`__________________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`___________________
`
`ERICSSON INC. AND TELEFONAKTIEBOLAGET
`LM ERICSSON,
`Petitioner
`
`v.
`
`INTELLECTUAL VENTURES II LLC,
`Patent Owner
`___________________
`
`Patent 7,269,127
`
`Title: METHODS AND APPARATUS FOR MULTI-CARRIER
`COMMUNICATIONS WITH VARIABLE CHANNEL BANDWIDTH
`_____________________
`
`IPR Case No. IPR2014-01185
`
`SUPPLEMENTAL DECLARATION OF ZYGMUNT J. HAAS, PH.D.
`UNDER 37 C.F.R. § 1.68 ON BEHALF OF PETITIONER
`
`
`
`
`
`
`
`ERIC-1036
`Ericsson v. IV, IPR2014-01185
`Page 1 of 22
`
`
`
`Supp. Haas Decl.
`
`IPR2014-01185
`
`I, Zygmunt Haas, do hereby declare:
`
`1.
`
`My background and qualifications are detailed in ¶¶ 5-13 in my
`
`previous declaration submitted as exhibit ERIC-1009 in this IPR, i.e., IPR2014-
`
`01185.
`
`2.
`
`In the preparation for this declaration, I have studied:
`
`(1) The ’127 Patent, ERIC-1001;
`
`(2) U.S. Patent No. 5,732,113 (“Schmidl”), ERIC-1002;
`
`(3) U.S. Patent No. 6,411,649 (“Arslan”), ERIC-1003;
`
`(4) U.S. Provisional Application 60/322,786, filed Sept. 17, 2001 (“Mody
`
`Provisional”), ERIC-1035;
`
`(5) The excerpts from Ex. 2001, Declaration of Dirk Hartogs, provided by
`
`counsel for Ericsson in Appendix A; and
`
`(6) The excerpts from ERIC-1034, Deposition of Dirk Hartogs, Ph.D.,
`
`provided by counsel for Ericsson cited herein.
`
`3.
`
`My understanding of the relevant legal standards remain the same as
`
`submitted in my previous declaration, ERIC-1009, ¶¶ 14-17.
`Dr. Hartogs mischaracterizes the claim term “insert pilot symbols into data
`blocks”
`4.
`
`Patent Owner’s expert, Dr. Hartogs, interprets the claim term “insert
`
`pilot symbols into data blocks” as “insert pilot symbols into at least one data
`
`1
`
`ERIC-1036
`Page 2 of 22
`
`
`
`Supp. Haas Decl.
`
`block.” That is, Dr. Hartogs asserts that the claim requires that pilot symbols be
`
`IPR2014-01185
`
`
`
`inserted into a discrete data block, not into (or among) a set of data blocks. For
`
`example, Dr. Hartogs states that “[t]he pilot symbols are frequency domain
`
`symbols inserted into a data block in the frequency domain.” Hartogs Decl., ¶ 41.
`
`But Dr. Hartogs goes further and asserts that the term “insert pilot symbols into
`
`data blocks” excludes any embodiments that would result in a separate pilot
`
`symbol in the time domain. Dr. Hartogs is incorrect on this point.
`
`5.
`
`Dr. Hartogs, repeatedly explains that the ’127 patent does not support
`
`embodiments that would result in a separate pilot symbol in the time domain. For
`
`example,
`
`Q. So is it your opinion then that pilot symbols don't show up as
`complete symbols in the time domain in the ‘127 patent?
`
`…
`
`A. I believe that’s correct.
`
`Hartogs Depo, 136:19-23. And also:
`
`Q. So is it then your opinion that, unlike training symbols, which
`show up in the time domain as time domain OFDM symbols, the pilot
`symbols would never show up in the time domain as a separate and
`distinct OFDM time domain symbol?
`A. As long as data was being provided to the system, that’s correct.
`
`Hartogs Depo, 139:21-140:4.
`
`
`
`2
`
`ERIC-1036
`Page 3 of 22
`
`
`
`Supp. Haas Decl.
`
`
`
`
`IPR2014-01185
`
`6.
`
`Dr. Hartog’s position that there would never be a separate and distinct
`
`pilot symbol in the time domain is in direct contradiction to the ’127 patent for
`
`several reasons. First and most glaringly, U.S. Provisional Application 60/322,786
`
`(“Mody Provisional”), which is incorporated by reference into the ’127 patent,
`
`explicitly shows that the inventors intended for there to be separate time-domain
`
`pilot symbols. For example, Mody Provisional explains that “[p]ilots in the form
`
`of known OFDM symbols are sent for at least Q symbol periods (QTs) in order to
`
`obtain a unique solution for the channel coefficient estimates. … The OFDM
`
`symbol period is given by Ts = NT+Tg, where 1/T is the sample rate into the
`
`OFDM modulator.” ERIC-1035, p. 2. Since the pilot symbols are “known OFDM
`
`symbols,” this implies that they are known to the receiver, do not contain any user
`
`data in the frequency domain, and contain only pilot symbols in the frequency
`
`domain. Thus, Mody Provisional discloses separate OFDM pilot symbols in the
`
`time domain.
`
`7. Mody Provisional’s description of pilot symbols as separate time-
`
`domain symbols is consistent with the ’127 patent specification. For example,
`
`referring to the time domain, the ’127 patent explains:
`
`Training symbols are typically added as prefixes to the data
`structures (e.g., at the beginning of frame structure) to enable training
`(i.e., time and frequency synchronization) between the transmitter and
`receiver of a MIMO communication system. These training symbols
`
`
`
`3
`
`ERIC-1036
`Page 4 of 22
`
`
`
`Supp. Haas Decl.
`
`
`
`
`IPR2014-01185
`
`can be referred to as preambles and are part of the preamble
`structures. Space-time signal structures are constructed using STP for
`training symbols and data symbols individually. Furthermore, pilot
`structures (or pilots) are symbols that are also constructed by STP
`and have the same structure as preambles [training symbols].
`However, instead of being placed as a prefix to the data structure, the
`pilot structures are periodically arranged within groups of data
`symbols.
`ERIC-1001, 2:10-25. Thus, pilot symbols have the same structure as training
`
`symbols (also known as preambles) in the time domain, but the pilot symbols are
`
`arranged within groups of data symbols, as opposed to being at the beginning of a
`
`transmission.
`
`8.
`
`Dr. Hartogs explains that it is conceivable that during transmission of
`
`data blocks, an entire block in the frequency domain could be filled with pilots, but
`
`he would call the resulting time-domain symbol a training symbol, which is in
`
`direct contradiction to the passage from the ’127 patent quoted above.
`
`Q. Is there a limit to how many bills [sic – pilots] could be adjacent to
`one another in a particular data block the way you've represented it
`here?
`A. Only -- I'd say the only limitations here are put up by the
`cleverness of the implementer. Obviously, if you get to the point
`where you have the entire block filled with pilots, then it really has
`just become another training symbol and you probably have enough
`information to just reinitialize your transmission.
`
`
`
`4
`
`ERIC-1036
`Page 5 of 22
`
`
`
`Supp. Haas Decl.
`
`Id., 148:10-21 (emphasis added); see also id, 166:17-169:6.
`
`
`
`
`
`IPR2014-01185
`
`9. My interpretation that “insert pilot symbols into data blocks” would
`
`lead to pilot symbols in the time domain is supported by other parts of the ’127
`
`patent. For example, the ’127 patent recites:
`
`Although omitted from FIG. 6 for simplicity, pilot symbols may also
`be intermittently inserted into the data symbols 80 by the pilot/training
`symbol inserter 46, as discussed above.
`
`ERIC-1001, 11:44-47.
`
`10. The ’127 patent states that pilot symbols are “omitted from FIG. 6 for
`
`simplicity,” but Dr. Hartogs does not know how pilot symbols would be shown in
`
`the time domain in Fig. 6 if they were shown.
`
`Q. Okay. Well, the patent states that the pilot symbols are omitted
`from figure 6 for simplicity. How would one represent pilot symbols
`in figure if one wished to do so?
`A. I don't know.
`Q. Could you represent a pilot symbol in figure 6?
`A. Nothing -- certainly, nothing comes to mind.
`Id., 130:11-19.
`
`11. Dr. Hartogs also presents Figure E in his declaration, which illustrates
`
`his opinion that a time domain data symbol includes a combination of frequency
`
`domain pilot symbols and data symbols. See Hartogs Decl., ¶ 46, Fig. E. Dr.
`
`Hartogs’ Fig. E implies that pilot symbols are not actually omitted from Fig. 6 of
`
`
`
`5
`
`ERIC-1036
`Page 6 of 22
`
`
`
`Supp. Haas Decl.
`
`the ’127 patent (even though the ’127 patent says they are omitted), but rather are
`
`IPR2014-01185
`
`
`
`embedded within data symbols 80.
`
`12. By contrast, “inserting pilot symbols into data block” in claim 1
`
`includes an embodiment in which pilot symbols are inserted in the frequency
`
`domain in such a way that pilot symbols are produced in the time domain. The
`
`figure below that illustrates my understanding.
`
`
`
`
`
`6
`
`ERIC-1036
`Page 7 of 22
`
`
`
`Supp. Haas Decl.
`
`IPR2014-01185
`
`FIG. A
`
`13. As shown above, the pilot/training symbol inserter “combines training
`
`symbols with data blocks,” per claim 1, in the frequency domain in such a way as
`
`to result in training blocks in the time domain that consist solely of training
`
`symbols in the frequency domain. A training block forms part of a training symbol
`
`in the time domain (the other part being a cyclic prefix), as shown in the annotated
`
`and modified Fig. 6 of the ’127 patent. Likewise, the pilot/training symbol inserter
`
`7
`
`ERIC-1036
`Page 8 of 22
`
`
`
`Supp. Haas Decl.
`
`“inserts pilot symbols into data blocks” in the frequency domain in such a way to
`
`IPR2014-01185
`
`
`
`result in pilot blocks in the time domain that consist solely of pilot symbols in the
`
`frequency domain. A pilot block forms part of a pilot symbol in the time domain,
`
`as shown in the annotated and modified Fig. 6 of the ’127 patent. Fig. A is
`
`consistent with how the ’127 patent describes pilot symbols. The pilot symbols
`
`and training symbols are inserted in a similar manner so as to result in separate
`
`OFDM symbols (which are in the time domain).
`
`Reason for the claimed “pilot symbols”
`
`14. The ’127 patent discusses the time variation of the channel in relation
`
`to pilot symbols and training symbols.
`
`Pilot blocks are typically transmitted with data blocks to calibrate
`(i.e., synchronize) the receiver 16 to the transmitter 14 on a small
`scale. This calibration, or synchronization, accounts for the time
`varying nature of the channel 12, for example. Training symbols,
`however, are typically used to periodically calibrate the receiver 16 to
`the transmitter 14.
`ERIC-1001, 7:40-45. Training symbols are used for initial synchronization at the
`
`beginning of a transmission. This is why, for example, multiple training symbols
`
`are shown at the beginning of a transmission in Fig. 6 of the ’127 patent, with the
`
`first one of the training symbols being an “enhanced” training symbol, whereas for
`
`periodic calibration after initial synchronization as few as one pilot symbol in the
`
`time domain (from multiple pilot symbols in the frequency domain) may be used
`
`
`
`8
`
`ERIC-1036
`Page 9 of 22
`
`
`
`Supp. Haas Decl.
`
`for periodic calibration. The “small scale” referred to in the above passage from
`
`IPR2014-01185
`
`
`
`the ’127 patent is a small-time scale. By contrast, the training symbols
`
`synchronize on a “large (time) scale” because a transmitter and receiver are
`
`typically not synchronized before a transmission starts.
`
`Schmidl in view of Arslan renders the “pilot/training symbol inserter” of
`claim 1 obvious
`
`
`a. Disclosure of Schmidl
`
`15. Schmidl discloses an encoder that inserts training symbols in the
`
`frequency domain that then become training symbols in the time domain. Schmidl
`
`discloses the structure of an OFDM symbol in Fig. 4 and a transmitted data frame
`
`in Fig. 6. The combined teachings of Schmidl’s Figs. 4 and 6 are annotated and
`
`presented below.
`
`16. Schmidl’s encoder 14 in Fig. 1 creates the training symbols: “since a
`9
`
`
`
`
`
`ERIC-1036
`Page 10 of 22
`
`
`
`Supp. Haas Decl.
`
`2m -ary QAM encoder 14 is used to create first OFDM training symbol 134...”
`
`IPR2014-01185
`
`
`
`ERIC-1002, 13:16-18. This description together with the frequency contents of the
`
`symbols in the transmitted frame in Fig. 6 inform a POSA how the encoder 14
`
`creates the training symbols. A POSA would have understood that the encoder 14
`
`necessarily incorporates circuitry that inserts training symbols in the frequency
`
`domain such that training symbols in the time domain are produced. This circuitry
`
`is an example of a training symbol inserter.
`
`b. Disclosure of Arslan
`
`17. Arslan discloses
`
`that “data sequences are commonly called
`
`synchronizing sequences or training sequences and are typically provided at the
`
`beginning of a frame of data.” ERIC-1003, 1:57-59. Arslan also discloses that “a
`
`channel tracker may be synchronized using a synchronization sequence [or training
`
`sequence] and then periodically retrained using known pilot symbols” (id., 3:4-6).
`
`Arslan further describes the structure of its frames, using Fig. 4 as an illustration.
`
`FIG. 4 illustrates a frame structure which may be utilized with the
`present invention. As seen in FIG. 4, the frame includes a
`synchronizing portion 0 to A using training symbols, information
`portions B-C, F-G and J-K and pilot portions D-E, H-I and Y-Z.
`According to the present invention, the pilot portions are interspersed
`between information portions so as to allow retraining of the
`adaptive channel estimator 30 during the frame. The synchronizing
`portion of the frame is a series of predefined symbols, from 0 to A in
`
`
`
`10
`
`ERIC-1036
`Page 11 of 22
`
`
`
`Supp. Haas Decl.
`
`
`
`
`IPR2014-01185
`
`FIG. 4, which are the same for each received frame. The information
`portion of the frame, from B to C, F to G and J to K, contains symbols
`which may vary from frame to frame and contain the information to
`be transmitted in the frame. The pilot portions of the frame, from D
`to E, H to I and Y to Z, contain predefined symbols which may be
`used to retrain the adaptive channel estimator 30.
`Id., 6:7-21 (emphasis added). Fig. 4 of Arslan is presented below for reference.
`
`
`18. The modulations described in Arslan are not OFDM, but rather single
`
`carrier modulations. However, Arslan is not intended to be limited to single-
`
`carrier modulations:
`
`Furthermore, while the present invention has been described with
`respect to frames and symbols, these terms are used in a generic sense.
`Thus, a frame generically refers to a predefined data sequence. A
`symbol generically refers to an information unit such as, for example,
`a bit.
` ERIC-1003, 10:7-11.
`
`19. Arslan also discloses that pilot symbols and training symbols are
`
`inserted in a frame of data: “known data sequences are inserted periodically into
`
`the transmitted information sequences… [and] are commonly called synchronizing
`
`
`
`11
`
`ERIC-1036
`Page 12 of 22
`
`
`
`Supp. Haas Decl.
`
`sequences or training sequences and are typically provided at the beginning of a
`
`IPR2014-01185
`
`
`
`frame of data,” (id., 1:55-59) and “[t]he present invention utilizes pilot symbols
`
`inserted in a frame of data” (id., 5:26-27).
`
`20.
`
`In summary, Arslan discloses that a frame structure beginning with
`
`training symbols and containing pilot symbols subsequently interspersed among
`
`data symbols in the time domain is useful to provide the “ability to track the
`
`channel parameters for fast time-varying systems” (id., 2:13-15).
`
`c. The combination of Schmidl and Arslan renders claim 1 obvious
`
`21. The only element of claim 1 in dispute is “the pilot/training symbol
`
`inserter … configured to insert pilot symbols into data blocks.” First, I have
`
`shown that Schmidl discloses circuitry that inserts known symbols to calibrate or
`
`synchronize the receiver to the transmitter. Specifically, training symbols are
`
`inserted in the frequency domain to form time-domain training symbols with the
`
`characteristics required by claim 1. Second, I have shown that Arslan discloses a
`
`frame structure beginning with training symbols and containing pilot symbols
`
`subsequently interspersed among time-domain data symbols that provides the
`
`ability to track the channel parameters for fast time-varying systems to calibrate or
`
`synchronize the receiver to the transmitter.
`
`22. Arslan recognizes the benefit of including pilot symbols within the
`
`data transmission to maintain calibration and synchronization. Accordingly, it
`
`
`
`12
`
`ERIC-1036
`Page 13 of 22
`
`
`
`Supp. Haas Decl.
`
`would have been obvious to use pilot symbols in Schmidl’s system as separate
`
`IPR2014-01185
`
`
`
`time-domain symbols for tracking time variations to calibrate or synchronize the
`
`receiver to the transmitter by using time-domain pilot symbols. As discussed
`
`above, Schmidl discloses inserting at the beginning of data structures known
`
`symbols as training symbols for calibration and synchronization. It would have
`
`been obvious to create time-domain pilot symbols in Schmidl in the same manner
`
`as the time-domain training symbols are created in the Schmidl’s encoder – that is,
`
`by inserting blocks of known pilot symbols in the frequency domain that would
`
`result in time-domain pilot symbols. The same circuitry in Schmidl’s encoder 14
`
`used to insert training symbols would be used to insert pilot symbols, resulting in
`
`the claimed “pilot/training symbol inserter configured to insert pilot symbols into
`
`data blocks.”
`
`
`
`
`
`
`
`13
`
`ERIC-1036
`Page 14 of 22
`
`
`
`Supp. Haas Decl.
`
`CONCLUSION
`
`IPR2014-01185
`
`23.
`
`This declaration and my opinions herein are made to the best of my
`
`knowledge and understanding, and based on the material available to me, at the
`
`time of signing this declaration. I declare that all statements made herein on my
`
`own knowledge are true and that all statements made on information and belief are
`
`believed to be true, and fiirther, that these statements were made with the
`
`knowledge that Wlllfill false statements and the like so made are punishable by fine
`
`or imprisonment, or both, under Section 1001 or Title 18 of the United States
`
`Code.
`
`Pm? J2!
`
`.2015’
`
`Date
`
`Zygmunt Haas, Ph.D.
`
`ERIC-1036
`Page 15 of 22
`
`
`
`
`
`IPR2014-01185
`
`Appendix A
`
`
`
`
`15
`
`Supp. Haas Decl.
`
`
`
`
`
`
`
`ERIC-1036
`Page 16 of 22
`
`
`
`Supp. Haas Decl.
`
`
`
`
`IPR2014-01185
`
`
`
`
`
`16
`
`ERIC-1036
`Page 17 of 22
`
`
`
`Supp. Haas Decl.
`
`
`
`
`IPR2014-01185
`
`
`
`
`
`17
`
`ERIC-1036
`Page 18 of 22
`
`
`
`Supp. Haas Decl.
`
`
`
`
`IPR2014-01185
`
`
`
`18
`
`
`
`ERIC-1036
`Page 19 of 22
`
`
`
`Supp. Haas Decl.
`
`
`
`
`IPR2014-01185
`
`
`
`19
`
`
`
`ERIC-1036
`Page 20 of 22
`
`
`
`Supp. Haas Decl.
`
`
`
`
`IPR2014-01185
`
`
`
`20
`
`
`
`ERIC-1036
`Page 21 of 22
`
`
`
`Supp. Haas Decl.
`
`
`
`
`IPR2014-01185
`
`
`
`21
`
`
`
`ERIC-1036
`Page 22 of 22