis added)
`
`4
`
`Kim 2006 (EX 2010) at 1 (emphasis added)
`Patent Owner Caltech’s Oral Argument
`IPR2017-00728
`
`
`
`IRA codes solved a decades-old problem
`
`POR 1, 55-63
`
`5
`
`MM ¶¶ 186, 187
`Patent Owner Caltech’s Oral Argument
`IPR2017-00728
`
`
`
`IRA codes solved a decades-old problem
`
`POR1, 55-63
`
`Kim 2006 (EX 2010) at 23 (emphasis added)
`
`Patent Owner Caltech’s Oral Argument
`IPR2017-00728
`
`6
`
`
`
`Academia and industry praised IRA codes
`POR1, 55-63
`(cid:1)The IEEE 802.20 Working Group proposed the use of IRA
`codes in a new standard. (MM ¶¶189.)
`(cid:1)The named inventors’ contemporaneous paper, titled
`“Irregular Repeat-Accumulate Codes,” has been cited over
`Lestable2007 (EX 2012) at 2 (emphasis added)
`700 times. (MM ¶191.)
`(cid:1)Named inventors Dr. Jin’sand Dr. Khandekar’sPh.D. theses
`regarding IRA codes have been cited 75 and 54 times,
`respectively. (MM ¶191.)
`Patent Owner Caltech’s Oral Argument
`7
`IPR2017-00728
`
`
`
`The Field Of Error Correction Is
`Highly Unpredictable
`
`8
`
`
`
`Unpredictability equates with nonobviousness
`Reply isoMTE 2
`“Unpredictabilityof results equates more
`with nonobviousnessrather than
`obviousness, whereas that which is
`predictable is more likely to be obvious.”
`Honeywell Int’l, Inc. v. MexichemAmancoHolding S.A.,
`865 F.3d 1348, 1356 (Fed. Cir. 2017).
`
`9
`
`Patent Owner Caltech’s Oral Argument
`IPR2017-00728
`
`
`
`Mathematical analysis of new codes is
`difficult or impossible
`
`POR 4-6, 44-47; Sur. 6; Reply isoMTE 2
`
`EX2033 (Davis Dep.) 256:22-258:12
`Reply 14
`Patent Owner Caltech’s Oral Argument
`IPR2017-00728
`
`10
`
`
`
`Discovery of new codes requires
`guesswork and experimentation
`
`POR 4-6, 44-47; Sur. 6;
`Reply isoMTE 2
`
`EX2033 (Davis Dep.) 258:8-12
`
`11
`
`MM ¶128
`
`EX1202(MacKay) at 1450
`Patent Owner Caltech’s Oral Argument
`IPR2017-00728
`
`
`
`Discovery of new codes requires
`guesswork and experimentation
`(cid:1)Divsalar’s1998 paper shows that leading scientists were still
`POR 46
`seeking to understand how turbo codes achieved near-Shannon-
`limit performance years after Berrou’sdiscovery.
`EX1217 (Divsalar) p. 201
`
`12
`
`EX2031 (DivsalarDecl.) ¶18
`
`Patent Owner Caltech’s Oral Argument
`IPR2017-00728
`
`
`
`Petitioner’s Proposed Motivations to
`Combine are Flawed
`
`13
`
`
`
`MacKay discloses “irregular” parity-check
`matrices
`(cid:1)The ’032 patent claims are directed to the encoding and
`POR 6-7, 18-21
`decoding of information bits to form codewords. MM ¶¶
`29-37.
`(cid:1)In contrast, MacKay is directed to “irregular” parity-check
`matrices that represent codewords. MM ¶¶ 57-58.
`“[MacKay’s] 93 variations don’t give any sense of what the
`encoding would befor the code associated to that.”
`EX2033 (Davis Dep.) 300:5-6
`
`14
`
`Patent Owner Caltech’s Oral Argument
`IPR2017-00728
`
`
`
`MacKay discloses “irregular” parity-check
`matrices
`(cid:1)Regular codes: “[V]ery
`sparse random parity check
`matriceswith uniformweight
`t per column and trper row.”
`EX1202, p. 1449.
`(cid:1)Irregular codes: “[P]arity
`check matrices[that] have
`nonuniformweight per
`column.” Id.
`15
`
`EX1202, p. 1450
`
`POR 6-7, 20-21, 27-33
`
`EX1202, p. 1450
`Patent Owner Caltech’s Oral Argument
`IPR2017-00728
`
`
`
`Ping’s code is already irregular
`POR 27-33; Sur. 2
`“[A]ftercomparing ‘regular’ Gallagercodes (like those of
`Ping)against ‘irregular Gallagercodes’, MacKay concludes
`that making an LDPC code irregular improves performance.”
`(cid:1)Ping’s parity-check matrix has nonuniform column weights, and
`so is not a “regular” code per MacKay.
`Pet. 42-43
`(cid:1)There is no motivation to combine Ping with MacKay because
`EX 2033 (Davis Dep.) 231:11-14
`Ping’s parity-check matrix is already irregular.
`POPR 11
`Patent Owner Caltech’s Oral Argument
`16
`IPR2017-00728
`
`
`
`Ping’s code is already irregular
`POR 27-33; Sur. 2-3
`(cid:1)Ping’s parity check matrix has column weights of t, 2, and 1.
`
`Ping’s parity-check matrix; MM ¶92
`
`MacKay’s regular code; EX1202, p. 1450
`
`17
`
`MacKay’s irregular code; EX1202, p. 1450
`Patent Owner Caltech’s Oral Argument
`IPR2017-00728
`
`
`
`Ping’s code is not “nearly uniform”
`POR 27-33; Sur. 2
`“Thus Ping’s parity-check matrix is not regular and its
`column weights are non-uniform. MacKay’s teaching that
`non-uniform party-check matrices perform better than
`regular parity-check matrices provide no motivation to
`modify Ping.”
`MM ¶93
`Q. So would MacKay considers codes which have some
`weight 2 columns and some weight 4 columns regular or
`irregular?
`A. And that, I didn’t form any opinionsthat in the –in my
`declaration. It is unclear to me. It could go either way.
`EX2033 (Davis Dep.) 246:2-7
`Patent Owner Caltech’s Oral Argument
`IPR2017-00728
`
`18
`
`
`
`Ping’s Hdsubmatrix is not a parity-check matrix
`POR 29-32; Sur. 2-3
`(cid:1)Petitioner’s combination fails because MacKay defines
`“irregularity” in terms of an entire parity-check matrix,
`not a portion thereof.
`(cid:1)Ping’s Hdis not a parity-check matrix because it cannot
`be used to determine whether a codeword is valid.
`EX1204 (Davis Decl.) ¶47 (cited by Pet. 14)
`
`19
`
`EX1203 (Ping) 38 (cited by MM ¶91)
`
`Patent Owner Caltech’s Oral Argument
`IPR2017-00728
`
`
`
`Ping’s Hdsubmatrix is not a parity-check matrix
`POR 31-32, Sur. 2-3
`(cid:1)Petitioner’s expert admits that Ping’s Hdis not a parity-
`check matrix
`
`20
`
`EX 2033 (Davis Dep.) 217:18-218:5
`Patent Owner Caltech’s Oral Argument
`IPR2017-00728
`
`
`
`Ping’s Hdimproves on random LDPC codes
`(MacKay)
`POR 7-9, 23-35, 49-51
`(cid:1)Petitioner’s proposed modification ignores that Ping
`presents itself as an improvement over randomly
`generated parity check matrices, such as those found in
`MacKay.
`
`21
`
`EX1203, p. 38
`Patent Owner Caltech’s Oral Argument
`IPR2017-00728
`
`
`
`Ping imposes specific “rules” for its parity-
`check matrix
`(cid:1)Ping’s proposed parity-check matrix is split into two sub-
`POR 7-9, 23-35, 49-51
`matrices, Hpand Hd, each of which is structured
`according to specific “rules.”
`
`22
`
`EX1203, p. 38
`Patent Owner Caltech’s Oral Argument
`IPR2017-00728
`
`
`
`Petitioner selectively disregards Ping’s rules
`(cid:1)Petitioner’s argument is inconsistent.
`POR 7-9, 23-35, 49-51
`(cid:1)Petitioner expressly acknowledges that a POSA would notbe
`motivated to modify the constrained structure of Hp.
`(cid:1)Yet Petitioner inconsistently contends that a POSA wouldbe
`Reply, p. 8
`motivated to modify the constrained structure of Hd.
`Reply, p. 13
`Patent Owner Caltech’s Oral Argument
`IPR2017-00728
`
`23
`
`
`
`Petitioner’s proposal violates Ping’s “rules to
`create Hd”
`(cid:1)Ping imposes specific “rules to create Hd”for
`performance reasons
`POR 7-9, 23-35, 49-51; Sur. 4
`(cid:1)Petitioner’s proposal violates these rules and ignores their
`benefits without providing any rationale to do so.
`EX1203, p. 38
`(cid:1)Petitioner does not address that its proposal would not
`“best increase the recurrence distance” or “reduce[] the
`correlation during the decoding process.”
`Patent Owner Caltech’s Oral Argument
`24
`IPR2017-00728
`
`
`
`Petitioner’s proposal violates Ping’s
`principle of operation
`POR 33-35
`(cid:1)Petitioner’s combination modifies Ping’s principle of
`operation.
`“Our predecessor court held that if a proposed modification
`or combination of the prior art would change the principle of
`operation of the prior art invention being modified, then the
`teachings of the references are not sufficient to render the
`claims prima facie obvious.” In re Gardner, 449 F. App'x914, 916 (Fed. Cir. 2011)
`(citing In re Ratti, 270 F.2d810 (C.C.P.A. 1959))
`“[A]nyproposed combination is unreasonable if it modifies
`the principle of operation of the reference relied upon.”
`Blue CoatSystems v. Finjan, Inc., IPR2017-00996, Paper7, p. 15
`Patent Owner Caltech’s Oral Argument
`IPR2017-00728
`
`25
`
`
`
`Petitioner’s rationale is illogical
`POR 28-33, 46-49; Sur. 3-5
`“[S]traightforwardfor a [POSA] to change Ping’s generator Hd
`matrix such that not all columns had the same weight …
`This would have been an easy way for one of ordinary skill to
`incorporate the irregularity disclosed by MacKay into Ping.”
`(cid:1)There is no reason to incorporate irregularity into Ping
`because Ping’s parity check matrix is already irregular.
`Pet. 44
`(cid:1)MacKay does not teach modifying a submatrix.
`(cid:1)Petitioner’s modification disregards and violates Ping’s
`“rules.”
`Patent Owner Caltech’s Oral Argument
`26
`IPR2017-00728
`
`
`
`Petitioner provides no Tanner graph analysis
`POR 22-23; Sur. 3
`(cid:1)Claim 18 requires encoding parity bits “in accordance
`with the following Tanner graph.”
`(cid:1)Petitioner does not analyze how the proposed
`combination compares with Claim 18’s Tanner graph.
`Petitioner’s proposed construction of “Tanner graph” (Pet. 55)
`Pet. 55-57
`Patent Owner Caltech’s Oral Argument
`IPR2017-00728
`
`27
`
`
`
`Petitioner provides no rationale to combine decoders
`POR 23-25; Sur. 4-5
`(cid:1)Claim 18 requires a “message passing decoder.”
`(cid:1)Petitioner provides norationale to incorporate Divsalar,
`MacKay or Luby97’s decoders into Ping.
`
`28
`
`MM ¶163
`Patent Owner Caltech’s Oral Argument
`IPR2017-00728
`
`
`
`Petitioner provides no rationale
`to combine Divsalar’srepeater
`POR 51-55; Sur. 8
`(cid:1)Incorporating Divsalar’srepeater into Ping adds needless
`complexity and requires additional memory. MM ¶¶141-42.
`(cid:1)This is confirmed by Petitioner’s attorney-made diagram
`EX1203, p. 38
`
`29
`
`Annotated Petitioner’s EX1272
`Patent Owner Caltech’s Oral Argument
`IPR2017-00728
`
`
`
`Petitioner provides no rationale
`to make Ping non-systematic
`POR 26-27; Sur. 5
`(cid:1)Claim 23 requires a non-systematic code.
`(cid:1)Making Ping non-systematic would destroy its code.
`“[T]his would require removing the Hdsubmatrix from the
`parity-check matrix H.”
`(cid:1)Hdis the very thing petition seeks to modify.
`MM ¶156
`
`30
`
`Patent Owner Caltech’s Oral Argument
`IPR2017-00728
`
`
`
`Petitioner’s proposed modification lacks
`particularity
`POR 42-45; Reply isoMTE 1
`“A large problem with the Petition and Dr. Davis’s obviousness argument
`is that they do not provide much if any specificityregarding how a
`person of ordinary skill would modify Ping in view of MacKay.”
`MM ¶118
`“A petition must identify its challenge ‘in writing and with particularity…’
`35 U.S.C. §312(a)(3); see also 37 C.F.R. §§42.22(a), 42.104(b)(4)-(5).
`(EX2004, ¶¶110-117).”
`(cid:1)Proposal: modify Ping’s Hdby “setting some columns to
`weight 9 and others to weight 3.” (Pet. 44)
`POR 42
`(cid:1)For a small codewordof 10,000 bits, there are 101500
`possibilities. MM ¶¶118-19.
`Patent Owner Caltech’s Oral Argument
`31
`IPR2017-00728
`
`
`
`Petitioner’s Reply is Improper
`
`
`
`I
`
`Petitioner’s Reply is Improper
`32
`
`
`
`Petitioner cannot introduce new improper
`evidence in Reply
`“A reply may only respond to arguments raised in the
`corresponding opposition. §42.23. While replies can help
`crystalize issues for decision, a reply that raises a new issue
`or belatedly presents evidence will not be considered and may
`be returned. The Board will not attempt to sort proper from
`improper portions of the reply. Examples of indications that a
`new issue has been raised in a reply include new evidence
`necessary to make out a prima faciecase for the patentability
`or unpatentability of an original or proposed substitute claim,
`and new evidence that could have been presented in a prior
`filing.”
`Office Patent Trial Practice Guide, 77 Fed. Reg. 48756, 48767 (Aug. 14, 2012) (emphasis added)
`
`Sur. 6
`
`33
`
`Patent Owner Caltech’s Oral Argument
`IPR2017-00728
`
`
`
`Petitioner’s new reply theories are untimely
`Sur. 6-8
`(cid:1)Petitioner tries to cure the defects identified in the POR
`with with brand new evidence.
`(cid:1)The Reply includes new weight distribution patterns not
`found in the Petition.
`(cid:1)The pattern on the left has nonuniform row weights.
`(cid:1)The pattern on the right has column weights of 4, 5, and 9.
`
`34
`
`EX1265 ¶48
`
`EX1265 ¶52
`Patent Owner Caltech’s Oral Argument
`IPR2017-00728
`
`
`
`Petitioner’s new experimental data is irrelevant
`Sur. 6-8
`(cid:1)Petitioner tested codes that were not proposed in the
`petition.
`(cid:1)Had 18+ years of hindsight, including the knowledge of
`the patents and Dr. Jin’ssource code.
`(cid:1)Matlabreceived 35 updates since May 2000.
`(cid:1)Irregular codes used twiceas many decoding iterations.
`(cid:1)Did not discuss whether experiments were comparable
`to what a POSA at the relevant timeframe would have
`done.
`Patent Owner Caltech’s Oral Argument
`35
`IPR2017-00728
`
`
`
`Petition Articulates No Reasonable
`Expectation of Success
`
`36
`
`
`
`Obviousness inquiry requires REOS
`POR 16-17, 46
`“The combination of familiar elements according to
`known methods is likely to be obvious when it does
`no more than yield predictable results.”
`KSR Intern. Co. v. Teleflex, Inc., 550 U.S. 398 (2007)
`“Although predictability is a touchstone of obviousness, the
`‘predictable result’ discussed in KSR refers not only to the
`expectation that prior art elements are capable of being
`physically combined, but also that the combination would
`have worked for its intended purpose.”
`DePuySpine, Inc. v. Medtronic SofamorDanek, Inc.,
`567 F.3d 1314, 1326 (Fed. Cir. 2009)
`“Obviousness requiresa reasonable
`expectation of success.”
`MPEP 2143.2.I
`Patent Owner Caltech’s Oral Argument
`IPR2017-00728
`
`37
`
`
`
`Petition provided no analysis of REOS
`POR 16-17, 46-51, 55; Sur. 5-6; Reply isoMTE 1
`(cid:1)The petitions provided no analysis as to whether the
`proposed combinations would have a reasonable
`expectation of success or yield predictable results.
`“JTEKT’s rationale that an artisan of ordinary skill would have
`been motivated to reduce weight, without further persuasive
`evidence why such a weight loss would have been predictably
`realized, or at least expected, amounts to an unsupported
`conclusory assertion.”
`JTEKT Corp. v. GKN Automotive, Ltd., IPR2016-00046,
`Paper No. 27 at 28-29 (Jan. 23, 2017)
`Patent Owner Caltech’s Oral Argument
`IPR2017-00728
`
`38
`
`
`
`Unpredictability is undisputed
`POR 4-6, 46-47; Sur. 6; Reply isoMTE 2
`“[W]hat you would really like to be able to do is a
`formal mathematical analysis of the strength of the
`codes that you are working with, but that’s often
`really hard … [I]t might even be impossible to do
`the mathematical analysis.”
`EX2033 (Davis Dep.) 256:21-258:12
`(cid:1)Unpredictability contradicts any assertion that suggested
`modification would have been “straightforward.”
`Reply 14
`(cid:1)See also MM ¶¶46-49, 126-131.
`Patent Owner Caltech’s Oral Argument
`39
`IPR2017-00728
`
`
`
`Unpredictability is undisputed
`POR 7, 48-49
`(cid:1)Even in MacKay’s small test sample, several irregular
`codes resulted in highly undesirable error floors.
`
`EX1202 (MacKay) 1450, 1452
`
`40
`
`EX 2033 (Davis Dep.) 261:17-262:13
`Patent Owner Caltech’s Oral Argument
`IPR2017-00728
`
`
`
`Objective Indicia of Non-obviousness
`
`41
`
`
`
`Objective Indicia of Non-Obviousness
`POR 55-56
`“As this court has repeatedly explained, this evidence is not
`just a cumulative or confirmatory part of the obviousness
`calculus but constitutes independent evidence of
`nonobviousness.”
`Ortho-McNeil Pharm., Inc. v. Mylan Labs., Inc., 520 F.3d1358, 1365 (Fed. Cir. 2008) (emphasis added)
`“This case illustrates a good reason for considering objective
`indicia as a critical piece of the obviousness analysis:
`Objective indicia ‘can be the most probative evidence of
`nonobviousness in the record, and enables the court to avert
`the trap of hindsight.’”
`Leo Pharm. Prod., Ltd. v. Rea, 726 F.3d1346, 1358 (Fed. Cir. 2013) (emphasis added)
`Patent Owner Caltech’s Oral Argument
`IPR2017-00728
`
`42
`
`
`
`IRA codes are commercially successful
`
`POR 55-56
`
`43
`
`MM ¶168
`Patent Owner Caltech’s Oral Argument
`IPR2017-00728
`
`
`
`Reply isoMTE 5
`
`J. Pfaelzerdid not find noninfringement
`(cid:1)Caltech was moving party.
`(cid:1)Judge Pfaelzerleft issue of infringement for jury to decide.
`EX1267, *2
`Id., *5.
`44
`
`Patent Owner Caltech’s Oral Argument
`IPR2017-00728
`
`
`
`J. Pfaelzerdecision based on narrow
`construction
`MTE 9-10
`(cid:1)Judge Pfaelzernarrowly construed “repeat” to preclude
`“re-use” of a bit.
`(cid:1)Judge Wu in current case rejected this construction.
`California Institute of Technology v. Hughes Communications Inc.,
`No. 2:13-cv-07245-MRP-JEM, Dkt. 105, p. 14 (C.D. Cal. Aug. 6, 2014)
`California Institute of Technology v. Broadcom Limited, et al.,
`No. 2:16-cv-03714-GW-AGRx, Dkt. 213, p. 1 (C.D. Cal. Jul. 12, 2017)
`Patent Owner Caltech’s Oral Argument
`IPR2017-00728
`
`45
`
`
`
`Praise for IRA codes
`
`46
`
`POR 61-63
`
`EX2008, p. 196
`
`EX2010, p. 23
`Patent Owner Caltech’s Oral Argument
`IPR2017-00728
`
`
`
`DVB-S2 performance credited to IRA codes
`
`POR 56-57
`
`EX2006, p. 1
`
`EX2007, p. 1
`Patent Owner Caltech’s Oral Argument
`IPR2017-00728
`
`47
`
`
`
`DVB-S2’s commercial success
`
`POR 64-66
`
`EX2013, p. 2
`
`48
`
`EX2014, p. 1
`Patent Owner Caltech’s Oral Argument
`IPR2017-00728
`
`