Paper No. ___
`Filed: April 7, 2017
`
`
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`_____________________________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`_____________________________
`
`
`
`APPLE INC.,
`Petitioner,
`
`v.
`
`CALIFORNIA INSTITUTE OF TECHNOLOGY,
`Patent Owner.
`_____________________________
`
`Case IPR2017-00297
`Patent No. 7,916,781
`
`_____________________________
`
`
`
`PATENT OWNER’S PRELIMINARY RESPONSE
`PURSUANT TO 37 C.F.R. § 42.107
`
`

`

`TABLE OF CONTENTS
`
`I.INTRODUCTION .............................................................................................. 1
`II.THE PRESENT PETITION RECYCLES PREVIOUS CHALLENGES
`REJECTED BY THE OFFICE ................................................................... 2
`
`A. Ground 2 Recycles the Same Ping-Based Anticipation
`Previously Rejected .......................................................................... 4
`Grounds 1, 3 and 4 are Substantially the Same.................................. 5
`
`B.
`
`III.THREE OF FOUR GROUNDS IN THE PETITION CAN BE DENIED
`BECAUSE THEY RELY ON IMPROPER EVIDENCE AND
`UNSUBSTANTIATED ALLEGATIONS ................................................ 10
`IV.THE PROPOSED GROUNDS OF CHALLENGE FAIL .............................. 12
`
`A. Ground 1 Fails ................................................................................ 12
`
`1. The single equation of Ping fails to teach the first and second
`encoding operations as recited in claim 1 ..................................................... 13
`2. The petition does not demonstrate motivation to combine ........... 17
`3. The references fail to disclose limitations in claim 5 ................... 24
`4. The references fail to disclose limitations in claim 8 ................... 25
`
`B.
`
`Ground 2 Fails ................................................................................ 26
`
`1. Ping fails to disclose the “receiving” steps of independent claims
`19-21………………………………………………………………………….27
`2. Ping fails to disclose “wherein at least two of the information bits
`appear in three subsets of the information bits,” as recited in independent claim
`19……………………………………………………………………………..30
`
`C.
`
`Ground 3 Fails ................................................................................ 30
`
`1. The references fail to disclose “wherein the information bits appear
`in a variable number of subsets” ................................................................... 31
`2. There is no reason to combine the teachings of Ping, MacKay, and
`Divsalar as proposed .................................................................................... 37
`
`D. Ground 4 Fails ................................................................................ 39
`
`V.CONCLUSION ............................................................................................... 40
`
`-i-
`
`

`

`I.
`
`INTRODUCTION
`
`The Board should not institute inter partes review (IPR) on claims 3-12 and
`
`19-21 of U.S. Patent No. 7,916,781 (“the ’781 patent”) because petitioner Apple
`
`Inc. (“Petitioner” or “Apple”) has not met its burden of showing that it has a
`
`reasonable likelihood of prevailing on any of its proposed grounds of
`
`unpatentability.
`
`As an initial matter, review should be denied on the basis that the present
`
`petition rehashes substantially the same art and arguments that have already been
`
`presented to the Office and rejected by the Board in a previous IPR challenge.
`
`Petitioner acknowledges that the ’781 patent was already “challenged in one
`
`petition for inter partes review.” Pet. at 1; see also IPR2015-00059. The previous
`
`petition similarly relied on Ping alone or in view of the Luby ’909 Patent or the
`
`Patterson ’999 Patent. The Board denied institution as to all grounds that relied on
`
`Ping alone or in view of other references. Hughes Network Systems, LLC v.
`
`California Institute of Tech., Case No. IPR2015-00059, Paper 18 at 14-16. In this
`
`instance, Petitioner essentially recycles the same Ping-based arguments. The
`
`current Ground 2 (anticipation based on Ping) has no substantive differences
`
`compared to what the Board previously considered and rejected. For the remaining
`
`grounds, the petition adds Divsalar as a contingency plan to the same
`
`Ping/anticipation argument, and substitutes MacKay for the Luby ’909 Patent and
`
`-1-
`
`

`

`Coombes for the Patterson ’999 Patent to present substantially the same disclosures
`
`and arguments that the Board considered and rejected in the prior petition.
`
`The proposed grounds of challenge should also be rejected on the merits.
`
`The petition materials fail to demonstrate that each feature of claims 3-12 and 19-
`
`21 of the ’781 patent is found in the cited art. Multiple aspects of the claimed
`
`subject matter are missing from the asserted references.
`
`Moreover, the Petitioner has failed to demonstrate that one of ordinary skill
`
`in the art would reasonably have combined the references as proposed. The
`
`arguments advanced in the petition essentially amount to little, if anything, beyond
`
`assertions that the cited references are analogous art. The Board has rejected such
`
`assertions as insufficient motivation to specifically combine teachings in particular
`
`ways. Any additional explanation provided in the petition is insufficient, lacks the
`
`requisite logical underpinnings, or both, and should be dismissed.
`
`Accordingly, institution of inter partes review should be denied.
`
`II. THE PRESENT PETITION RECYCLES PREVIOUS
`CHALLENGES REJECTED BY THE OFFICE
`
`The instant petition presents one in a series of challenges to the ’781 patent,
`
`but rehashes substantially the same art and arguments already presented to the
`
`Office and rejected by the Board. Accordingly, the Board should exercise its
`
`discretion in denying institution on all grounds in the petition. 35 U.S.C. § 325(d)
`
`-2-
`
`

`

`(“In determining whether to institute or order a proceeding under this chapter,
`
`chapter 30, or chapter 31, the Director may take into account whether, and reject
`
`the petition or request because, the same or substantially the same prior art or
`
`arguments previously were presented to the Office.”) (emphasis added).
`
`The present petition fails to offer any art or arguments substantially different
`
`from what the Board has already rejected. Petitioner acknowledges that the’781
`
`patent was already “challenged in one petition for inter partes review.” Pet. at 1. In
`
`the prior petition, the Board rejected grounds substantially the same as the grounds
`
`Petitioner presents in this instance. See Hughes Network Systems, LLC v.
`
`California Institute of Tech., Case No. IPR2015-00059, Paper 18 (Apr. 27, 2015).1
`
`For example, the rejected grounds in the Hughes petition challenged claims
`
`1-7, 13-16, and 19 as either anticipated by Ping or obvious over Ping in view of the
`
`’999 patent (U.S. Patent No. 4,623,999) and/or the Luby ’909 patent (U.S. Patent
`
`No. 6,081,909). Id. at 8. Here, each of the grounds in the instant petition challenges
`
`claims 1-8, 10-12, and 19-21 as either anticipated by the same Ping reference or
`
`obvious over the Ping reference in view of Divsalar. Compare Pet. at 34-70 with
`
`
`
`1 Petitioner has also filed another pending petition challenging claims of the
`
`’781 patent. See Apple Inc. v. Cal. Inst. of Tech., Case No. IPR2017-00423, Paper
`
`5 (PTAB Dec. 12, 2016).
`
`-3-
`
`

`

`IPR2015-00059, paper 1 at 16-47. For certain dependent claims, Petitioner in this
`
`instance substitutes MacKay for the Luby ’909 Patent and Coombes for the
`
`Patterson ’999 Patent. As addressed below, the asserted art and arguments are not
`
`meaningfully different compared to what was advanced in the Hughes petition, and
`
`Petitioner fails to explain why consideration of the same or substantially argument
`
`is warranted here.
`
`A. Ground 2 Recycles the Same Ping-Based Anticipation Previously
`Rejected
`
`As an initial matter, Hughes’s Ping-based anticipation ground of claim 19
`
`has been recycled completely here as an anticipation ground against claims 19-21.
`
`As can be seen in current Ground 2, the petition (Pet. at 57-60) relies on the same
`
`Ping anticipation ground for claims 19-21 as did Hughes for claim 19 (see Hughes
`
`IPR2015-00059, Paper 4 at 27-31). Petitioner, however, does not explain how its
`
`challenge is meaningfully different compared to the challenge advanced by Hughes
`
`and rejected by the Board.
`
`To the extent there is any meaningful difference between Hughes’ Ping-
`
`based anticipation attack and the one advanced here, it is only that the present
`
`Petitioner has less to say on the topic. That is, while Hughes devoted nearly four
`
`full pages to its failed challenge of claim 19 (see Hughes IPR2015-00059, Paper 4
`
`at 27-31), Petitioner here devotes one page to the same, with one sentence for each
`
`-4-
`
`

`

`element. Pet. at 57. These sentences assert without explanation that Petitioner’s
`
`reasoning was described in the claims of Ground 1 (which is an obviousness
`
`ground, not an anticipation ground, and addresses different claims containing non-
`
`identical language). Id. Petitioner takes the same tack for claims 20 and 21,
`
`dispensing with each in slightly more than a single page. Id. at 58-60.2 In the
`
`context of a §325(d) analysis, Petitioner’s brevity underscores the lack of argument
`
`beyond what has already been considered and rejected by the Board.
`
`B. Grounds 1, 3 and 4 are Substantially the Same
`
`Petitioner’s obviousness theory in Ground 1 (challenging claims 1-3, 5-8, 10
`
`and 12) is constructed based on the same application of Ping to the challenged
`
`claims as was previously asserted in the Hughes case and rejected by the Board.
`
`Petitioner alleges that Ping teaches all aspects of the challenged claims, but adds
`
`Divsalar as a contingency in case the Ping does not teach its encoding in two
`
`distinct steps. See Pet at 38 (“…even if Ping, alone, did not teach encoding in two
`
`distinct steps (which it does)…. Divsalar teaches encoding in two distinct steps.”).
`
`
`
`2 The failure to sufficiently explain its mapping between challenged claim
`
`elements and disclosure of Ping at least fails to comply with the requirements of
`
`Bd.R. 42.104(b)(2) (“The petition must specify where each element of the claim is
`
`found in the prior art patents or printed publications relied upon.”).
`
`-5-
`
`

`

`The current petition provides no reason to disturb the Board’s previous
`
`findings regarding Ping vis-à-vis the ’781 patent claims. The Board in Hughes
`
`concluded that “we are not persuaded by Petitioner that Ping’s LDPC codes meets
`
`limitation [b] and [c] of claim 1” IPR2015-00059, Paper 18 at 16. These limitations
`
`are the first and second encoding operations recited in claim 1. Id. at 7. Despite the
`
`Board’s previous conclusion that Ping does not disclose the claimed two-step
`
`encoding process, Petitioner nonetheless repeatedly asserts that “Ping teaches two
`
`encoding steps” in the manner claimed. (Pet. at 37; see also id. at 25, 42, 66.) The
`
`only support for this assertion is Ping’s Equation (4). Id. at 25-26. Yet this is the
`
`same disclosure that the Board rejected as not showing the claimed two-step
`
`encoding process in Hughes. See IPR2015-00059, 18 at 15-16; see also id. Paper 4
`
`at 19-21. Petitioner draws boxes around the terms of Ping’s Equation (4) (Pet. at
`
`25) that were not explicitly drawn in the Hughes petition, but drawing boxes
`
`around the same terms of Ping’s equation does not change the disclosure of Ping.
`
`Accordingly, instead of remedying the deficiencies of the Hughes IPR, Petitioner
`
`has repeated the same failed art and argument with a different veneer.
`
`Moreover, Petitioner fails to explain how the addition of Divsalar to Ping
`
`remedies the deficiencies the Board identified with respect to substantially the
`
`same claims in Hughes. For example, Petitioner here identifies the same portion of
`
`Ping’s parity check matrix as did Hughes and concludes, with insufficient
`
`-6-
`
`

`

`explanation, that portion of a parity check matrix is a generator matrix. E.g., Pet. at
`
`35. Yet, as explained below in Section IV.A.1, Petitioner’s own definition of
`
`“generator matrix” contradicts its assertions. Petitioner thus fails to establish why
`
`Ping in view of Divsalar overcomes the deficiencies of the Ping-based grounds in
`
`the rejected Hughes petition.
`
`With respect to current Grounds 3 and 4, as explained above, the present
`
`petition differs in part from the Hughes IPR petition in that the Luby ’909 Patent
`
`(Hughes Ex. 1016) has been replaced with the MacKay reference (Ex. 1002), and
`
`the ’999 Patent (Hughes Ex. 1027) has been replaced with the Coombes reference
`
`(Ex. 1018).
`
`However, the substitution of Luby ’909 with MacKay in this case does not
`
`meaningfully distinguish from what the Board already considered and rejected in
`
`the Hughes case. That is because MacKay is cited here for the same “irregularity”
`
`teaching as was Luby ’909 previously. See Pet. at 30 (“As MacKay concluded,
`
`‘[t]he excellent performance of irregular Gallager codes…’”). The petition fails to
`
`explain whether there is any meaningful difference between the “irregularity”
`
`disclosed in MacKay compared to Luby ’909. Indeed, overlap is explicit in
`
`MacKay as, with regard to “irregularity,” that reference specifically cites the
`
`corresponding Luby article describing the same subject matter as the Luby ’909
`
`patent. See Ex. 1102, at 1449 (“The excellent performance of irregular Gallager
`
`-7-
`
`

`

`codes is the motivation for this paper… The irregular codes of Luby,
`
`Mitzenmacher, Shokrollahi and Spielman [5]…”). As such, Petitioner fails to
`
`adequately explain why it is not just presenting the same or substantially the same
`
`art and argument, despite swapping Luby ’909 with MacKay for the very same
`
`technical disclosure.
`
`The replacement of the ’999 patent with Coombes also provides no
`
`meaningful difference to the Hughes case as each is provided for substantially the
`
`same disclosure.
`
`While the petition acknowledges the previous Hughes IPR, the petition lacks
`
`any explanation regarding the similarity of its challenges compared to those
`
`previously advanced in the Hughes cases. Petitioner has not explained why either
`
`the present petition or the -00059 petition substantially differ either from each
`
`other, or from the -00423 petition against the ’781 patent, such that the additional
`
`burden on Caltech and Board resources is reasonable. The Board has previously
`
`rejected IPR challenges where facially redundant challenges were advanced with
`
`no explanation why additional Board review is warranted. Cf. Maxell Co., Ltd.,
`
`Case No. IPR2015-01511, Paper 10 at 15-16 (factoring in failure to explain the
`
`need for redundant proceedings).
`
`Although the instant petition challenges some claims not challenged by the
`
`Hughes petition or the -00423 petition, the Board has indicated that overlap in
`
`-8-
`
`

`

`challenged claims is not a determinative factor and has denied institution under 35
`
`U.S.C. § 325(d) in similar scenarios. See, e.g., Ford Motor Co. v. Paice LLC, Case
`
`No. IPR2015-00767, Paper 14 at 7 (PTAB Aug. 18, 2015) (“We also have
`
`considered Petitioner’s arguments that because it presents a new set of claims, e.g.,
`
`consisting of previously challenged claims, but also including newly challenged
`
`claims, we must consider the Petition…. We are not persuaded by this argument
`
`because the express language of 35 U.S.C. § 325(d) does not mention claims as
`
`being a factor in deciding whether to institute trial.”). See also Blue Coat Systems
`
`LLC v. Finjan, Inc., Case No. IPR2016-01443, Paper No. 13 at 10 (serial petition
`
`being directed to different claims denied institution under §325(d)).
`
`Given the significant overlap between the asserted art and corresponding
`
`argument in the present petition compared to the previous Hughes petition, it is
`
`immaterial that the earlier petition was filed by a different party.3 See, e.g., Apple,
`
`Inc. v. Limestone Memory Sys. LLC, Case No. IPR2016-01567, Paper 11 at 9-12
`
`(PTAB Jan 18, 2017). In Limestone Memory Sys., Apple similarly filed an IPR
`
`petition following denial of a petition against the same patent that had been filed
`
`by an unrelated party. The Board rejected Apple’s challenge to a claim based on
`
`
`
`3 Counsel for Petitioner Apple also represented Hughes in the previous Hughes
`
`district court litigation.
`
`-9-
`
`

`

`§ 325(d), noting Apple’s failure to “identify new arguments based on additional
`
`references,” to “make persuasive assertions that differentiate its Petition from the
`
`[earlier] Petition,” or to explain why the prior art relied upon by Apple “which is
`
`substantially the same prior art as that previously presented, does not have the
`
`earlier noted deficiency.” Id. at 12.
`
`Accordingly, because the instant petition presents the same or substantially
`
`the same prior art or arguments previously presented to the Office and because
`
`Petitioner appears to have used previously rejected grounds as a roadmap to bolster
`
`its petition, with the effect of harassing Caltech, the Board should exercise its
`
`authority under § 325(d) to deny the petition.
`
`III. THREE OF FOUR GROUNDS IN THE PETITION CAN BE
`DENIED BECAUSE THEY RELY ON IMPROPER EVIDENCE
`AND UNSUBSTANTIATED ALLEGATIONS
`
`The petition identifies Divsalar as prior art under 35 U.S.C. §102(b), relying
`
`on two sources of support: (1) a Board decision in IPR2015-00059; and (2) Ex.
`
`1019, which is a declaration from a librarian (Ms. Fradenburgh) that Hughes
`
`submitted in the IPR2015-00059 case. Pet. at 31 (citing Ex. 1011 and Ex. 1019).
`
`However, Petitioner has not established that Divsalar is actually a prior art printed
`
`publication.
`
`As an initial matter, Ms. Fradenburgh is not a witness in this proceeding, and
`
`Petitioner offers the report as evidence to prove the truth of the matter asserted—
`
`-10-
`
`

`

`that Divsalar is 102(b) prior art to the ’781 patent. As such, Ex. 1019 is
`
`inadmissible hearsay. Fed. R. Evid. 801(c) (defining hearsay as “a statement that:
`
`(1) the declarant does not make while testifying at the current trial or hearing; and
`
`(2) the party offers in evidence to prove the truth of the matter asserted in the
`
`statement”). Ms. Fradenburgh’s availability as a witness, in fact, was a source of
`
`significant contention in the Hughes IPR, when Ms. Fradenburgh refused to appear
`
`for cross-examination following institution in the Hughes IPR.
`
`Furthermore, even if considered, Ex. 1019 is woefully deficient in content.
`
`Ex. 1019 ¶ 6 refers to an “acquisition record,” said to be attached as “Exhibit A,”
`
`but no Exhibit A is attached. At ¶ 9, Ms. Fradenburgh states she has reviewed the
`
`reference in Exhibit B, yet no Exhibit B is attached. While an email is attached to
`
`the declaration, it contains only several pages with numerous unexplained dates
`
`and data entries. The declaration also does not address whether the reference was
`
`catalogued or indexed in the library system.
`
`Finally, Petitioner here misapprehends the scope of the Board’s decision in
`
`IPR2015-00059. The Board’s decision to apply Divsalar as prior art relied at least
`
`partially on a waiver theory. See, e.g., IPR2015-00059, Paper 42, at 31 (denying
`
`Caltech’s motion to strike Ms. Fradenburgh’s testimony on the basis of no timely
`
`-11-
`
`

`

`evidentiary objection). No such waiver theory applies in the present case. Ex. 1019
`
`is inadmissible hearsay here, a point Caltech will vigorously pursue.
`
`Accordingly, the Board should deny grounds 1, 3, and 4 on the additional
`
`basis that Petitioner has not established Divsalar as prior art to the ’781 patent.
`
`IV. THE PROPOSED GROUNDS OF CHALLENGE FAIL
`
`A. Ground 1 Fails
`
`Regardless of whether Divsalar is disqualified as prior art, the petition fails
`
`to demonstrate that claims 1-3, 5-8, 10, and 124 would have been obvious over the
`
`combination of Ping in view of Divsalar as asserted in Ground 1.
`
`As discussed above, Petitioner’s obviousness theory in Ground 1 is
`
`constructed based on applications of Ping’s disclosure already considered and
`
`rejected by the Board in the Hughes IPR. Recycling those charges, Petitioner
`
`alleges that Ping teaches all aspects of the challenged claims. Petitioner then adds
`
`Divsalar, and the corresponding obviousness theory, as a contingency—in the
`
`factual event the Ping does not teach its encoding in two distinct steps. See Pet at
`
`38 (“…even if Ping, alone, did not teach encoding in two distinct steps (which it
`
`does)…. Divsalar teaches encoding in two distinct steps.”).
`
`
`
`4 The Petition does not challenge claims 1-2, but purports to show how those
`
`claims are invalidated by Ping in view of Divsalar. Pet. at 35.
`
`-12-
`
`

`

`The petition misconstrues the disclosure of Ping and fails to demonstrate that
`
`a person of ordinary skill in the art would have been motivated to combine the
`
`references in the manner proposed.
`
`1. The single equation of Ping fails to teach the first and
`second encoding operations as recited in claim 1
`
`As an initial matter, Ping does not teach encoding in two distinct steps as
`
`argued in the petition. The petition fails to establish that Ping teaches or suggests
`
`“performing a first encoding operation on at least some of the information bits”
`
`and “performing a second encoding operation using the L transformed input bits,”
`
`as recited in claim 1.
`
`Petitioner points to a single equation in Ping but fails to adequately explain
`
`how Ping’s Equation (4) discloses two distinct encoding operations as claimed.
`
`Pet. at 25-29. As discussed above, this is the same disclosure that the Board
`
`previously rejected as not showing the recited two-step encoding process in the
`
`Hughes IPR. See IPR2015-00059, 18 at 15-16; see also id. Paper 4 at 19-21. While
`
`the present petitioner has drawn boxes around the terms of Ping’s Equation (4)
`
`(Pet. at 25) that were not explicitly drawn in the Hughes petition, simply drawing
`
`boxes around the same terms of Ping’s equation does not change the disclosure of
`
`Ping. Accordingly, Petitioner has repeated the same failed art and argument that
`
`was previously advanced and rejected.
`
`-13-
`
`

`

`Specifically, the petition alleges that Equation (4) is implemented by first
`
`computing all of the summations of the information bits, and then accumulating
`
`the summations in order to calculate the parity bits. Pet. at 27-29. However, this
`
`interpretation is not in accordance with Ping’s teachings. Ping represents the parity
`
`bit calculation using a single equation in
`
`which each parity bit is computed by adding
`
`the previous parity bit and the summation of
`
`information bits corresponding to that
`
`particular parity bit. See Ex. 1003, at 38
`
`(Equation (4) (annotated here)). Equation (4) is consistent with each summation of
`
`information bits being performed when computing a particular parity bit, rather
`
`than performing all of the summations prior to any parity bit calculations as
`
`proposed by Petitioner. Nothing in Ping, either in Equation (4) or elsewhere,
`
`however, provides any indication of two distinct encoding steps as argued in the
`
`petition. Any suggestion to the contrary is pure conjecture.
`
`As further evidence that Petitioner improperly interprets Ping’s single-parity
`
`bit equation as disclosing the claimed first and second encoding steps, the portion
`
`of the equation identified by Petitioner as corresponding to the claimed first
`
`encoding step produces only a single bit, characterized by Petitioner as a “sum of
`
`information bits.” Even if Petitioner is correct in its assertion that the portion of
`
`-14-
`
`

`

`the equation above boxed in red produces a sum of information bits, this does not
`
`disclose generating L transformed bits as required by claim 1, “wherein L is two or
`
`more.” Ex. 1001 cl. 1 (emphasis added). Petitioner acknowledges “the first
`
`encoding operation must generate ‘two or more’ transformed bits” (Pet. at 44) yet
`
`provides no explanation as to how production of a single summation bit in the
`
`portion of the relied upon equation satisfies this limitation.
`
`Petitioner’s argument additionally fails adequately explain its labeling as a
`
`“generator matrix” a portion of a parity check matrix (i.e., “Hd”). As discussed
`
`above, a similar argument was considered by the Board in the Hughes IPR2015-
`
`00059. Paper 18 at 15-16 (petitioner “improperly equates Ping’s LDPC codes with
`
`the ’781 use (Fig. 4) of a low density generator matrix (LDGM) in order to imply
`
`that Ping meets limitations [b] and [c] of claim 1.”).
`
`Here, Petitioner asserts that a portion of the parity check matrix (i.e., “Hd”)
`
`is a generator matrix as a basis for its assertion that Ping discloses the first and
`
`second encoding steps recited in the challenged claims. Pet. at 37-38. Ping,
`
`however, expressly describes Hd is a portion of the parity check matrix, but does
`
`not use the terminology of generator matrix. The petition materials do not
`
`adequately explain the basis for recasting the terminology.
`
`In fact, the petition’s own discussion of generator matrices seems to
`
`contradict Petitioner’s identification of Hd as a generator matrix. The petition
`
`-15-
`
`

`

`defines a generator matrix as a matrix that generates a codeword x from a set of
`
`information bits. See Pet. at 12. However, the Hd portion of the parity check matrix
`
`of Ping fails to generate a codeword x from a set of information bits. No codeword
`
`at all is generated by the Hd portion of the parity check matrix of Ping; it instead
`
`“generates” a vector of length n-k (i.e., the number of parity bits). For instance, for
`
`a code such as Ping’s with n codeword bits generated from k information bits, a
`
`generator matrix G (as defined by Petitioner) must have dimensions n×k. Pet. at
`
`12. Yet as explained above, Ping’s submatrix Hd is an (n-k)×k matrix. See also Pet.
`
`at 26 (showing Hd with n-k rows and k columns).
`
`Petitioner’s argument that Ping teaches two distinct encoding steps fails for
`
`additional reasons. For example, the petition also alleges that because “Ping would
`
`not teach storing Hd in the encoder’s memory unless it were used,” a person of
`
`ordinary skill would therefore have understood the summations to be computed
`
`prior to the accumulation step. Pet. at 37. The fact that Hd is stored in the encoder,
`
`however, has no bearing on whether the alleged summation step is performed prior
`
`to the accumulation step. The matrix merely specifies a particular format for
`
`storing data values, and does not impose any requirements on how and when those
`
`data values would be retrieved and used in an encoding operation. Moreover, as Hd
`
`is a part of the parity check matrix, it would be logical to store it for performing
`
`parity checks.
`
`-16-
`
`

`

`The petition also alleges that a single generator matrix that combined both
`
`steps of Ping’s Equation (4) into a single step “would not be low density and would
`
`require significantly more memory.” Pet. at 38. This statement is entirely
`
`conclusory and is not supported by any reasoning or citations to the references of
`
`record. Petitioner does not produce any evidence to substantiate its assertion that
`
`encoding with a single step would necessarily require a high-density generator
`
`matrix and require more memory. Ping expressly attributes the reduction in
`
`memory usage and complexity to the semi-random code design, rather than the
`
`number of steps involved in performing the encoding operation.
`
`Accordingly, Ping fails to disclose the two-step encoding as alleged.
`
`2. The petition does not demonstrate motivation to combine
`
`The petition fails to provide the requisite “articulated reasoning with some
`
`rational underpinning” to support the asserted conclusion of obviousness. KSR Int’l
`
`v. Teleflex, Inc., 550 U.S. 398, 419 (2007) (citing In re Kahn, 441 F.3d 977, 988
`
`(Fed. Cir. 2006)). Petitioner’s reasoning is superficial, overly simplistic, and lacks
`
`any meaningful explanation about how or why the combination of Ping’s and
`
`Divsalar’s teachings would be combined. The petition additionally fails to address,
`
`let alone establish, that there would be any reasonable expectation of success to
`
`form an operable and useful code. The petition proposes several “motivations to
`
`combine Ping and Divsalar” at pages 37-40, each of which can be dismissed.
`
`-17-
`
`

`

`Petitioner first advances a confusing and nonsensical argument that Ping and
`
`Divsalar would be combined because both teach two encoding steps. See Pet. at
`
`37-38 (“Ping teaches two encoding steps…. In view of Divsalar’s teaching
`
`encoding in two distinct steps, one of ordinary skill would have implemented
`
`Ping’s encoder using two steps.”)5 Petitioner’s logic is circular and self-defeating.
`
`Divsalar is relied on to teach a modification of Ping—encoding in two steps.
`
`Petitioner’s proffered rationale for said modification is that Ping already discloses
`
`encoding in two distinct steps. Id. at 37 (“Ping teaches two encoding steps…. One
`
`of ordinary skill would have understood Ping to disclose encoding in two distinct
`
`steps for several reasons.”) It is entirely unclear why there would be a motivation
`
`to modify Ping to include two distinct steps if Ping already includes two distinct
`
`steps as argued in the petition. Conversely, if Ping does not teach encoding in two
`
`distinct steps, the stated motivation of both references teaching two-step encoding
`
`is factually incorrect.
`
`
`
`5 Caltech does not acquiesce to the characterization of Divsalar in the petition
`
`materials, as numerous inaccuracies have been noted. For example, Petitioner
`
`inaccurately states that Divsalar’s interleaving step is “optional.” Pet. at 37
`
`Divsalar does not state that the interleaving step is optional.
`
`-18-
`
`

`

`In other words, Petitioner’s motivation to combine Ping with Divsalar
`
`defeats itself: if Ping does not teach two steps, then the motivation does not apply,
`
`whereas if Ping does teach two steps, then the proposed combination makes no
`
`sense.
`
`Moreover, Ping’s purported steps of summation and accumulation are
`
`different from Divsalar’s steps of repetition and accumulation. The petition does
`
`not describe in what manner one of skill would have combined the steps of
`
`summation and repetition to arrive at the claimed subject matter. Nor does it
`
`explain why a person of ordinary skill would have thought to combine these steps.
`
`Petitioner instead relies on improper hindsight bias and concludes that combining
`
`two distinct elements in a patent claim would have been obvious from independent
`
`disclosures in different references without providing any supporting analysis.
`
`Para-Ordnance Mfg. v. SGS Importers Int’l, Inc., 73 F.3d 1085, 1086 (Fed. Cir.
`
`1995) (“Obviousness may not be established using hindsight . . . .”) (citing W.L.
`
`Gore & Assocs., Inc. v. Garlock, Inc., 721 F.2d 1540, 1551, 1553 (Fed. Cir. 1983))
`
`(affirming district court’s finding of invalidity).
`
`Next, Petitioner argues that even if Ping did not teach two encoding steps,
`
`Divsalar would have motivated a skilled artisan to implement Ping’s method using
`
`two distinct steps. Pet at 38. Petitioner advances two reasons for the combination,
`
`both of which can be rejected.
`
`-19-
`
`

`

`First, the petition states that the combination would “simplify the encoder’s
`
`implementation.” Pet. at 38. Petitioner’s argument is a bare conclusion
`
`unsupported by any explanation. There is no explanation of what is meant by
`
`“implementation,” how such implementation would be simplified, or why this
`
`would have been recognized by a person of ordinary skill. Dr. Davis merely
`
`parrots the same conclusory statement presented in the petition6. Ex. 1004, ¶102.
`
`See Kinetic Techs., Case No. IPR2014-00529, Paper 8 at 15 (“Merely repeating an
`
`argument from the Petition in the declaration of a proposed expert does not give
`
`that argument enhanced probative value.”); Corning Inc. v. DSM IP Assets B.V.,
`
`Case No. IPR2013-00048, Paper 94 at 33 (P.T.A.B. May 9, 2014) (finding that an
`
`expert’s verbatim repeating of attorney argument warrants “little weight in the
`
`absence of objective evidentiary support”). Pursuant to 37 C.F.R. § 42.65(a),
`
`“[e]xpert testimony that does not disclose the underlying facts or