`
`_______________
`
`
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`
`
`_______________
`
`
`
`MICROSOFT CORPORATION, MICROSOFT MOBILE INC., SAMSUNG
`ELECTRONICS AMERICA, INC. AND SAMSUNG ELECTRONICS CO. LTD.,
`Petitioners
`
`v.
`
`FASTVDO LLC,
`Patent Owner
`
`_______________
`
`Case IPR2016-01179
`Patent 5,850,482
`
`_______________
`
`
`
`DECLARATION OF KENNETH A. ZEGER, PH.D., IN SUPPORT OF
`PATENT OWNER’S RESPONSE TO PETITION
`
`
`
`
`IPR2016-01179 Ex. 2004
`Page 1 of 35
`
`
`
`I.
`
`INTRODUCTION
`
`1. I, Dr. Kenneth A. Zeger, have been retained by Patent Owner FASTVDO
`
`LLC (“FASTVDO” or “Patent Owner”) through Zunda LLC to provide my
`
`opinions in support of their Response to the Petition for Inter Partes Review
`
`of U.S. Patent No. 5,850,482 to Meany et al., issued on December 15, 1998
`
`(“’482 Patent,” Ex. 1001) pursuant to the legal standards set forth below.
`
`Zunda LLC is being compensated for my time at the rate of $790 per hour
`
`for time spent on non-deposition tasks and for deposition time. I have no
`
`interest in the outcome of this proceeding, and no part of my compensation
`
`is contingent upon the outcome of this proceeding.
`
`2. I have also been asked to provide my technical review, analysis, insights, and
`
`opinions regarding the Declaration of Professor Robert L. Stevenson, Ph.D.
`
`(“Stevenson Declaration,” Ex. 1005) on the patentability of claims 1-3, 5, 6,
`
`12-14, 16, 17, and 28 of the ’482 Patent and Microsoft Corporation,
`
`Microsoft Mobile Inc, Samsung Electronics America, Inc. and Samsung
`
`Electronics Co. Ltd.’s (“Petitioners”) Petition that relies on the Stevenson
`
`Declaration. I have also reviewed the deposition transcript of Dr. Stevenson
`
`from March 9, 2017 (Ex. 2006).
`
`3. In preparing this Declaration, I have also reviewed U.S. Patent No. 5,392,037
`
`to Kato (“Kato,” Ex. 1002) and U.S. Patent No. 5,243,629 to Wei (“Wei,”
`
`IPR2016-01179 Ex. 2004
`Page 2 of 35
`
`
`
`Ex. 1004), portions of the file history of the ‘482 patent (Ex. 1003), as well
`
`as other documents referenced below in this Declaration.
`
`4. The statements made herein are based on my own knowledge and opinions.
`
`II. BACKGROUND AND QUALIFICATIONS
`
`5. I have studied, taught, and practiced electrical and computer engineering for
`
`more than thirty years.
`
`6. I received a Bachelor’s degree in Electrical Engineering and Computer
`
`Science from the Massachusetts Institute of Technology in 1984.
`
`7. I received a Master of Science degree in Electrical Engineering and
`
`Computer Science from the Massachusetts Institute of Technology in 1984.
`
`8. I received a Master of Arts degree in Mathematics from the University of
`
`California, Santa Barbara, CA in 1989.
`
`9. I received a Ph.D. degree in Electrical and Computer Engineering from the
`
`University of California, Santa Barbara, CA in 1990.
`
`10. I am currently a Full Professor of Electrical and Computer Engineering at
`
`the University of California, San Diego (UCSD). I have held this position
`
`since 1998, having been promoted from Associated Professor after two years
`
`at UCSD. I have been an active member of the UCSD Center for Wireless
`
`Communications for 18 years. I teach courses full-time at UCSD in the
`
`fields of Electrical and Computer Engineering, and specifically in subfields
`
`IPR2016-01179 Ex. 2004
`Page 3 of 35
`
`
`
`including communications and information theory at the undergraduate and
`
`graduate levels. Prior to my employment at UCSD, I taught and conducted
`
`research as a faculty member at the University of Illinois, Urbana-
`
`Champaign for four years, and at the University of Hawaii for two years.
`
`11. My twenty-plus years of industry experience includes consulting work for
`
`the United States Department of Defense as well as for private companies
`
`such as Xerox, Nokia, MITRE, ADP, and Hewlett-Packard. The topics upon
`
`which I provide consulting expertise include data communications for
`
`wireless networks, digital communications, information theory, computer
`
`software, and mathematical analyses.
`
`12. I have authored approximately 73 peer-reviewed journal articles, the
`
`majority of which are on the topic of communications, information theory,
`
`or signal processing. I have also authored over 100 papers at various
`
`conferences and symposia over the past twenty-plus years, such as the: IEEE
`
`International Conference on Communications; IEEE Radio and Wireless
`
`Symposium; Wireless Communications and Networking Conference; IEEE
`
`Global Telecommunications Conference; International Symposium on
`
`Network Coding; IEEE International Symposium on Information Theory;
`
`UCSD Conference on Wireless Communications; International Symposium
`
`on Information Theory and Its Applications; Conference on Advances in
`
`IPR2016-01179 Ex. 2004
`Page 4 of 35
`
`
`
`Communications and Control Systems; IEEE Communication Theory
`
`Workshop; Conference on Information Sciences and Systems; Allerton
`
`Conference on Communications, Control, and Computing; Information
`
`Theory and Its Applications Workshop; Asilomar Conference on Signals,
`
`Systems, and Computers. Roughly half of those papers relate to data
`
`compression. I also am co-inventor on a US patent disclosing a memory
`
`saving technique for image compression.
`
`13. I was elected a Fellow of the IEEE in 2000, an honor bestowed upon only a
`
`small percentage of IEEE members. I was awarded the National Science
`
`Foundation Presidential Young Investigator Award in 1991, which included
`
`$500,000 in research funding. I received this award one year after receiving
`
`my Ph.D.
`
`14. I have served as an Associate Editor for the IEEE Transactions on
`
`Information Theory and have been an elected member of the IEEE
`
`Information Theory Board of Governors for three, three-year terms. I
`
`organized and have been on the technical advisory committees of numerous
`
`workshops and symposia in the areas of communications and information
`
`theory. I regularly review submitted journal manuscripts, government
`
`funding requests, conference proposals, student theses, and textbook
`
`IPR2016-01179 Ex. 2004
`Page 5 of 35
`
`
`
`proposals. I also have given many lectures at conferences, universities, and
`
`companies on topics in communications and information theory.
`
`15. I have extensive experience in electronics hardware and computer software,
`
`from academic studies, work experience, and supervising students. I
`
`personally program computers on an almost daily basis and have fluency in
`
`many different computer languages.
`
`16. My curriculum vitae, attached to this declaration as Exhibit 2005 (“Zeger
`
`CV”), lists my publication record in archival journals, international
`
`conferences, and workshops.
`
`III. LEGAL PRINCIPLES
`
`17. I am not an attorney and therefore I offer no opinions on the law. I have
`
`been advised of the following general principles of patent law to be
`
`considered in formulating my opinions as to the issues of the validity of the
`
`challenged claims.
`
`18. Anticipation: I have been informed by counsel that for a claim to be invalid
`
`as anticipated, the challenger in an inter partes review must show, by a
`
`preponderance of the evidence, that all the elements of a claim are present in
`
`a single previous device or method, or sufficiently described in a single
`
`previous printed publication or patent. To anticipate the claim, the prior art
`
`does not have to use the same words as the claim, but all the requirements of
`
`IPR2016-01179 Ex. 2004
`Page 6 of 35
`
`
`
`the claim must have been disclosed expressly or inherently, so that looking
`
`at that one reference, a person of ordinary skill in the art could make and use
`
`the claimed invention. A requirement of a claim that is missing from a prior
`
`art may be disclosed inherently if that missing requirement is necessarily
`
`present in that prior art. All elements of the claim must also be disclosed in
`
`the reference as they are arranged in the claim.
`
`19. I have been informed by counsel that if a prior art reference discloses two or
`
`more embodiments, none of which disclose all elements of a claim as
`
`arranged in the claim, those embodiments cannot be combined for
`
`anticipation purposes. Instead, the party asserting invalidity must present an
`
`obviousness analysis to support the combination of the embodiments.
`
`20. Obviousness: I have been informed by counsel that for a claim to be invalid
`
`as obvious, it must be obvious to a person of ordinary skill in the field of
`
`technology of the patent at the relevant time. The existence of every element
`
`of the claimed invention in multiple prior art references or systems does not
`
`necessarily prove obviousness. Most, if not all, inventions rely on building
`
`blocks of prior art. Obviousness may be found in an inter partes review
`
`proceeding only where there is a preponderance of evidence that the
`
`differences between the subject matter sought to be patented and the prior art
`
`are such that the subject matter as a whole would have been obvious at the
`
`IPR2016-01179 Ex. 2004
`Page 7 of 35
`
`
`
`time the invention was made to a person having ordinary skill in the art to
`
`which said subject matter pertains. Obviousness analysis involves
`
`determining the scope and content of the prior art; ascertaining the
`
`differences between the prior art and the claims at issue; and analyzing
`
`secondary considerations of non-obviousness (as explained below).
`
`21. I have been informed by counsel that in determining whether any of the
`
`challenged claims are obvious, I should consider whether the prior art
`
`discloses or suggests all the elements of the challenged claims. I understand
`
`that I should also consider whether there was a reason that would have
`
`prompted a person having ordinary skill in the art to combine the known
`
`elements (whether those elements are disclosed in different prior art
`
`references or in different embodiments in a single reference) in a way the
`
`claimed invention does, taking into account such factors as (1) whether the
`
`claimed invention was merely the predictable result of using prior art
`
`elements according to their known function(s); (2) whether the claimed
`
`invention provides an obvious solution to a known problem in the relevant
`
`field; (3) whether the prior art teaches or suggests the desirability of
`
`combining elements claimed in the invention; (4) whether the prior art
`
`teaches away from combining elements in the claimed invention; (5)
`
`whether it would have been obvious to try the combinations of elements,
`
`IPR2016-01179 Ex. 2004
`Page 8 of 35
`
`
`
`such as when there is a design need or market pressure to solve a problem
`
`and there are a finite number of identified, predictable solutions; and (6)
`
`whether the change resulted more from design incentives or other market
`
`forces. To render a claim obvious, the prior art must have provided a
`
`reasonable expectation of success.
`
`22. I have been informed by counsel that I should also consider any objective
`
`evidence (sometimes called “secondary considerations”) that may have
`
`existed at the time of the invention and afterwards that may shed light on the
`
`obviousness of the claims, such as:
`
`a. Whether the invention was commercially successful as a result of the
`
`merits of the claimed invention (rather than the result of design needs
`
`or market-pressure advertising or similar activities);
`
`b. Whether the invention satisfied a long-felt need;
`
`c. Whether others had tried and failed to make the invention;
`
`d. Whether others invented the invention at roughly the same time;
`
`e. Whether others copied the invention;
`
`f. Whether there were changes or related technologies or market needs
`
`contemporaneous with the invention;
`
`g. Whether the invention achieved unexpected results;
`
`h. Whether others in the field praised the invention;
`
`IPR2016-01179 Ex. 2004
`Page 9 of 35
`
`
`
`i. Whether persons having ordinary skill in the art of the invention
`
`expressed surprise or disbelief regarding the invention;
`
`j. Whether others sought or obtained rights to the patent from the patent
`
`holder; and
`
`k. Whether the inventor proceeded contrary to accepted wisdom in the
`
`field.
`
`l. I have been informed by counsel that differences between the prior art
`
`reference and a claimed invention, however slight, invoke the
`
`question of obviousness, not anticipation. Thus, the question of
`
`obviousness is invoked if a prior art reference, or a single embodiment
`
`of a prior art reference, discloses part but not all of the claimed
`
`invention, or if the prior art reference includes multiple, distinct
`
`teachings or embodiments that separately fail to disclose the claimed
`
`invention as arranged in the claims.
`
`IV. LEVEL OF ORDINARY SKILL
`
`23. Dr. Stevenson, Petitioner’s declarant, submitted a declaration stating that
`
`the person of ordinary skill in the art (“POSITA”) of the patented technology
`
`at the time of the invention of the ‘482 Patent, or April 17, 1996, would have
`
`a bachelor’s degree in electrical engineering, computer engineering, or the
`
`IPR2016-01179 Ex. 2004
`Page 10 of 35
`
`
`
`equivalent and 2-3 years of work experience with compression, encoding,
`
`and decoding.
`
`24. The Stevenson Declaration specifically states: “an undergraduate degree in
`
`computer science, computer engineering, or electrical engineering, as well as
`
`two years’ experience or a graduate degree with focus in the area of data
`
`compression, encoding, and decoding.” Ex. 1005, ¶40.
`
`25. I disagree with Dr. Stevenson’s definition of a POSITA in this case. I note
`
`that under Dr. Stevenson’s definition, a person of ordinary skill in the art
`
`would have experience with data compression and encoding, but would not
`
`necessarily have any experience or training in the subjects of error
`
`protection, including unequal error protection, error-resilient data
`
`transmission, or the equivalent.
`
`26. Based on my education and experience, as of April 1996, it was not
`
`standard practice in engineering to use unequal error protection (“UEP”)
`
`even when systems transmitted data of varying levels of importance. A
`
`POSITA as of the time of the ‘482 patent would have understood the costs
`
`and complexities that would have been incurred as a result of incorporating
`
`UEP in a system. The actual design of a UEP application would have
`
`generally required additional training and/or education, and would have been
`
`performed by advanced inventors or Ph.D.-level researchers.
`
`IPR2016-01179 Ex. 2004
`Page 11 of 35
`
`
`
`27. My opinion is that a POSITA would have an undergraduate degree in
`
`computer science, computer engineering, or electrical engineering, or the
`
`equivalent, as well as at least two to three years of work experience with
`
`error protection and error-resilient data transmission, or the equivalent.
`
`Equivalent knowledge and/or experience could have been acquired through
`
`other means. For example, a POSITA could have an undergraduate degree
`
`in a different field, but could have acquired the requisite knowledge in the
`
`subject of error protection and error-resilient data transmission through work
`
`experience.
`
`V.
`
`THE ‘482 PATENT
`
`28. A digital transmission system typically provides the function of
`
`communicating information (e.g., source bits) across a distance at a desired
`
`rate through a given transmission channel. A transmission channel can be
`
`noisy (i.e., subject to transmission errors), and therefore would have a
`
`limited capacity, known as the channel capacity. In the field of information
`
`theory, if the source entropy is below the channel capacity, all channel errors
`
`can in principle be overcome, and the source bits can be received entirely
`
`intact. To achieve this, channel encoding is applied to the source bits. This
`
`means that the source bits are encoded into another corresponding set of bits
`
`by an operation known as error control coding (e.g., block coding or
`
`IPR2016-01179 Ex. 2004
`Page 12 of 35
`
`
`
`convolutional coding). This operation increases the overall number of bits
`
`being transmitted by adding redundancy to the information. These channel
`
`encoded bits are transmitted and subject to the noisy transmission channel
`
`errors discussed previously. But despite the introduction of such errors in the
`
`bits received from the transmission channel, the redundancy introduced at
`
`the channel encoding stage can be utilized to accurately reproduce the source
`
`bits despite the channel noise and signal degradation. In general, more
`
`redundancy introduced in the channel encoding process confers greater
`
`robustness and resistance to channel errors, but also comes at a cost.
`
`29. The approach described above provides for equal error protection to the
`
`source bits. However, not all transmission errors due to transmission channel
`
`noise have the same effect on the receipt and correct decoding of the source
`
`bits. Some errors may be unnoticeable or inconsequential, while other errors
`
`may propagate and corrupt many source bits, or may result in the loss of
`
`synchronization. One of the benefits of the ’482 invention is the provision of
`
`greater error protection to bits more susceptible to consequential or
`
`catastrophic errors, i.e. those bits including some information that represents
`
`a property of the less or unprotected bits.
`
`30. According to the prior art, encoded code words could generally be highly
`
`susceptible to channel errors. If certain portions of the code word were
`
`IPR2016-01179 Ex. 2004
`Page 13 of 35
`
`
`
`corrupted, serious or catastrophic corruption could result. Error protection
`
`can include the transmission of additional redundant bits enabling the
`
`recipient to identify whether and how transmitted data differs from the
`
`source bits. For example, additional redundant data may be transmitted in a
`
`way that permits the original data to be reconstructed even if multiple errors
`
`are introduced. Another approach for alerting the system to errors is to
`
`append to the data cyclic redundancy check (“CRC”) bits, which represent
`
`the result of a particular mathematical operation performed upon the data.
`
`The recipient may perform the same mathematical operation upon the
`
`incoming data, and if the result is different from the appended CRC bits, the
`
`recipient can become aware of one or more errors in the data.
`
`31. Adding error protection information increases bandwidth requirements in
`
`the case of transmission over a noisy channel or storage capacity
`
`requirements in the case of storage in a storage medium. Therefore, the ’482
`
`invention contemplates the selective application of unequal error protection.
`
`It does this by distinguishing between those portions of the code word or
`
`data that are susceptible to errors and those portions that are resilient to
`
`errors, and applying more error protection to the portions susceptible to
`
`errors while applying less or no error protection to the portions resilient to
`
`errors. For example, one portion (e.g., “first portion” or “prefix”) may
`
`IPR2016-01179 Ex. 2004
`Page 14 of 35
`
`
`
`contain information about how to decode other portions (e.g., “second
`
`portion” or “suffix”). In this case, it can be beneficial to apply greater error
`
`protection to the first portion since an error in that portion would
`
`compromise the receiver’s ability to accurately decode the contents of the
`
`second portion. Thus, the invention prioritizes, and applies more protection
`
`to, those portions of the data that are more sensitive to errors (i.e., those
`
`areas that represent some property of the less protected, or unprotected, bits),
`
`as shown in Fig. 1 of the ’482 Patent reproduced below.
`
`
`
`32. The ’482 Patent discloses and claims devices and methods for organizing
`
`and encoding the data or code word and selectively applying error
`
`protection. This is beneficial in several ways, including the ability to
`
`minimize necessary transmission bandwidth or storage capacity.
`
`IPR2016-01179 Ex. 2004
`Page 15 of 35
`
`
`
`Furthermore, by structuring the data or code words into vulnerable portions
`
`and resilient portions, and then unequally protecting only those vulnerable
`
`portions against errors, the ’482 Patent reduces the need for more
`
`comprehensive error protection of all the data regardless of importance or
`
`error resilience of the data. These portions receiving higher levels of error
`
`protection contain information useful to decoding and/or decompressing the
`
`information contained in the error-resilient portions.
`
`VI. CLAIM CONSTRUCTION
`
`33. I have been informed that the ‘482 patent is expired as of April 17, 2016. I
`
`have also been informed that the claims of an expired patent are construed in
`
`inter partes review according to a district court’s standard. I understand that
`
`standard states that words of a claim are generally given their ordinary and
`
`customary meaning as understood by a person of ordinary skill in the art. I
`
`have reviewed the District Court’s claim construction order (Ex. 1014) and
`
`the Patent Trial and Appeal Board’s Institution Decision (Paper 15) and have
`
`applied the constructions set forth in those documents for the purpose of
`
`rendering my opinions below. I note that the District Court’s construction of
`
`“code word” differs slightly, and in my view, insignificantly, from the Patent
`
`Trial and Appeal Board’s construction of “code word.” I take no position on
`
`whether the constructions issued by the District Court and Patent Trial and
`
`IPR2016-01179 Ex. 2004
`Page 16 of 35
`
`
`
`Appeal Board represent the ordinary and customary meaning as understood
`
`by a POSITA as of the filing date of the ’482 patent.
`
`34. I note that neither the Petitioners nor the District Court construed the term
`
`“storage medium” as it appears in claims 5, 16, and 28. I have reviewed the
`
`‘482 patent and believe that based on the contents of the patent, the term
`
`“storage medium” should be interpreted as a physical storage device. This
`
`term should not be construed to cover transmission channels or data links,
`
`which generally do not store data.
`
`35. The specification and drawings of the ‘482 patent support the proper
`
`interpretation of a “storage medium” as a physical storage device. In FIG. 1
`
`of the ‘482 patent, the storage medium 18 is arranged parallel to a
`
`transmitter block 20 for transmitting via data links 22 and 24. Ex. 1001,
`
`FIG. 1. The ‘482 specification explains that the storage medium 18 may be
`
`a “magnetic disk storage which is error protected as shown in FIG. 6.” Ex.
`
`1001, 17:15-19. Instead of storing the data in a storage medium 18, the ’482
`
`patent discloses that “the compressed and encoded data can be efficiently
`
`transmitted, such as via first and second data links.” Ex. 1001, 17:26-27. I
`
`read this disclosure in the ‘482 patent to indicate that storing data in a
`
`storage medium should be interpreted differently from transmitting data via
`
`a data link. While I do not take the position that the preamble of claim 28 is
`
`IPR2016-01179 Ex. 2004
`Page 17 of 35
`
`
`
`a limitation, I have used the preamble of claim 28 to inform my construction
`
`of “storage medium.” Specifically, the preamble of claim 28 states a
`
`“computer readable memory.” ‘482 patent, 22:23. A POSITA would not
`
`characterize the transmission of data as a “computer readable memory,” and
`
`would not characterize the transmission of data as a “storage medium.” The
`
`‘482 patent claims also clearly distinguish between the concept of “storing”
`
`data “in a first data block of a storage medium” and the concept of
`
`“transmitting” data “via a first data link,” such as shown in the different
`
`word choices used in claims 5 and 6. Ex. 1001, 18:57-59; 19:1-2.
`
`VII. NON-OBVIOUSNESS OF THE CHALLENGED CLAIMS
`
`36. The Stevenson Declaration includes unsupported assertions as to the
`
`combinability of Kato’s embodiments. Ex. 1005, ¶¶ 80-88. But Dr.
`
`Stevenson does not provide rationale to support the specific combinations of
`
`embodiments except to state that the different embodiments allegedly
`
`disclose specific claim elements (i.e., unequal error protection. Ex. 1005, ¶¶
`
`83-84, 87-88). As I will explain below, I disagree with Dr. Stevenson’s
`
`conclusion that the challenged claims are obvious over Kato alone, and Kato
`
`in view of Wei (for claims 6 and 17).
`
`Kato
`
`IPR2016-01179 Ex. 2004
`Page 18 of 35
`
`
`
`37. The Stevenson Declaration argues that there was motivation for a POSITA
`
`to combine completely distinct embodiments in Kato. Ex. 1005, ¶¶80-88. I
`
`disagree. For example, the Stevenson Declaration purports to modify the
`
`first embodiment of Kato with the fourth embodiment of Kato in order to
`
`provide unequal protection against error propagation. But according to
`
`Kato, the first embodiment already contains this benefit. I view this
`
`oversight by the Stevenson Declaration as an indication that Dr. Stevenson
`
`was relying upon impermissible hindsight to support the combination of
`
`embodiments. I will explain my opinions in greater detail below.
`
`38. For at least the reasons discussed in this Declaration, it is my opinion that
`
`the Stevenson Declaration does not establish that Kato renders obvious the
`
`challenged claims.
`
`39. Kato separately discloses four separate embodiments. Three of the
`
`embodiments relied upon by Dr. Stevenson are introduced in Kato as a
`
`“Description of the First Preferred Embodiment” (6:55-16:28), a
`
`“Description of the Third Preferred Embodiment” (19:36-23:43), and a
`
`“Description of the Fourth Preferred Embodiment” (23:44-33:12). Likewise,
`
`Figures 1(a) and 1(b) of Kato relate to the first embodiment; Figures 4(a),
`
`4(b), and 5 relate to the third embodiment; and Figures 6(a), 6(b), and 7
`
`relate to the fourth embodiment.
`
`IPR2016-01179 Ex. 2004
`Page 19 of 35
`
`
`
`40. The fourth embodiment describes an “encoding circuit 602” that “encodes
`
`the input data Di into variable-length code words, which are sequentially
`
`arranged in one or more data store regions within a transmission format.
`
`Thus, the variable-length code words are converted into bitserial-format
`
`data, which are outputted from the encoding circuit 602.” Kato, 24:24-30.
`
`Then, after briefly describing the error correcting code (ECC) encoder 603’s
`
`operation, which includes the addition of error correcting code “to the output
`
`data from the encoding circuit 602” (Kato, 24:32-3), Kato resumes the
`
`discussion of the encoding circuit 602: “The operation of the encoding
`
`circuit 602 will now be described in more detail. As described previously,
`
`the encoding circuit 602 encodes the input data Di into the variable-length
`
`code word Ci, and locates the variable-length code word Ci in the data store
`
`region within the data transmission format.” Kato, 24:40-45. Kato then
`
`describes the arrangement of the data store region(s) with respect to the
`
`encoding circuit 602. Kato, 24:46-27:31.
`
`41. Specifically, Kato describes the RAM 617 as being “formed with one or
`
`more data store regions corresponding to the data store region or regions in
`
`the transmission format.” Kato, 25:31-34. The writing of the first portion Pi
`
`and second portion Ri into the RAM 617 is described at 25:45-26:4, and the
`
`conclusion of this process is described by Kato as “the arrangement of one
`
`IPR2016-01179 Ex. 2004
`Page 20 of 35
`
`
`
`variable-length code word Ci into the data store region.” Kato, 26:5-7. The
`
`arrangement of code words stored in the data store region(s) of RAM 617 by
`
`encoding circuit 602 is described by Kato as a “transmission format.” Kato,
`
`25:33-34; 23:54-57; 24:41-45; 24:46-48.
`
`42. To the extent Dr. Stevenson’s opinion is that Kato’s first and fourth
`
`embodiments should be combined, I disagree. Nothing in Kato suggests that
`
`the method of the first embodiment should be combined with the apparatus
`
`of the fourth embodiment. In fact, Kato specifically teaches that the data
`
`output from the first embodiment are code words that are inherently resilient
`
`to error propagation, while Kato’s fourth embodiment provides an alternate
`
`solution for preventing error propagation. As a result, Kato’s first
`
`embodiment has no need for unequal error protection because Kato’s first
`
`embodiment already “prevents error propagation, and … is advantageous
`
`from the standpoint of error-propagation suppressing characteristics.” Kato,
`
`16:4-7. Kato explains that the first embodiment decodes and recovers the
`
`correct original data Di “even when there is a certain error between the
`
`estimate Pi used in the equations (5)-(7) and the estimate used in the
`
`encoding.” Kato, 15:66-16:1.
`
`43. Thus, Kato discloses that the first embodiment “prevents error propagation,”
`
`but Dr. Stevenson is attempting to modify Kato’s first embodiment to add
`
`IPR2016-01179 Ex. 2004
`Page 21 of 35
`
`
`
`unequal error protection, which is also intended to avoid error propagation.
`
`Dr. Stevenson does not offer any engineering analysis to justify the increase
`
`costs in bandwidth or power that come with transmitting additional bits due
`
`to error protection (and unequal error protection). Kato, 16:4-5. Dr.
`
`Stevenson also does not offer any analysis explaining the benefit that this
`
`modification would achieve, if any, over the natural error protection already
`
`taught in Kato’s first embodiment. He also does not explain what results he
`
`believes would result from this modification.
`
`44. I believe Dr. Stevenson fundamentally misinterprets Kato’s first
`
`embodiment. I understand he gave the following testimony at deposition:
`
`“The way I view Kato is – something called the first embodiment is this
`
`predictive coding technique they have. And then the other three things they
`
`add are enhancements to that addition of that first embodiment, the
`
`predictive coding section.” Ex. 2006, 50:19-51:2. But Dr. Stevenson’s view
`
`of the first embodiment fails to take into account Kato’s full disclosure. In
`
`particular, Kato believes that the predictive coding technique of the first
`
`embodiment includes characteristics of error protection to avoid error
`
`propagation, and I agree with Kato’s characterization of the first
`
`embodiment in this regard. Kato, 15:66-16:7. Dr. Stevenson’s analysis
`
`assumes that the incorporation of the ECC encoder 603 from Kato’s fourth
`
`IPR2016-01179 Ex. 2004
`Page 22 of 35
`
`
`
`embodiment will improve the risk of error propagation in Kato’s first
`
`embodiment. This analysis ignores what Kato explains about the error
`
`resiliency of the first embodiment. A POSITA would not overlook these
`
`disclosures, and would not incur the cost of adding unequal error protection
`
`to Kato’s first embodiment to avoid error propagation where Kato’s first
`
`embodiment already possesses this ability.
`
`45. In conclusion, a person of ordinary skill would not be motivated to combine
`
`Kato’s first embodiment with the fourth embodiment, or vice versa. Since
`
`the code words in the first embodiment are inherently resilient to error
`
`propagation, there is no reason that a POSITA would have combined the
`
`first embodiment with the fourth embodiment, which is also directed to
`
`preventing error propagation. There would be no benefit for this
`
`modification.
`
`46. Additionally, Dr. Stevenson has not explained or remedied the
`
`incompatibilities that would result from his proposed combination of
`
`embodiments. For example, Kato’s first embodiment sub-encodes category
`
`index Ji and remainder data Ei using sub encoding circuits 109 and 110, and
`
`then combines the coded data into the multiplexed coded data Ci using
`
`multiplexing circuit 111. Kato, Fig. 1(a); 10:44-46; 11:1-2; 11:5-9. From
`
`the multiplexing circuit 111, the output data is transmitted via output
`
`IPR2016-01179 Ex. 2004
`Page 23 of 35
`
`
`
`terminal 108. Kato, 11:9-11. Unlike the first embodiment, the data stored in
`
`RAM 617 and then passed to ECC encoder 603 in the fourth embodiment
`
`does not require multiplexing. This is because the fourth embodiment stores
`
`the first and second portions in the RAM 617 in a “transmission format.”
`
`Kato, 25:31-34. The output from the ECC encoder 603 is sent to modulator
`
`604, and then to an output terminal 605. Kato, 23:57-63. Dr. Stevenson has
`
`not explained how the ECC encoder 603 would be incorporated into the first
`
`embodiment, where it would be located, and how it would interact with the
`
`sub encoding circuits 109 and 110, multiplexing circuit 111, and output
`
`terminal 108.
`
`47. Further, I understand from Dr. Stevenson’s deposition testimony that he has
`
`suggested the disclosure in Kato of