`
`Paper No. 21
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`
`
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`
`
`TERADATA OPERATIONS, INC.
`Petitioner,
`
`v.
`
`REALTIME DATA LLC d/b/a IXO
`Patent Owner.
`
`
`
`CASE: IPR2017-00557
`PATENT: 7,358,867 B2
`ATTORNEY DOCKET NO.: 079826.0102
`
`
`
`PETITIONER’S REPLY
`
`
`
`
`
`
`
`

`

`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`IPR2017-00557
`U.S. Patent No. 7,358,867
`
`B.
`
`2.
`
`3.
`
`2.
`
`TABLE OF CONTENTS
`
`REPLY TO PATENT OWNER’S ARGUMENTS .......................... 1
`A. Hsu teaches the “data blocks” of the challenged claims (Grounds 1, 2,
`5, and 6) ................................................................................................. 1
`1.
`The discussion of Hsu’s “blocks” in the Petition and Dr.
`Creusere’s declaration is extensive. ............................................ 2
`The Petition and Dr. Creusere’s declaration both map Hsu’s
`processing of “blocks” to the “data blocks” of Claim 16 ........... 6
`Dr. Creusere opines that the “heterogeneous files” of Hsu also
`can be viewed to correspond with the “data blocks” of Claim 16
`and the Challenged Claims ......................................................... 8
`Hsu determines “whether or not to compress each one of said plurality
`of data blocks with a particular one or more of several encoders”
`(Grounds 1, 2, 5, and 6) ....................................................................... 11
`1.
`Hsu’s determination not to compress a data block is not made
`“independently of any particular encoder” ............................... 11
`Even if PO were correct, the broadest reasonable interpretation
`of the claim language is still met .............................................. 16
`The combination of Hsu and Franaszek satisfies the limitations of
`[D2] and [D3] to “provide” and “output” a “null data compression
`type descriptor” ................................................................................... 17
`Patent Owner does not dispute that Hsu includes a plurality of lossless
`encoders (Grounds 2 and 5) ................................................................ 22
`The Petition satisfies the requirements of an obviousness analysis
`(Grounds 2, 5, 6, and 9) ....................................................................... 24
`CONCLUSION .................................................................................. 24
`
`C.
`
`D.
`
`E.
`
`i
`
`I.
`
`II.
`
`
`
`

`

`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`IPR2017-00557
`U.S. Patent No. 7,358,867
`
`PETITIONER’S EXHIBIT LIST
`
`December 5, 2017
`
`EX. NO.
`
`BRIEF DESCRIPTION
`
`1001
`
`1002
`1003
`1004
`
`1005
`
`1006
`
`1007
`
`1008
`
`1009
`
`1010
`
`1011
`
`1012
`
`1013
`
`U.S. Patent No. 7,358,867
`
`Declaration of Dr. Charles Creusere
`Curriculum Vitae of Dr. Charles Creusere
`File History for U.S. Patent No. 7,358,867
`W.H. Hsu, et al., Automatic Synthesis of Compression
`Techniques for Heterogeneous Files, Software Practice
`& Experience, Vol. 25, No. 10 pp. 1097-1116 (Oct.
`1995) (“Hsu”)
`U.S. Patent No. 5,870,036 to Franaszek et al.
`(“Franaszek”)
`U.S. Patent No. 5,097,261 to Langdon, Jr. et al
`(“Langdon, Jr.”)
`
`Reserved
`
`MCGRAW-HILL DICTIONARY OF SCIENTIFIC AND
`TECHNICAL TERMS, Fifth Ed. (1993) (excerpts)
`MICROSOFT PRESS COMPUTER DICTIONARY, Third Ed.
`(1997) (excerpts)
`Petition for Inter Partes Review of U.S. Patent No.
`9,054,728 (IPR2017-00108)
`Petition for Inter Partes Review of U.S. Patent No.
`9,054,728 (IPR2017-00179)
`Complaint, Realtime Data, L.L.C. v. Teradata
`Operations, Inc., No. 2:16-cv-02743 (C.D. Cal.)
`(“District Court Action”)
`
`ii
`
`
`
`
`
`

`

`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`IPR2017-00557
`U.S. Patent No. 7,358,867
`
`EX. NO.
`
`1014
`
`1015
`
`1016
`
`1017
`
`1018
`
`1019
`
`1020
`
`1021
`
`1022
`
`1023
`
`1024
`
`1025
`
`BRIEF DESCRIPTION
`Patent Owner’s Infringement Contentions and Asserted
`Claims, served in the District Court Action
`Trial Transcript Vol. 5 from Realtime Data, LLC d/b/a
`IXO v. T-Mobile U.S.A., Inc., No. 6:10-cv-493(E.D. Tex.
`Feb. 8, 2013)
`Order Granting in Part Plaintiff’s Motion for Judgment
`as a Matter of Law as to Invalidity, Realtime Data, LLC
`d/b/a IXO v. T-Mobile U.S.A., Inc., No. 6:10-cv-49, Dkt.
`662 (E.D. Tex. Mar. 4, 2013)
`Reserved
`Memorandum Opinion and Order, Realtime Data, LLC
`d/b/a IXO v. Packeteer, Inc., No. 6:08-cv-144, Dkt. 371
`(E.D. Tex. Jun. 22, 2009)
`Final Judgment, Realtime Data, LLC d/b/a IXO v. T-
`Mobile U.S.A., Inc., No. 6:10-cv-493, Dkt. 664 (E.D.
`Tex. Mar. 28, 2013)
`Decision on Appeal in Blue Coat Systems, Inc. v.
`Realtime Data LLC, Appeal 2012-002371,
`Reexamination Control No. 95/000,479
`Decision instituting Inter Partes Review of U.S. Patent
`No. 7,378,992 (IPR2016-00980) (instituted on Nov. 1,
`2016)
`Decision instituting Inter Partes Review of U.S. Patent
`No. 8,643,513 (IPR2016-00978) (instituted on Nov. 1,
`2016)
`Petition for Inter Partes Review of U.S. Patent No.
`7,161,506 (IPR2017-00176)
`Decision instituting Inter Partes Review of U.S. Patent
`No. 7,378,992 (IPR2016-00373) (instituted on Jun. 27,
`2016)
`Decision instituting Inter Partes Review of U.S. Patent
`
`
`
`iii
`
`

`

`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`IPR2017-00557
`U.S. Patent No. 7,358,867
`
`EX. NO.
`
`1026
`
`1027
`
`1028
`
`1029
`
`1030
`
`1031
`
`1032
`
`1033
`
`BRIEF DESCRIPTION
`No. 8,643,513 (IPR2016-00374) (instituted on Jun. 27,
`2016)
`Declaration of Mr. Scott Bennett and Attachments 1a-h
`William Underwood, Extensions of the UNIX File
`Command and Magic File for File Type Identification,
`Technical Report ITTL/CSITD 09-02, Georgia Tech
`Research Institute (Sept. 2009).
`AT&T UNIX® PC UNIX System V User’s Manual,
`Volume 1 (1986)
`File(1): FreeBSD General Commands Manual (Dec. 8,
`2000)
`Mark Nelson, The Data Compression Book, M&T
`Books, 1992 (“Nelson”)
`Annotated Petition filed in IPR2017-00557 (Paper 1)
`Annotated Declaration of Dr. Charles Creusere filed in
`IPR2017-00557 (Ex. 1002)
`Transcript of Deposition of Dr. Kenneth A. Zeger taken
`on November 1, 2017
`
`
`
`
`
`iv
`
`
`
`
`
`
`
`

`

`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`IPR2017-00557
`U.S. Patent No. 7,358,867
`
`
`TABLE OF AUTHORITIES
`
`
`
`Page(s)
`
`CASES
`CRFD Research, Inc. v. Dish Network Corp.,
`No. 16-2198, slip op ........................................................................................... 21
`
`Mars Inc. v. H.J. Heinz Co.,
`377 F.3d 1369 (Fed. Cir. 2004) .......................................................................... 23
`
`OTHER AUTHORITIES
`
`37 C.F.R. § 42.23 ....................................................................................................... 1
`
`MPEP, § 2111.03 ..................................................................................................... 23
`
`
`
`
`
`
`
`v
`
`

`

`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`IPR2017-00557
`U.S. Patent No. 7,358,867
`
`Petitioners submit this reply under 37 C.F.R. § 42.23 to Patent Owner’s
`
`Response (“Response,” Paper 19).
`
`I.
`
`REPLY TO PATENT OWNER’S ARGUMENTS
`A. Hsu teaches the “data blocks” of the challenged claims (Grounds
`1, 2, 5, and 6)
`
`Disregarding the text of the Petition, the Petitioner’s supporting evidence,
`
`and basing its arguments entirely on out-of-context selections from the transcript of
`
`Dr. Creusere’s deposition, PO argues that the Petition does not present evidence
`
`that “Hsu teaches the claimed ‘data blocks.’” Response, p.6. In particular, PO
`
`argues that Dr. Creusere “provides no opinion on what aspect of Hsu teaches
`
`Claim 16’s ‘data blocks,’ or how any aspect of Hsu does so.” Id., p.10-11. And,
`
`thus, PO reasons that “the Petition fails to state what in Hsu teaches the ‘data
`
`blocks’ of Claim 16.” Id. The opposite is true.
`
`Both the Petition and Dr. Creusere’s declaration repeatedly describe how
`
`Hsu’s “blocks” 1 correspond with the “data blocks” of Claim 16 and the other
`
`challenged claims. E.g. Petition, p.17, Ex.1002 ¶ 91. Dr. Creusere also explained
`
`in his deposition that in addition to Hsu’s blocks themselves, Hsu’s
`
`1
`Hsu uses the terms “block,” “block of data,” and “5k block,” amongst
`
`others, to refer to the “blocks” that make up a heterogeneous file. See Ex. 1005,
`
`pp.1102-03, 1108.
`
`1
`
`

`

`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`IPR2017-00557
`U.S. Patent No. 7,358,867
`
`“heterogeneous files” may also qualify as the claimed “data blocks.” Ex.2002
`
`107:18-108:5; 111:18-112:3, 112:19-24. Therefore, Dr. Creusere rejected PO’s
`
`repeated attempts during his deposition to characterize his declaration testimony as
`
`being limited to a single one of those two approaches to invalidity of the claims.
`
`1.
`
`The discussion of Hsu’s “blocks” in the Petition and Dr.
`Creusere’s declaration is extensive.
`
`As Dr. Creusere explains in his declaration, “Hsu discloses a system and
`
`method for compressing ‘heterogeneous files’—files that contain ‘multiple types of
`
`data such as text, images, binary, audio, or animation.’” Ex. 1002, ¶¶ 62 & 91
`
`(citing Ex. 1005, p.1098); see Petition, p.9. Each heterogeneous file can be treated
`
`by Hsu “as a collection of fixed size [data] blocks, each containing a potentially
`
`different type of data.” Id. (citing Ex. 1005, p.1102) (bracketing in original)2; see
`
`Petition, p.9. Hsu also teaches that the system operates on “variable-length
`
`block[s] of a file.” Id. ¶ 100; (citing Ex. 1005, p.1098); see Petition, pp.19-20.
`
`Dr. Creusere points out that “because ‘no [single] algorithm is effective in
`
`compressing all files,’ significant savings in memory and storage use can be
`
`obtained by compressing each of a file’s data blocks with an algorithm (i.e.
`
`encoder) that is optimal for the specific data type of the data in each block.” Id.
`
`¶ 94 (citing Ex. 1005, p.1097). And “Hsu explains that ‘applying the appropriate
`
`
`2 Emphasis supplied throughout, unless noted.
`
`2
`
`

`

`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`IPR2017-00557
`U.S. Patent No. 7,358,867
`
`algorithm to each block’ permits ‘better [space] savings than possible by using a
`
`single algorithm for compressing the file.’” Id. ¶ 91 (citing Ex. 1005, p.1097); see
`
`Petition, p.17. Thus, “Hsu’s process for characterizing the data blocks is
`
`performed ‘one 5K block at a time,’ and for ‘each block.’” Id. ¶ 62 (citing Ex.
`
`1005, pp.1108-09). In particular, Dr. Creusere explains that “Hsu uses statistical
`
`methods based on the analysis of the data within each block to determine a data
`
`type of the data block and ‘to determine the best algorithm to use in compressing
`
`each block of data in a file (possibly a different algorithm for each block).’” Id.
`
`¶ 63 (citing Ex. 5, p.1097).
`
`Hsu’s compression, Dr. Creusere writes, “is accomplished in two phases: the
`
`pre-compression phase and the compression phase.” Id. ¶ 64 (citing Ex. 1005,
`
`p.1102); see Petition, p.10. For the pre-compression phase “the Hsu system first
`
`determines the data type and compressibility of each data block, and then selects
`
`the optimal algorithm based on the block’s data type and compressibility.” Id.; see
`
`id. ¶ 96 (“Hsu teaches that the ‘compressibility of each block’ of data is
`
`determined in the first phase of compression”); Petition, p.10. Dr. Creusere says:
`
`Specifically, Hsu explains that “[t]he compressibility of a block of
`data and the appropriate algorithm to do so are determined by the
`type of data contained in the block and the type of redundancy (if
`any) in the data” and that these properties are “represented by four
`parameters: the block type, and the three redundancy metrics.”
`
`3
`
`

`

`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`IPR2017-00557
`U.S. Patent No. 7,358,867
`
`Id. ¶ 95 (citing Ex. 1005, p.1103); see Petition, p.18.
`
`Dr. Creusere further states that the “block type” of a data block in Hsu is
`
`“the predominant data type of data found in a single, fixed-size data block” and
`
`that Hsu “uses a procedure called ‘new-file’” to determine the data type of that
`
`data block. Id. ¶ 65 (citing Ex. 1005, p.1104). As for the “compressibility” of
`
`each data block, Dr. Creusere writes that:
`
`To evaluate the compressibility properties of each data block, Hsu
`computes and compares the results of various analyses of the contents
`of the data blocks (which Hsu calls “redundancy metrics”) to
`determine which of the following three different types of data
`redundancies are most prevalent in a given data block . . . Ex. 1005 at
`1104.
`
`Id. ¶ 67; see Petition, p.18. And Dr. Creusere explains that each of these
`
`“redundancy metrics” are calculated on a “block” basis. Id. ¶¶ 68-70.
`
`The highest redundancy metric (out of the three) for a data block indicates
`
`“the most prevalent type of redundancy in the data block.” Id. ¶ 71 & (citing Ex.
`
`1005, p.1106); see Petition, p.10. By using “the data type for the block and the
`
`largest redundancy metric,” Dr. Creusere states that Hsu’s system identifies “the
`
`appropriate compression algorithm for the data block.” Id. ¶ 72; see Petition,
`
`pp.18-19. And Dr. Creusere also describes that “[a]fter an optimal algorithm is
`
`selected for each data block of a file, the system creates a record of each data
`
`4
`
`

`

`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`IPR2017-00557
`U.S. Patent No. 7,358,867
`
`block and its optimal algorithm, which Hsu refers to as the file’s ‘compression
`
`plan.’” Id. (citing Ex. 1005, p.1109); see Petition, p.20. In the circumstance that
`
`“the current block is deemed uncompressible, a ‘skip’ instruction is recorded[]” in
`
`the compression plan. Id. ¶ 75 (citing Ex. 1005, p.1109); see Petition, pp.21-22.
`
`As to the second phase (the compression phase), Dr. Creusere explains that
`
`“Hsu provides that ‘[t]he compressor determines and then applies the selected
`
`algorithms to the blocks separately.’” Id. ¶ 95 (citing Ex. 1005, p.1098); see
`
`Petition, pp.18, 20. Dr. Creusere’s declaration delves deeper and discusses that
`
`Hsu includes “a simple optimization step which circumvents the overhead of
`
`restarting the compression after each fixed length block by merging contiguous
`
`blocks that are to be compressed using the same compression algorithms.” Id. ¶ 74
`
`(citing Ex. 1005, p.1109); see Petition pp.22-23. Dr. Creusere’s declaration details
`
`how the compression occurs to data blocks:
`
`Specifically, Hsu provides that “[t]he compressed data is written to an
`output buffer, while the compressed length (which indicates where in
`the compressed file a compressed block begins and ends) and
`compression method are recorded in a separate history for reference at
`decompression time. If negative compression or no compression is
`achieved, or if the block was already marked uncompressible, then the
`data is copied directly to the output buffer (the full block length and a
`code for 'no compression' are recorded in the compression history).
`Upon reaching the end of the blocks, the system writes out the
`
`5
`
`

`

`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`IPR2017-00557
`U.S. Patent No. 7,358,867
`
`compressed data from the output buffers and prepends the encoded
`compression history to produce the final output file.” Id. at 1109.
`
`Id. ¶ 105 (citing Ex. 1005, p.1109); see Petition 20-22. …
`
`Thus, both Dr. Creusere’s declaration and the Petition provide a detailed
`
`discussion of how the “blocks” of Hsu are processed.
`
`2.
`
`The Petition and Dr. Creusere’s declaration both map Hsu’s
`processing of “blocks” to the “data blocks” of Claim 16
`
`The Petition and Dr. Creusere’s declaration showed – via a direct mapping –
`
`how Hsu’s processing of the “blocks” anticipates the challenged claims. Attached
`
`as Exhibit 1031 is an annotated copy of the discussion of Claim 16 in the Petition
`
`and attached as Exhibit 1032 is an annotated copy of Dr. Creusere’s discussion of
`
`Claim 16 in his declaration.3 As is seen in the highlighting, the Petition and Dr.
`
`Creusere’s declaration discuss, on an element by element basis, how Hsu’s
`
`processing of “blocks” corresponds with Claim 16’s processing of “data blocks.”
`
`Ex. 1031, pp.16-22; Ex. 1032, ¶¶ 90-105.
`
`The way that the Petition and Dr. Creusere’s declaration discusses the
`
`“blocks” of Hsu is instructive as well. As indicated above, Hsu uses the terms
`
`“block,” “block of data,” and “5k block,” amongst others, to refer to the blocks that
`
`3
`The only annotation is highlighting to show the references to Hsu’s “blocks”
`
`throughout the discussion of Claim 16.
`
`6
`
`

`

`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`IPR2017-00557
`U.S. Patent No. 7,358,867
`
`make up a heterogeneous file. See Ex. 1005, pp.1102-03, 1108. But Hsu does not
`
`use the term “data block,” which appears in the challenged claims. However, to
`
`make clear that the “blocks” of Hsu are, in fact, “data blocks” as required by the
`
`claims, the Petition and Dr. Creusere’s declaration added the word “data” in
`
`brackets when describing that the files of Hsu are comprised of a collection of
`
`blocks:
`
`Hsu’s “compressor treats a file as a collection of fixed size
`[data] blocks, each containing a potentially different type of
`data and thus best compressed using different algorithms.” Ex.
`1005, p.1102.
`
`Petition, p.9 (in the “Brief Summary of Hsu”); Ex. 1002 (Creusere Declaration), ¶
`
`62 (in the “Overview of Hsu”). The Petition and Dr. Creusere’s declaration also
`
`add the word “data” in brackets when mapping Hsu to the first instance of “data
`
`blocks” in the claims. See Petition, P.17; Ex. 1002 (Creusere Declaration), ¶ 91.
`
`Accordingly, it is simply not true that Dr. Creusere’s declaration “provides
`
`no opinion on what aspect of Hsu teaches Claim 16’s ‘data blocks’” or that “the
`
`Petition fails to state what in Hsu teaches the ‘data blocks’ of Claim 16.”
`
`Response, pp.11. Rather, the Petition and Dr. Creusere’s declaration very clearly
`
`explain that the “blocks” of Hsu correspond with the “data blocks” of claim 16 and
`
`how Hsu’s blocks are processed such that Hsu anticipates (and renders obvious)
`
`the challenged claims.
`
`7
`
`

`

`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`IPR2017-00557
`U.S. Patent No. 7,358,867
`
`3.
`
`Dr. Creusere opines that the “heterogeneous files” of Hsu also
`can be viewed to correspond with the “data blocks” of Claim
`16 and the Challenged Claims
`
`In addition to the fact that Hsu’s processing of “blocks” anticipates (or
`
`renders obvious) the challenged claims, Dr. Creusere also opines that Hsu’s
`
`processing of heterogenous files (which are made up of those blocks) shows
`
`anticipation as well:
`
`Q. …Your opinion that Hsu meets this limitation is based on Hsu
`determining whether or not to compress the fixed size 5K data blocks.
`Correct?
`
`A. No, that's not correct. I believe that Hsu meets this opinion either if
`one considers the plurality of data blocks to be a plurality of files,
`where each file is representing a data block.
`
`But I also believe that Hsu meets this limitation of the claim if one
`considers the blocks instead to be the 5K blocks that we're discussing.
`
`Ex. 2002, 107:18-108:5. That is, one can consider either the granular “blocks” of
`
`Hsu or the heterogeneous files, which are collections of the granular “blocks,” to
`
`also correspond to the claimed the “data blocks” of the challenged claims. Under
`
`either view, Hsu’s compression of heterogeneous files anticipates the claims.
`
`PO’s Response attempts to draw wild conclusions from Dr. Creusere’s
`
`testimony about the fact that both the “heterogeneous files” and the “blocks”
`
`making up those “heterogeneous files” can be viewed to correspond with the “data
`
`8
`
`

`

`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`IPR2017-00557
`U.S. Patent No. 7,358,867
`
`blocks” of the challenged claims. In fact, PO’s primary support for its argument
`
`that “the Petition does not explain how Hsu teaches the ‘data blocks’ of Claim 16”
`
`(Response, p.9) comes from PO’s citation to the deposition testimony of Dr.
`
`Creusere. See Response, p.9-10. And it is in that testimony where Dr. Creusere
`
`testified that in his opinion both the blocks of Hsu correspond with the “data
`
`blocks” of Claim 16 and the heterogeneous files of Hsu can also independently
`
`correspond with the “data blocks” of Claim 16:
`
`Q. (By Mr. Rothwell) You don't have anything in your opinion
`that states what in Hsu constitutes a data block, correct?
`
`A. Again, as I say, I believe that with respect to Hsu, both files
`can be considered data blocks in terms of Claim 16A as well
`as the fixed size blocks that Hsu actually works with.
`
`Ex. 2002 at 117:6-17. But even in the portions of Dr. Creusere’s testimony that
`
`PO points to, it is clear that Dr. Creusere does not disavow his opinion that Hsu
`
`discloses “data blocks.” Rather, what Dr. Creusere said is that his declaration does
`
`not include an “explicit statement” that “a file is a data block as defined in Claim
`
`16A” or that “Hsu's 5K data blocks is a data block with respect to 16A”:
`
`A. Again, I do not state, to my recollection anywhere in my
`opinion, that either a file is a data block as defined in Claim
`16A; nor do I state in my opinion that one of Hsu's 5K data
`blocks is a data block with respect to 16A. I do not believe I
`make that -- I do not recall making that explicit statement
`anywhere in my declaration.
`
`9
`
`

`

`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`IPR2017-00557
`U.S. Patent No. 7,358,867
`
`Response, p. 10 (citing Ex. 2002, 116:6-118:6). But just because Dr. Creusere
`
`does not make an explicit statement that a “block is a data block” or that a “file is a
`
`data block” does not mean that the Petition fails to show how Hsu’s block-based
`
`processing of heterogeneous files anticipates the challenged claims.
`
`And PO does not allege any conflict in Dr. Creusere’s discussion of Hsu or
`
`any confusion arising from Dr. Creusere’s declaration. For example, PO does not
`
`allege that Dr. Creusere considers Hsu’s blocks to correspond with the “data
`
`block” for one element and yet consider a heterogeneous file as the “data block”
`
`for a different element. Nor does PO allege that it is confused as to any particular
`
`element. Indeed, PO fully understood the allegations of the Petition and Dr.
`
`Creusere’s declaration sufficient to prepare its preliminary response (Paper 10) and
`
`has now made additional arguments in its Response that make clear that PO fully
`
`understands the Petition’s argument that Hsu anticipates the challenged claims (see
`
`Response, pp.15-20). The ID itself captures this issue: “Hsu’s disclosure of
`
`heterogeneous files that are best compressed on a block-by-block basis requires
`
`‘receiving a plurality of data blocks.’” Paper 14 (“ID”), p.15 (quoting Petition,
`
`p.17).
`
`10
`
`

`

`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`IPR2017-00557
`U.S. Patent No. 7,358,867
`
`B. Hsu determines “whether or not to compress each one of said
`plurality of data blocks with a particular one or more of several
`encoders” (Grounds 1, 2, 5, and 6)
`
`Patent Owner argues that Hsu does not disclose Limitation 16[B]
`
`(“determining whether or not to compress each one of said plurality of data blocks
`
`with a particular one or more of several encoders”). Response, pp.15-20. To be
`
`sure, the factual underpinnings of PO’s argument are incorrect. But assuming
`
`arguendo that PO’s allegations were true, Hsu discloses Limitation 16[B] as
`
`construed with its broadest reasonable interpretation.
`
`1.
`
`Hsu’s determination not to compress a data block is not made
`“independently of any particular encoder”
`
`Limitation 16[B] recites “determining whether or not to compress each one
`
`of said plurality of data blocks with a particular one or more of several
`
`encoders.” PO does not argue that the entirety of Limitation 16[B] is missing
`
`from Hsu. PO does not allege that Hsu fails to disclose “whether or not to
`
`compress each one of said plurality of data blocks” generally. See Response,
`
`pp.15-22. Nor does PO allege that Hsu fails to disclose determining to compress
`
`“each one of said plurality of data blocks with a particular one or more of several
`
`encoders.” Id. Rather, PO argues only that Hsu fails to disclose the “or not”
`
`portion of Limitation 16[B]. Response, p.17 (“limitation 16[B] requires making a
`
`determination not to compress a data block ‘with a particular one or more of
`
`11
`
`

`

`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`IPR2017-00557
`U.S. Patent No. 7,358,867
`
`several encoders,’ and would not be satisfied by a general determination not to
`
`compress a data block at all that is made independently of any particular encoder.”)
`
`With respect to Hsu, PO alleges that “Hsu’s determination not to compress
`
`does not take any particular encoders into consideration at all.” Response, p.17.
`
`PO acknowledges that “Hsu calculates three redundancy metrics for each fixed-
`
`size file segment” and that the “metrics provide ‘a method of statistical analysis of
`
`the compressibility of a file (its redundancy types).’” Id., pp.17-18. PO points out
`
`that, if the redundancy metrics are below 2.5, then “the segment ‘is considered
`
`fully compressed (uncompressible) and the program continues on to the next’
`
`segment.”4 Id., p.18. “Otherwise, using the block type and largest metric, the
`
`appropriate compression algorithm (and possible heuristic) are chosen from the
`
`compression algorithm database.” Ex. 1005, p.1102. PO argues that Hsu’s
`
`decision not to compress is “made solely with respect to the calculated metrics and
`
`irrespective of the particular encoders available in Hsu’s system.” Response,
`
`p.18; see p.22 (“Hsu’s ultimate determination not to compress was a determination
`
`
`4
`Petitioner notes that PO elected to rewrite Hsu’s disclosure to remove
`
`references to “block.” The actual quote from Hsu is “the block is considered fully
`
`compressed (uncompressible) and the program continues on to the next block.”
`
`Ex. 1005, p. 1102.
`
`12
`
`

`

`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`IPR2017-00557
`U.S. Patent No. 7,358,867
`
`as to the data’s compressibility writ large, and not with respect to any particular
`
`encoder.”) But this is not true.
`
`Hsu chooses the compression algorithm to apply to a given block based on
`
`the data type and the redundancy metrics. The mapping of those data types to the
`
`redundancy metrics are shown in Table 1 of Hsu, as annotated in Dr. Creusere’s
`
`declaration:
`
`Ex. 1002, p.35 (annotated Table 1 provided in Dr. Creusere’s declaration). As Dr.
`
`Creusere explains in his declaration, after the metrics are calculated:
`
`
`
`13
`
`

`

`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`IPR2017-00557
`U.S. Patent No. 7,358,867
`
`“If at least one of the parameters is greater than 2.5, the file is
`considered compressible,” and the “maximum of the normalized
`metrics is then selected and used in conjunction with the file type to
`select the appropriate compression algorithm from the lookup table”
`taught in the Hsu reference.
`
`Ex. 1002, ¶ 71 (quoting Ex. 1005, pp.1106). Thus, for example, if a data block
`
`contained the “hexadecimal” data type and the highest redundancy metric above
`
`2.5 was for MSR (string repetition ratio), then Hsu would choose to apply a
`
`Lempel-Ziv compression algorithm to that block. On the other hand, if the highest
`
`redundancy metric above 2.5 was for MRL (average run-length), then Hsu would
`
`choose to apply a run-length encoding compression algorithm.
`
`As is also clear from the table, the calculated redundancy metrics, which are
`
`used to pick the algorithm, are associated with specific compression algorithms.
`
`Ex. 1005, Table 1. Specifically, the Alphabetic Distribution Metric (MAD) is
`
`associated with “arithmetic coding,” while the Average Run Length Metric (MRL)
`
`is associated with “run length encoding,” and the String Repetition Ratio Metric
`
`(MSR) is associated with “Lempel-Ziv” encoding. Id. The table shows this where
`
`“arithmetic coding” appears only under the Alphabetic Distribution Metric, “run-
`
`length encoding” appears only under the Average Run Length Metric, and Lempel-
`
`14
`
`

`

`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`IPR2017-00557
`U.S. Patent No. 7,358,867
`
`Ziv appears only under the String Repetition Ratio Metric. Id.5 This is because
`
`“each of the redundancy types is exploited by different compression algorithms.”
`
`Id., p.1104. As Hsu explains:
`
`Frequency of characters [(or alphabetic distribution)] is exploited by
`arithmetic or alphabetic encoding algorithms. … Contiguous strings
`[or runs], long strings of identical units occurring next to one another,
`are exploited by run length encoding algorithms. … Recurrent strings
`[or repetitive strings], which occur repeatedly in the input stream with
`any number of interleaved symbols, are exploited by textual
`substitution algorithms such as Lempel-Ziv.
`
`Id., p.1104 (cited at e.g. Petition, p.18; Ex. 1002, ¶¶ 67 & 95). Accordingly, Hsu’s
`
`statement that when the “metrics are all below a certain threshold, then the block is
`
`considered fully compressed (uncompressible)” does not mean that the block could
`
`never be compressed with any algorithm, but that the block is “uncompressible”
`
`with the algorithms available to Hsu (arithmetic coding, run-length encoding,
`
`Lempel-Ziv, and JPEG).
`
`
`5
`Petitioner acknowledges that the “high resolution color graphic” data type
`
`can use either JPEG or run-length encoding. But that does not discount the fact
`
`that the MAD, MRL, and MSR metrics are associated and correlated with specific and
`
`particular encoders.
`
`15
`
`

`

`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`IPR2017-00557
`U.S. Patent No. 7,358,867
`
`Moreover, Dr. Zeger agrees that there is a “correlation” and “relation”
`
`between the redundancy metrics and the available compression options. Ex. 1033,
`
`146:6-21. Accordingly, it is simply not true that when Hsu decides not to
`
`compress a data block it is “irrespective of the particular encoders available in
`
`Hsu’s system.” See Response, p.18. Rather, by using metrics that are tied to the
`
`available compression algorithms, Hsu’s decision whether or not to compress a
`
`data block with the available compression algorithms is based on the particular
`
`encoders available.
`
`2.
`
` Even if PO were correct, the broadest reasonable
`interpretation of the claim language is still met
`
`Even assuming arguendo that the use of Hsu’s redundancy metrics to
`
`determine not to compress a data block “does not consider the particular encoders
`
`available in Hsu’s system,” Hsu would still meet limitation 16[B]. PO does not
`
`propose that Limitation 16[B] requires construction. See Response, pp.5-6.
`
`Nevertheless, PO appears to be arguing that Limitation 16[B] has a special
`
`construction: determining whether [] to compress each one of said plurality of
`
`data blocks with a particular one or more of several encoders [and, determining
`
`whether not to compress each one of said plurality of data blocks with a particular
`
`one or more of several encoders]. But that is not what is required by the broadest
`
`reasonable interpretation of the claim language. All that is required by the element
`
`16
`
`

`

`
`
`Petitioner’s Reply to Patent Owner’s Response
`
`IPR2017-00557
`U.S. Patent No. 7,358,867
`
`is either a determination to compress or a determination not to compress. Hsu
`
`“determin[es] whether or not to compress each one of said plurality of data
`
`blocks.” And when Hsu decides to compress a block, it is with “a particular one or
`
`more of several encoders.”6 Hsu anticipates.
`
`C. The combination of Hsu and Franaszek satisfies the limitations of
`[D2] and [D3] to “provide” and “output” a “null data
`compression type descriptor” (Ground 9)
`
`Limitation 16[D] recites:
`
`[D1] if said determination is to not compress said particular one of
`said plurality of data blocks:
`
`[D2] providing a null data compression type descriptor
`representative of said determination not to compress; and
`
`[D3] outputting said null data compression type descriptor and
`
`6
`PO includes an argument that “the system of Hsu is analogous to other
`
`systems that may have a general rule to never compress blocks of certain data
`
`types or sizes, regardless of which encoders are installed, e.g., a determination not
`
`to compress any