`
`Paper No. ___
`Filed: March 19, 2018
`
`
`
`
`
`
`
`
`
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`_____________________________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`_____________________________
`
`
`APPLE INC.,
`Petitioner,
`
`v.
`
`CALIFORNIA INSTITUTE OF TECHNOLOGY,
`Patent Owner.
`_____________________________
`
`Case IPR2017-00219
`Patent No. 7,116,710
`
`
`
`_____________________________
`
`
`
`PATENT OWNER’S SURREPLY
`
`
`
`TABLE OF CONTENTS
`
`I.STATEMENT OF PRECISE RELIEF REQUESTED .......................................... 1
`II.ARGUMENT ...................................................................................................... 1
`A.
`Luby does not teach irregular repetition of information bits ................ 1
`B.
`hindsight, and the corresponding argument unavailing........................ 3
`C.
`The proposed combination does not disclose “partitioning” ................ 5
`D.
`Caltech does not bear the burden of proving no motivation................. 5
`E.
`Reasonable expectation of success was never addressed ..................... 7
`The new experimental data should be rejected .................................... 7
`F.
`III.CONCLUSION .................................................................................................. 8
`
`The attorney-generated graphs are erroneous and tainted with
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`-i-
`
`
`
`
`I.
`
`STATEMENT OF PRECISE RELIEF REQUESTED
`
`In view of new argument and evidence submitted in Petitioner’s Reply
`
`briefing, the Board (Papers 50, 52, 54) authorized a short sur-reply but prohibited
`
`submission of rebuttal evidence. The Reply materials are replete with untimely and
`
`improper new argument and evidence—including submission of new data,
`
`attorney-generated Tanner graphs, and a declaration from a new witness. The
`
`Reply (2) provides no reasonable justification for replacing Dr. Davis with a new
`
`witness. Dr. Davis was aware of his Fulbright commitment since at least February
`
`2017 and he testified he remains available for deposition in the U.S. EX1056, ¶3.
`
`Accordingly, the Reply materials should be disregarded and given no weight.
`
`II. ARGUMENT
`
`A. Luby does not teach irregular repetition of information bits
`
`The Petition presented the misguided theory that “the irregularity of Luby”
`
`could be incorporated into the RA codes by modifying Divsalar’s repeater. See,
`
`e.g., Pet. 37, 46. The POR explains, inter alia, that (1) the “irregularity of Luby”
`
`refers to a bipartite graph in which the codeword is irregular (repetition of
`
`information bits is a fundamendally different concept); (2) Divsalar’s codeword
`
`already exhibits “irregularity of Luby”; and (3) modifying the repeater would not
`
`affect the codeword’s irregularity. E.g., EX2004 ¶¶84-86; POR 2, 19-26.
`
`The Reply (1-3) asserts for the first time that Luby teaches irregular
`
`repetition of information bits. This is an improper shift away from the case in the
`-1-
`
`
`
`
`petition. Moreover, the contention is wrong, relying on flawed analysis, wholly
`
`unsupported assumptions and a linguistic shell game to confuse the terminology.
`
`First, Petitioner fails to acknowledge that its original expert Dr. Davis flatly
`
`conceded that Luby does not disclose irregular repetition of information bits.
`
`EX2033 196:2-198:15 (testifying, e.g., that a message node’s degree “doesn’t
`
`answer the question of whether that information bit is repeated or not”).
`
`Beyond that, the entire argument is based on the faulty assumption that
`
`“Luby’s codewords contain 8,000 information bits and 8,000 parity bits.” Reply 1.
`
`The statement is a conclusory and unexplained nonsequitur to the preceding
`
`citation to Luby, which refers neither to parity bits or information bits. And the
`
`Reply neglects to mention that Dr. Mitzenmacher (a co-author of Luby)
`
`specifically rejected the assertion at his deposition. EX1262 215:21-217:25; see
`
`also, EX2004 ¶77; EX1204, 256 (“We do not perform an actual encoding, but
`
`instead … use an initial message consisting entirely of zeroes.”). To the extent the
`
`Reply is conflating “check bits” with “parity bits,” Dr. Mitzenmacher explained
`
`this is erroneous. EX1262 216:25-217:21. Of course, the Reply forces Caltech and
`
`the Board to speculate as to the basis of its stated assumption since no reason for
`
`the conclusion is ever given. But the petitioner bears the unshifting burden of
`
`proof, and the entire argument collapses absent the unfounded assumption of 8,000
`
`information bits in Luby’s codeword.
`
`-2-
`
`
`
`
`The Reply (2-3) argues that Luby discloses irregular repetition because “a
`
`POSA [would] preferentially use higher degrees for information bits.” Again, this
`
`relies on the erroneous assumption that Luby discloses codewords containing 8,000
`
`parity bits and 8,000 information bits (see above).1 And Dr. Mitzenmacher rebutted
`
`this hypothetical during his deposition, explaining that it would not be beneficial,
`
`and likely harmful, to preferentially resolve information bits relative to parity bits
`
`because the entire codeword must still be decoded. EX1262 232:25-233:22.
`
`B.
`
`The attorney-generated graphs are erroneous and tainted with
`hindsight, and the corresponding argument unavailing
`
`Divsalar’s code already has variance in its codeword (i.e., “the irregularity
`
`of Luby”) and modifying Divsalar’s repeater would have no effect on its codeword
`
`variance—this is plainly evident in the modified Khandekar graph in the petition.
`
`EX2004 ¶¶85-86; POR 28-29, 40-41. The Reply (3-5) offers no response on this
`
`point.2 Besides repeating the same misguided assertion about changing the
`
`
`1 The Reply (2-3) also confuses the terms “message bits” with “information
`
`bits.” The “message bits” in Luby are different than information bits. EX2004 ¶76;
`
`EX2033 190:1-6; EX1264 152:20-153:6. Moreover, the whole notion of
`
`“preferential assignment” in the Reply argument is fundamentally at odds with
`
`Luby’s “randomly choosen graphs.” EX1204 249.
`
`2 The Reply (11) statement that accumulators were known is nonresponsive.
`
`-3-
`
`
`
`
`repeater (which would not affect codeword irregularity), the Reply offers new
`
`attorney-created graphs and nonsensical commentary.
`
`As an initial matter, the graphs (Reply 5: EX1246, 1247 as well as EX1257,
`
`1258) are not in the petition and should be rejected as untimely new argument.
`
`Even if considered, simply because Petitioner’s lawyers generated similarly-styled
`
`graphs in preparing the Reply in no way demonstrates that the underlying codes
`
`would have been combined 18 years ago or that a POSA would have expected such
`
`a modification to be successful in improving error-correction. Moreover, the logic
`
`in the Reply (5) argument—i.e., that a POSA would desire circles on the left side
`
`of one graph to look more like circles on the left side of a different graph—is
`
`flawed. The Reply compares different “circles” representing different elements.
`
`Luby’s “message nodes” refer to the codeword bits—which for EX1246 are the
`
`nodes on the right labelled “X1…” This is a nonsensical argument based on
`
`attorney generated graphs, not the logic of a POSA.
`
`Furthermore, the new exhibits are tainted with impermissible hindsight.
`
`Prior to Caltech’s IRA patents, Tanner graphs were conventionally presented as
`
`bipartite graphs depicting the relationship between a codeword and parity check
`
`equations. See, e.g., Fig. 2 of Luby (EX1204 253); see also EX2004 ¶40; EX1206
`
`¶¶55-56; EX1264 81:6-23. Caltech’s inventors used Tanner graphs in a novel way
`
`to illustrate an encoding process. See ’710 patent, Fig. 3; EX2031 ¶15. Petitioner
`
`-4-
`
`
`
`
`provides no evidence that a POSA would have applied Caltech’s novel Tanner
`
`graph representation to any code at the time.
`
`In addition to the improper new graphs, the Reply (3-4) falsely claims that
`
`applying Luby’s “fundamental teachings” would result in irregular repetition. But
`
`Luby’s fundamental teaching is that its codeword bits have irregular degrees.
`
`EX2004 ¶63; EX204 249. To the extent the Reply presents “the irregularity of
`
`Luby” as some abstract concept, that is unsupported by the reference and a notion
`
`soundly rejected by Dr. Mitzenmacher. See, e.g., EX1262 249-250.
`
`Moreover, Frey (Reply 6) does not save the day for Petitioner. Frey presents
`
`a non-enabling disclosure that, if anything, establishes that finding irregular
`
`repetition was a difficult and unpredictable parameter. EX2004 ¶¶65-69.
`
`C. The proposed combination does not disclose “partitioning”
`
`The Reply (6-7) now argues that it is not making an inherency argument
`
`with regard to “partitioning.” Yet, the Reply continues to claim that “irregular
`
`repetition ‘de facto’ causes partitioning,” which could not be interpreted as
`
`anything but an inherency theory. Id. Dr. Mitzenmacher’s testimony on this point
`
`remains unrebutted. POR 29-30; EX2004 ¶ 94; EX1062 376:8-21. The Reply(7) is
`
`wrong to make it Caltech’s burden to show his example is “practical or likely,” and
`
`Petitioner’s suggestions of probabilities and possibilities are insufficient.
`
`D. Caltech does not bear the burden of proving no motivation
`
`-5-
`
`
`
`
`Remarkably, the Reply (8-11) argues that “Caltech has failed to show that a
`
`POSA would not have combined Divsalar and Luby.” But the petitioner bears the
`
`unshifting burden of demonstrating unpatentability—the burden is not on Caltech.
`
`The Reply (9, bottom of 10) states that “Divsalar could be made irregular.”
`
`While the assertion correctly characterizes Petitioner’s obviousness theory, the
`
`assertion is factually misguided and legally irrelevant. See, e.g., EX1262 240:19-
`
`241:21; EX2031 at, e.g., ¶¶33-37; EX1264 60:1-21. Petitioner clings to assertions
`
`of “could have” and the conceptually “simple,” but never establishes that a POSA
`
`would have so modified Divsalar with any reasonable expectation of success.
`
`The Reply (12) mischaracterizes “the basis” of Dr. Divsalar’s testimony and
`
`misses the point—there is no reason why. Modifying Divsalar’s RA code to add
`
`irregular repetition would have been adding complexity without expected benefit
`
`(or even expectation the code would function). EX2031¶¶33-35; EX1264 60:1-21
`
`(“[H]ypothetically someone can … do anything crazy. But there should be some
`
`motivation by doing so.”). One of issues with Petitioner’s modification identified
`
`by Dr. Divsalar (not “the basis”) was further slowing an already slow RA code.
`
`Petitioner’s pivot to a new RA code modification theory underscores the point. The
`
`petition originally proposed a modified RA code with a 1/3 rate (where q=3 on
`
`average), whereas the new theory proposes a modified RA code with a much
`
`slower 1/5 rate (where q=5 on average). And the Reply wholly ignores that Luby’s
`
`-6-
`
`
`
`
`codes had a rate of ½. EX1204, 256.
`
`E. Reasonable expectation of success was never addressed
`
`The POR (3-4, 18-19, 44-45) points out that the petition materials wholly
`
`lack discussion of reasonable expectation of success or the (now conceded)
`
`unpredictability in the field. As such, any discussion of reasonable expectation of
`
`success presented in the Reply is improperly new and should be ignored. 37 C.F.R.
`
`§42.23(b); Office Patent Trial Practice Guide, 77 Fed. Reg. 48767 (Aug. 14, 2012).
`
`F.
`
`The new experimental data should be rejected
`
`The new experimental data (EX1268) and corresponding testimony
`
`(EX1265) and argument (Reply 10) should be disregarded for a number of reasons.
`
`First, the material is improper as presenting an entirely new theory of
`
`unpatentability. The petition only proposed modifying a repeat-three RA code to
`
`one in which information bits are repeated 2 and 4 times, which Caltech addressed
`
`in detail (e.g., POR 39-44). Petitioner now abandons this theory and pivots to a
`
`new one—repeating information bits 3 and 7 times. Reply 10.
`
`Second, the experimental data materials are not contemporaneous with the
`
`“relevant time” and are irrelevant to the understanding of a POSA. It is completely
`
`irrelevant what Dr. Frey claims he can do in the year 2018 when armed with
`
`Caltech’s patent disclosures and publications, Dr. Jin’s original coding work,
`
`contemporary resources (e.g., Dr. Frey (¶51) used Matlab, a software program that
`
`-7-
`
`
`
`
`received over 35 version updates since May 2000), and 18 years of post-filing date
`
`knowledge. Dr. Frey’s testimony presents zero reflection of the environment in
`
`1999-2000, and provides no information as to why a POSA would make the
`
`proposed modification 18 years ago or reasonably expected success at that time.
`
`If surreply evidence was permitted, Caltech’s witnesses would have been
`
`able to provide rebuttal testimony explaining the data’s unreliability—including
`
`numerous technical flaws, evident cherry-picking of parameters, and unexplained
`
`selection of a single degree profile that in no way flows from any art reference at
`
`issue in this case. For example, the Reply materials provide no justification for the
`
`Eb/No Gaussian noise parameters highlighted in ¶47 of Dr. Frey’s declaration (-
`
`.8, .1, .8). And there is a disconnect between the identified simulation parameters
`
`and the “plot” depicted in the graphs in ¶53 (data points between Eb/No 0.2 and
`
`0.8). While the “plot” in ¶53 forms the basis of Petitioner’s comparison, the origin
`
`of the plotted points is entirely unknown and inconsistent with the identified
`
`simulation parameters (¶53 does not even assert the “plot” was generated from the
`
`identified simulation). EX1068 is misleadingly presented as a single document
`
`when it is, in fact, an amalgamation of five independent documents. The new data
`
`is untimely, unexplained, unreliable and fails to comply with 37 C.F.R. §42.65.
`
`III. CONCLUSION
`
`Accordingly, the Reply materials should be given no weight.
`
`-8-
`
`
`
`
`
`
`Date: March 19, 2018
`
`
`
`
`
`Respectfully submitted,
`
`
`
`/ Michael T. Rosato /
`Michael T. Rosato, Lead Counsel
`Reg. No. 52,182
`
`
`
`-9-
`
`
`
`
`CERTIFICATE OF SERVICE
`
`I certify that the foregoing Patent Owner’s Surreply was served on this 19th
`
`day of March, 2018, on the Petitioner at the electronic service addresses of the
`
`Petitioner as follows:
`
`
`Richard Goldenberg
`Dominic Massa
`Michael H. Smith
`James M. Dowd
`Mark D. Selwyn
`Arthur Shum
`WILMER CUTLER PICKERING HALE AND DORR LLP
`richard.goldenberg@wilmerhale.com
`dominic.massa@wilmerhale.com
`michaelh.smith@wilmerhale.com
`james.dowd@wilmerhale.com
`mark.selwyn@wilmerhale.com
`arthur.shum@wilmerhale.com
`
`
`
`
`
`
`Date: March 19, 2018
`
`
`
`
`
`
`
`
`Respectfully submitted,
`
`
`/ Michael T. Rosato /
`Michael T. Rosato, Lead Counsel
`Reg. No. 52,182
`
`
`
`-10-
`
`
`