UNITED STATES PATENT AND TRADEMARK OFFICE
`_____________________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`_____________________
`
`
`
`APPLE INC.
`Petitioner
`
`v.
`
`DSS TECHNOLOGY MANAGEMENT, INC.
`Patent Owner
`
`_____________________
`
`Case IPR2015-00369
`Patent 6,128,290
`_____________________
`
`PETITIONER’S REPLY TO PATENT OWNER RESPONSE
`
`
`
`Mail Stop “PATENT BOARD”
`Patent Trial and Appeal Board
`U.S. Patent & Trademark Office
`P.O. Box 1450
`Alexandria, VA 22313-1450
`
`
`
`
`
`
`
`
`

`
`IPR2015-00369
`U.S. Pat. No. 6,128,290
`
`EXHIBIT LIST
`
`Description
`
`Apple
`Exhibit No.
`APL 1001
`APL 1002
`
`APL 1003
`APL 1004
`APL 1005
`
`APL 1006
`
`APL 1007
`
`APL 1008
`
`APL 1009
`APL 1010
`APL 1011
`
`U.S. Patent No. 6,128,290 to Carvey (“the ʼ290 patent”)
`T. J. Barber, Jr., “BodyLAN™: A Low Power Communications
`System,” Masterʼs Thesis at Massachusetts Institute of Technol-
`ogy, 1996 (“Barber”)
`U.S. Patent No. 5,241,542 to Natarajan (“Natarajan”)
`U.S. Patent No. 4,887,266 to Neve (“Neve”)
`Prosecution History of U.S. Application No. 08/949,999 (now
`U.S. Patent No. 6,128,290) (“the ’999 application”)
`U.S. Application No. 08/611,695 (as-filed) (“the ’695 applica-
`tion”)
`Apple’s Claim Construction Brief in Case No. 6:13-cv-00919
`JDL (EDTX)
`Declaration of Jack D. Grimes, Ph.D. in Support of Petition for
`Inter Partes Review of U.S. Patent No. 6,128,290 (“Grimes
`Dec.”)
`Curriculum Vitae of Jack D. Grimes, Ph.D. (“Grimes CV”)
`INTENTIONALLY BLANK
`Deposition Transcript of Robert Dezmelyk, IPR2015-00369 and
`IPR2015-00373, December 15, 2015 (“Dezmelyk Depo.”)
`APL 1012 Mischa Schwartz, Telecommunications Networks: Protocols,
`Modeling and Analysis, Addison-Wesley, 1988 (“Schwartz”)
`Tom Sheldon, Encyclopedia of Networking & Telecommunica-
`tions, Lisa Wolters-Broder ed., McGraw Hill, 2001 (other ex-
`cerpts submitted as DSS 2010)
`Declaration of Dr. Jing Hu (“Hu Dec.”)
`Curriculum Vitae of Dr. Jing Hu
`
`APL 1013
`
`APL 1014
`APL 1015
`
`
`
`
`
`
`
`i
`
`

`
`IPR2015-00369
`U.S. Pat. No. 6,128,290
`
`C.
`
`2.
`
`
`TABLE OF CONTENTS
`Introduction ..................................................................................................... 1
`I.
`Natarajan teaches or suggests a server transmitter operating in “low duty
`II.
`cycle RF bursts,” as recited in claim 1 of the ’290 patent. ....................................... 2
`III. HDLC is consistent with low duty cycle RF bursts. ...................................... 3
`A.
`The preferred embodiment in the ’290 patent uses HDLC. .................. 4
`B.
`DSS relies on the testimony of Mr. Dezmelyk, who admits
`he is not an expert in HDLC. ................................................................. 5
`A POSA would have looked to Schwartz for information on
`Natarajan’s HDLC protocol and understood that it is
`consistent with low duty cycle RF bursts. ............................................. 6
` Mr. Dezmelyk did not consider Schwartz–the most 1.
`
`logical reference for information on Natarajan’s
`HDLC protocol–when forming his opinions. ............................ 6
`Natarajan’s HDLC protocol is consistent with low
`duty cycle RF burst communication. ......................................... 7
`D. DSS and Mr. Dezmelyk concoct inaccurate piecemeal
`arguments from excerpts of unrelated references that are
`inconsistent with each other and inconsistent with the
`operation of HDLC. ............................................................................... 9
`DSS’s “idle words” argument is a red herring. ...................................15
`1.
`Neve is cited to expressly show that synchronizing a
`
`base station and peripheral units was well-known. ................. 15
`Natarajan’s HDLC protocol does not use idle words. ............. 16
`2.
`
`IV. DSS’s “low duty cycle” argument is meritless............................................. 18
`A. Mr. Dezmelyk’s definition of “duty cycle” is nonsensical. ................18
`B.
`DSS’s proposed claim construction that “low duty cycle” is
`less than 10% is arbitrary and unduly narrow. ....................................20
`DSS improperly truncates the time period for calculating
`Natarajan’s duty cycle. ........................................................................22
`The Board should not give any weight to Mr. Dezmelyk’s testimony. ....... 23
`A. Mr. Dezmelyk’s testimony lacks credibility. ......................................23
`B. Mr. Dezmelyk admits he is not an expert in HDLC. ..........................24
`C. Mr. Dezmelyk bases his opinions on inaccurate assumptions. ...........24
`VI. Conclusion .................................................................................................... 25
`
`V.
`
`E.
`
`C.
`
`ii
`
`

`
`IPR2015-00369
`U.S. Pat. No. 6,128,290
`
`I.
`
`Introduction
`Claims 1-4 of the ’290 patent at issue in this inter partes review are merely a
`
`combination of well-known concepts. Each and every limitation is either expressly
`
`disclosed in the prior art or would have been plainly obvious to a person of ordi-
`
`nary skill in the art (“POSA”).
`
`DSS’s sole argument in its Patent Owner’s Response is that the combination
`
`of Natarajan and Neve does not teach or suggest a server transmitter that operates
`
`in “low duty cycle RF bursts.” Although this term was not commonplace, the tech-
`
`nical features it describes were well-known to those of ordinary skill in the art.
`
`DSS’s argument is flawed for at least four reasons. First, Natarajan teaches
`
`or suggests a server transmitter operating in “low duty cycle RF bursts.” Second,
`
`DSS bases its argument on the inaccurate premise that HDLC is inconsistent with
`
`low duty cycle RF bursts. In particular, DSS erroneously assumes that HDLC uses
`
`idle words. Third, DSS uses faulty logic to define “low duty cycle” and imposes an
`
`arbitrary 10% maximum threshold. And fourth, DSS bases its positions on the tes-
`
`timony Mr. Dezmelyk, which lacks credibility, particularly because he admits that
`
`he is not an expert in HDLC.
`
`Accordingly, DSS’s argument is meritless. Apple has shown by a prepon-
`
`derance of the evidence that claims 1-4 of the ’290 patent are unpatentable and the
`
`Board should enter judgment in accordance therewith.
`
`- 1 -
`
`

`
`IPR2015-00369
`U.S. Pat. No. 6,128,290
`II. Natarajan teaches or suggests a server transmitter operating in “low
`duty cycle RF bursts,” as recited in claim 1 of the ’290 patent.
`
`The vague term “low duty cycle RF bursts” is not defined in the ’290 patent.
`
`As the Board correctly recognized in the Institution Decision, under the broadest
`
`reasonable interpretation of this term, Natarajan’s “scheduled multi-access protocol
`
`in which time is divided into fixed-length frames, along with Natarajan’s descrip-
`
`tion of frames being divided into slots and multiple subframes” demonstrates that
`
`Natarajan discloses “said server and peripheral transmitters being energized in low
`
`duty cycle RF bursts.” (Institution Decision, p. 18; Hu Dec. ¶ 43.) Indeed, like the
`
`’290 patent, Natarajan discloses that “[s]cheduled access multiaccess protocols can
`
`be implemented to effectively conserve battery power by suitable control of the
`
`state of transmitter and receiver units at the portable units (i.e., by scheduling when
`
`they should be turned ON or OFF)…the transmitter (or receiver) consumes power
`
`only when it is actively transmitting a message (or actively receiving a message).”
`
`(Natarajan, 3:59-4:6.) This type of communication operates in “low duty cycle RF
`
`bursts.” (Grimes Dec. ¶ 115; Hu Dec. ¶ 43.)
`
`DSS argues that Natarajan only describes that the mobile units operate in
`
`this manner. (POR, p. 16.) But a POSA would have understood that, similarly,
`
`when the base station is not transmitting, its transmitter is powered off. (Grimes
`
`Dec. ¶¶ 27, 115-116; Grimes Depo. (DSS 2015), 68:5-12, 74:7-19, 75:21-76:3; Hu
`
`Dec. ¶ 44.) In Natarajan, “[m]ost users are very likely to be inactive (both Trans-
`
`- 2 -
`
`

`
`IPR2015-00369
`U.S. Pat. No. 6,128,290
`mit-Inactive and Receive-Inactive) most of the time for most applications. This is
`
`primarily due to the bursty nature of data communication traffic.” (Natarajan,
`
`6:41-44 (emphasis added).) A POSA would have understood this to mean that Na-
`
`tarajan’s base station and mobile units operate in low duty cycle RF bursts. (Hu
`
`Dec. ¶ 45.)
`
`Moreover, DSS acknowledges that Natarajan explicitly discloses that the
`
`mobile unit transmitters operate in “low duty cycle RF bursts.” (POR, p. 16;
`
`Dezmelyk Dec. ¶ 31.) So even if not expressly taught by Natarajan, it would have
`
`been plainly obvious to a POSA to have the base station operate in an analogous
`
`manner. (Hu Dec. ¶ 45.) Because the base and mobile stations have the same phys-
`
`ical structure, this would have been no more than using a known technique to im-
`
`prove similar devices in the same way. (Natarajan, 3:7-8; Hu Dec. ¶ 45.) The “low
`
`duty cycle RF bursts” limitation of claim 1 is not novel. (Hu Dec. ¶ 45.)
`
`III. HDLC is consistent with low duty cycle RF bursts.
`DSS alleges that the High-Level Data Link Control (HDLC) packet structure
`
`disclosed in Natarajan is inconsistent with a server transmitter being energized in
`
`low duty cycle RF bursts. (POR, pp. 20-22.) This is false. First, the ’290 patent’s
`
`preferred embodiment uses HDLC. Second, DSS relies on testimony from Mr.
`
`Dezmelyk, who admits he is not an expert in HDLC. Third, a POSA would have
`
`been directed to Schwartz for information on Natarajan’s HDLC protocol and un-
`
`- 3 -
`
`

`
`IPR2015-00369
`U.S. Pat. No. 6,128,290
`derstood that HDLC is consistent with low duty cycle RF bursts. Fourth, DSS con-
`
`cocts inaccurate, piecemeal arguments that are inconsistent with the HDLC proto-
`
`col. And fifth, Natarajan’s HDLC protocol does not use idle words.
`
`A. The preferred embodiment in the ’290 patent uses HDLC.
`DSS’s argument that HDLC is inconsistent with low duty cycle RF bursts is
`
`meritless because the preferred embodiment of the ’290 patent itself discloses us-
`
`ing HDLC for communication between the PDA and PEAs. If DSS’s argument is
`
`believed, then the ’290 patent’s preferred embodiment would not be enabled.
`
`The “basic scheme” of the ’290 patent’s frame structure is “a form of time
`
`division multiple access (TDMA).” (’290 patent, 5:45-50.) The ’290 patent states
`
`that “[a]s will be understood by those skilled in the art, the TDMA system is
`
`greatly facilitated by the establishment of a common frame time base between PEA
`
`and PDA.” (Id. at 7:63-65 (emphasis added).) This is accomplished using synchro-
`
`nization beacons (SBs). (Id. at 7:65-67.) Before receiving the SBs, a PEA is asso-
`
`ciated with the PDA using a succession of Attachment Beacons (ABs), which are
`
`“composed of RF bursts,” broadcast from the PDA to the PEAs. (Id. at 9:8-16,
`
`9:66-10:2.) “This succession of ABs forms an HDLC channel using bit-stuffing to
`
`delineate the beginning and end of a packet.” (Id. at 10:2-4 (emphasis added).)
`
`So, the ’290 patent uses HDLC to transmit and receive RF bursts. (Hu Dec.
`
`¶¶ 48-49.) Thus, the ’290 patent itself shows that DSS’s argument is fallacious.
`
`- 4 -
`
`

`
`IPR2015-00369
`U.S. Pat. No. 6,128,290
`B. DSS relies on the testimony of Mr. Dezmelyk, who admits he is
`not an expert in HDLC.
`
`DSS relies on Mr. Dezmelyk’s testimony to support its argument that HDLC
`
`is inconsistent with a server transmitter operating in low duty cycle RF bursts. (See
`
`e.g., POR, pp. 21-22 (citing Dezmelyk Dec. ¶ 34.).) But Mr. Dezmelyk admits he
`
`is not an expert in HDLC.
`
`Q.· Are you an expert in the HDLC protocol?
`
`A.· No.·I would not say I am an expert in that area.
`
`(Dezmelyk Depo., 26:15-16.)
`
`Mr. Dezmelyk is not an inventor on any patents related to HDLC protocol;
`
`he has not received any industry awards related to HDLC protocol; and he has
`
`never lectured on HDLC protocol. (Id. at 19:10-20:4.) In fact, this IPR and the re-
`
`lated district court case are the only matters he recollects working on that are even
`
`related more generally to wireless communication (id. at 21:1-22), which he asserts
`
`is “a hugely broad topic area” (id. at 26:1-2). Although his standard practice is to
`
`attach his Curriculum Vitae with a declaration, he did not here. (Id. at 16:5-7.)
`
`There is no evidence that Mr. Dezmelyk is qualified to opine on HDLC. To the
`
`contrary, by his own admission, Mr. Dezmelyk is not an expert in HDLC. And his
`
`understanding of the HDLC protocol is factually inaccurate. (Hu Dec. ¶ 50.) Ac-
`
`cordingly, DSS’s reliance on his testimony on HDLC must be disregarded.
`
`- 5 -
`
`

`
`IPR2015-00369
`U.S. Pat. No. 6,128,290
`C. A POSA would have looked to Schwartz for information on Nata-
`rajan’s HDLC protocol and understood that it is consistent with
`low duty cycle RF bursts.
`
`erence for information on Natarajan’s HDLC protocol–when
`forming his opinions.
`
` Mr. Dezmelyk did not consider Schwartz–the most logical ref-1.
`
`A POSA need not look any further than Natarajan itself for direction on
`
`more specifics about Natarajan’s HDLC protocol. Natarajan describes that infor-
`
`mation packets are transmitted between the base station and mobile stations:
`
`Packets received or to be sent are held in data storage 68 and commu-
`nicated to or from the RF transceiver 54 via interface 58 under control
`of serial channels…. The function of these serial channels is to encap-
`sulate data and control information in an HDLC (high-level data link
`control) packet structure and provide the packet in serial form to the
`RF transceiver 54. For more information on the HDLC packet struc-
`ture, see, for example, Mischa Schwartz, Telecommunication Net-
`works: Protocols, Modeling and Analysis, Addison-Wesley (1988).
`(Natarajan at 3:28-40 (emphasis added).)
`So, the Schwartz book (APL 1012) is the most logical resource for a POSA
`
`to consult for information on Natarajan’s HDLC packet structure. (Hu Dec. ¶ 51.)
`
`Indeed, Mr. Dezmelyk acknowledged that a POSA would have access to Schwartz.
`
`(Dezmelyk Depo., 72:12-22.) Yet Mr. Dezmelyk never looked at Schwartz when
`
`considering how Natarajan’s HDLC packet structure operates. (Id. at 71:11-73:13.)
`
`Nor does DSS reference Schwartz. Instead, DSS forgoes logic, formulating a con-
`
`voluted argument that defies technical accuracy. (See infra, Section III.D.)
`
`- 6 -
`
`

`
`IPR2015-00369
`U.S. Pat. No. 6,128,290
`Natarajan’s HDLC protocol is consistent with low duty cycle
`RF burst communication.
`
`2.
`
`
`Schwartz provides a concise summary of the HDLC frame format:
`
`The standard frame format for HDLC (ADCCP and SDLC have the
`same format) appears in Fig. 4-9. Note that the number of overhead
`(control) bits is ℓ' = 48, just the number used earlier for calculations.
`The eight-bit flag sequence 01111110 that appears at the beginning
`and end of a frame is used to establish and maintain synchroniza-
`tion. Because the flag appears at the beginning and end of the frame
`there is no need to prescribe an information field structure. The in-
`formation field (packet) delivered from the network layer above can
`be any desired number of bits. Extended versions of the frame struc-
`ture of Fig. 4-9 are available as well: The address, control, and
`block-check fields can all be increased to allow additional addressing,
`improved error detection, and increased sequence numbers. Since the
`flags appearing at the beginning and end of a frame contain six con-
`secutive ones, that sequence may not appear anywhere else in the
`frame. Bit stuffing is used to eliminate this possibility: a zero is in-
`serted at the transmitter any time that five ones appear outside the F
`fields. The zeros are removed at the receiver. If seven ones appear
`anywhere in the frame (six ones followed by an additional one), the
`frame is declared in error.
`
`(Schwartz, pp. 135-136 (emphasis added), Fig. 4-9.)
`
`- 7 -
`
`
`
`

`
`IPR2015-00369
`U.S. Pat. No. 6,128,290
`Schwartz also describes that “[w]hen the transmitter reaches its maximum
`
`sequence number it is forced to stop transmitting until a frame in the reverse direc-
`
`tion is received, acknowledging an outstanding packet.” (Id. at 136 (emphasis add-
`
`ed).) Thus, the transmitter does not continuously transmit under HDLC protocol.
`
`(Hu Dec. ¶ 53.) The discontinuous transmission of the HDLC protocol is also
`
`shown in Schwartz’s throughput calculation illustrated in Figure 4-13, showing pe-
`
`riods (e.g., between I10 and I30) where the transmitter is idle (i.e., not transmit-
`
`ting). (Id.) Figure 4-13 illustrates the maximum throughput, so the amount of time
`
`the transmitter is not transmitting is at a minimum. (Schwartz, p. 142 (“This station
`
`is assumed in addition to be in a saturated state: It always has frames to send. As
`
`noted earlier, this provides the maximum possible throughput.” (emphasis origi-
`
`nal)); Hu Dec. ¶ 53.) When the primary station is not saturated, there will be addi-
`
`tional periods where the transmitter is not transmitting. (Hu Dec. ¶ 53.)
`
`Annotated Figure 4-13 of Schwartz
`
`
`
`- 8 -
`
`

`
`IPR2015-00369
`U.S. Pat. No. 6,128,290
`Thus, the HDLC protocol described in Schwartz is consistent with low duty
`
`cycle communication. (Hu Dec. ¶ 54.) Moreover, as illustrated, for example by the
`
`I-frames I00-I30 in Figure 4-13, the transmissions occur in “bursts.” (Id.)
`
`Schwartz also describes the three modes of operation for the HDLC proto-
`
`col. (Schwartz, pp. 137-138.) The asynchronous balanced mode (ABM) is for
`
`point-to-point transmission only, and therefore inapplicable to Natarajan, which is
`
`point-to-multipoint communication. (Id.; Hu Dec. ¶ 55.) The normal response
`
`mode (NRM) and asynchronous response mode (ARM) are used for point-to-
`
`multipoint operation. (Schwartz, p. 137; Hu Dec. ¶ 55.)
`
`A POSA would have understood that Natarajan operates using the ARM.
`
`(Hu Dec. ¶ 55.) ARM provides that “the secondary station does not need permis-
`
`sion from the primary station to initiate transmission.” (Schwartz, p. 137.) This oc-
`
`curs in Period C in Natarajan, which allows for “bursty data traffic in a contention
`
`mode from mobile units to base station.” (Natarajan, 4:36-37 (emphasis added);
`
`Hu Dec. ¶ 55.) In Period C, the mobile units initiate transmission without permis-
`
`sion from the base station. (Hu Dec. ¶ 55.)
`
`D. DSS and Mr. Dezmelyk concoct inaccurate piecemeal arguments
`from excerpts of unrelated references that are inconsistent with
`each other and inconsistent with the operation of HDLC.
`
`DSS and Mr. Dezmelyk did not consider the most logical reference–
`
`Schwartz–for information on Natarajan’s HDLC protocol. Instead, they piece to-
`
`- 9 -
`
`

`
`IPR2015-00369
`U.S. Pat. No. 6,128,290
`gether a patchwork of references in an ill-fated effort alleging that Natarajan does
`
`not teach low duty cycle RF bursts because it employs HDLC. (POR, pp. 20-22;
`
`Dezmelyk Dec. ¶¶ 34-35.) This argument is meritless because the disparate refer-
`
`ences do not support the asserted premise. (Hu Dec. ¶ 56.)
`
`First, DSS’s reliance on the excerpt from DSS 2010 is misplaced. DSS as-
`
`serts that the cited definition of “bit-oriented framing” shows that HDLC “involves
`
`continuous outbound transmission.” (POR, pp. 20-21.) But DSS neglects to
`
`acknowledge the very first sentence of the cited section, stating that “[a] point-to-
`
`point connection between two computers or devices consists of a wire in which
`
`data is transmitted as a stream of bits.” (DSS 2010, p. 549 (emphasis added).) DSS
`
`2010 refers to point-to-point wired communication, not to a point-to-multipoint
`
`wireless system as taught in Natarajan. (Hu Dec. ¶ 56.) There are fundamental dif-
`
`ferences and unique challenges between point-to-point wired systems and point-to-
`
`multipoint wireless systems–the features are not simply interchangeable. (Id. at 57-
`
`60.) For example, a continuous transmission of so-called “idle words” to maintain
`
`synchronization when there is no data to transmit–suitable for an isolated point-to-
`
`point wired connection for design simplicity and reliability–would be detrimental
`
`to a point-to-multipoint wireless connection because it would interfere with the
`
`carefully designed scheduling, waste power, decrease the system data rate, and pol-
`
`lute the wireless channel potentially shared by many devices. (Id. at 59.)
`
`- 10 -
`
`

`
`IPR2015-00369
`U.S. Pat. No. 6,128,290
`DSS also asserts that the “HLDC [sic] packet structure is used to transmit
`
`‘long strings of data at one time,’” and therefore is inconsistent with “small” RF
`
`bursts. (POR, p. 21 (emphasis original).) DSS misquotes this passage, which actu-
`
`ally states: “[t]his type of framing allows the sender to transmit a long string of
`
`bits at one time.” (DSS 2010, p. 549 (emphasis added).) A POSA would have un-
`
`derstood that this does not mean that HDLC only sends a long string of bits; nor
`
`does it suggest that transmitting a string of bits is not a “burst.” (Hu Dec. ¶ 61.) Ra-
`
`ther, a sender with data bits to transmit “at one time” does not have to break the
`
`string into smaller packets before the string of bits reach the data link layer. (Id.)
`
`Moreover, this same reference discloses that the number of bits in the Infor-
`
`mation Field of an HDLC frame is “variable.” (APL 1013, p. 582, Figure H-2.) Mr.
`
`Dezmelyk, however, did not consider any portion of this reference other than what
`
`DSS provided to him.1 (Dezmelyk Depo., 101:4-102:7.)
`
`
`1 Regarding DSS 2010, Mr. Dezmelyk asserts he was “trying to get [a reference] in
`
`the right time frame” (Dezmelyk Depo., 101:18-19), but the copyright date for this
`
`reference is 2001–well after the ’290 patent’s filing date.
`
`- 11 -
`
`

`
`IPR2015-00369
`U.S. Pat. No. 6,128,290
`
`Annotated Figure H-2 of APL 1013
`
`
`
`This portion of DSS 2010 (APL 1013)–uncited by DSS–corroborates
`
`Schwartz. (Schwartz, p. 135 (“The information field (packet) delivered from the
`
`network layer above can be any desired number of bits.”) (emphasis added).) So,
`
`the HDLC frame format is compatible with a “burst” transmission. (Hu Dec. ¶ 62.)
`
`Next, Mr. Dezmelyk tries to tie the excerpt from DSS 2010 with an excerpt
`
`from DSS 2013, proposing that Natarajan teaches continuous transmission, which
`
`it does not. Mr. Dezmelyk notes that Natarajan’s HDLC packets are in “serial
`
`form.” (Dezmelyk Dec. ¶ 35 (citing APL 1003 at 3:36-37).) He alleges that serial
`
`communication systems “include a server transmitter transmitting idle words when
`
`no useful data is being transmitted.” (Dezmelyk Dec. ¶ 35.) But nowhere does DSS
`
`2013 use the term “idle words.”2 Mr. Dezmelyk instead provides a quote: “In syn-
`
`chronous transmission, groups of bits are combined into frames and frames are
`
`sent continuously with or without data to be transmitted.” (Id. (quoting DSS 2013,
`
`p. 2 (emphasis added)).) But this does not describe Natarajan’s HDLC protocol,
`
`
`2 Importantly, Natarajan’s system does not use idle words. (Section III.E., infra.)
`
`- 12 -
`
`

`
`IPR2015-00369
`U.S. Pat. No. 6,128,290
`which is asynchronous. (Hu Dec. ¶ 64.) Mr. Dezmelyk is wrong that a POSA
`
`would not have known whether Natarajan operates under a synchronous or asyn-
`
`chronous protocol. (Dezmelyk Depo., 99:12-20; Hu Dec. ¶ 64.) His assertion un-
`
`derscores his misunderstanding of HDLC, a subject on which he admits he is not
`
`an expert. (Dezmelyk Depo., 26:15-16; Hu Dec. ¶ 50.) As discussed above in Sec-
`
`tion III.C.2, a POSA would have understood that Natarajan operates under the
`
`asynchronous response mode (ARM) of HDLC. (Hu Dec. ¶ 64.)
`
`The very next sentences in DSS 2013 distinguish that “[i]n asynchronous
`
`transmission, groups of bits are sent as independent units with start/stop flags and
`
`no data link synchronization, to allow for arbitrary size gaps between frames.
`
`However, start/stop bits maintain physical bit level synchronization once detect-
`
`ed.” (DSS 2013, p. 2 (emphasis added).) These “start/stop flags” for synchroniza-
`
`tion corroborate Schwartz, that an “eight-bit flag sequence 01111110 that appears
`
`at the beginning and end of a frame is used to establish and maintain synchroniza-
`
`tion.” (Schwartz, p. 135; Hu Dec. ¶ 65.) Thus, in HDLC, transmission is not con-
`
`tinuous, as DSS asserts; rather, there are gaps between frames. (Hu Dec. ¶ 65.)
`
`Finally, Mr. Dezmelyk pieces an excerpt from DSS 2014 together with the
`
`excerpts from DSS 2010 and DSS 2013. (Dezmelyk Dec. ¶ 35 (quoting DSS 2014
`
`at Section 2.5.6. (“When transmitting, the Asynchronous HDLC controller will
`
`transmit IDLE characters (characters consisting of only “1”s) when no data is
`
`- 13 -
`
`

`
`IPR2015-00369
`U.S. Pat. No. 6,128,290
`available for transmission.”)).) This association is inapt for two reasons. First, DSS
`
`2014 relates to point-to-point communications, not point-to-multipoint communi-
`
`cations as in Natarajan. (DSS 2014, p. 4 (“This protocol is typically used as the
`
`physical layer for the Point-to-Point (PPP) protocol.”); Hu Dec. ¶ 66.) As discussed
`
`above, a point-to-point wired connection and a point-to-multipoint wireless con-
`
`nection are two fundamentally different types of data communication with distinct
`
`design challenges and principles. (See supra, Section III.D; Hu Dec. ¶¶ 57-60, 66.)
`
`Second, DSS 2014 is a technical manual for a proprietary Motorola product, which
`
`uses a modified “HDLC-like” protocol. (Hu Dec. ¶ 66.) This is evident because it
`
`can repeatedly transmit “characters consisting of only ‘1’s.” (Id.) This violates the
`
`standard HDLC protocol, where “[i]f seven ones appear anywhere in the frame
`
`(six ones followed by an additional one), the frame is declared in error.”
`
`(Schwartz, p. 136 (emphasis added); see also APL 1013, p. 582 (“If any portion of
`
`the data in the frame contains more than five 1 bits, a zero-bit insertion technique
`
`inserts a 0 bit to ensure that data is not mistaken for a flag.”).)
`
`Thus, DSS 2014 does not disclose the standard HDLC protocol and so it
`
`cannot be used to support DSS’s position. (Hu Dec. ¶¶ 67-68.) Yet Mr. Dezmelyk
`
`relies on this document as supposedly disclosing “the HDLC spec,” which it does
`
`not. (Dezmelyk Depo., 69:17-71:1; Hu Dec. ¶ 68.) Because Mr. Dezmelyk bases
`
`his opinions on DSS 2014, which does not disclose the standard HDLC protocol,
`
`- 14 -
`
`

`
`IPR2015-00369
`U.S. Pat. No. 6,128,290
`
`his opinions on HDLC are unsupported. (Hu Dec. ¶ 68.)
`
`The piecemeal argument cobbled together by DSS and Mr. Dezmelyk pro-
`
`vides misinformation about how HDLC operates. Indeed, excluded portions of the
`
`very references they cite belie their contentions regarding HDLC, and thus their
`
`assertions regarding Natarajan. Accordingly, DSS’s arguments are baseless.
`
`E. DSS’s “idle words” argument is a red herring.
`1.
`Neve is cited to expressly show that synchronizing a base sta-
`
`tion and peripheral units was well-known.
`
`Apple’s Petition includes Neve in combination with Natarajan because “Na-
`
`tarajan does not explicitly describe synchronizing the mobile units with the base
`
`station.” (Petition, p. 40.) However, Natarajan teaches numerous reasons that “co-
`
`ordinated timing of transmissions is important”. (See id. at 40-41 (citing Grimes
`
`Dec. ¶ 92); Hu Dec. ¶ 69.) Thus, a POSA would have been motivated to precisely
`
`synchronize the mobile units and base unit, leading a POSA to Neve–an example
`
`of a conventional synchronization technique. (Petition, pp. 41-42; Hu Dec. ¶ 69.)
`
`Indeed, although Natarajan does not explicitly disclose synchronizing the
`
`mobile units with the base station, the HDLC protocol, for example as described in
`
`Schwartz, explains that the “standard frame format for HDLC” has an “eight-bit
`
`flag sequence 01111110 that appears at the beginning and end of a frame [that] is
`
`used to establish and maintain synchronization.” (Schwartz, p. 135 (emphasis
`
`added).) Thus, Natarajan’s HDLC protocol contemplates a synchronization mech-
`
`- 15 -
`
`

`
`IPR2015-00369
`U.S. Pat. No. 6,128,290
`anism. (Hu Dec. ¶ 70.) To make clear that synchronization was well-known, Neve
`
`is combined with Natarajan. As Mr. Dezmelyk acknowledges, synchronization was
`
`within the skill of a POSA. (Dezmelyk Depo., 93:6-16; Hu Dec. ¶ 70.)
`
`Natarajan’s HDLC protocol does not use idle words.
`
`2.
`
`Again, Apple cites Neve to make clear that synchronization was convention-
`
`al in communication networks before the ’290 patent’s priority date. DSS uses
`
`Neve as a distraction, conjuring a theory that Natarajan uses idle words, as dis-
`
`closed in Neve. DSS’s theory is unfounded. (Hu Dec. ¶¶ 71-72.)
`
`To begin with, even Neve’s master station does not continuously transmit
`
`idle words. (Id. at 71.) The Board correctly found that Neve “do[es] not suggest
`
`continuous transmission from the master station, but instead transmission of idle
`
`words in the event that there is no data required to be transmitted in the time slots
`
`specifically allocated for transmission by the server” and that “the master station
`
`performs different functions during different time slots, only certain of which in-
`
`volve transmission.” (Institution Decision, p. 20 (emphasis added); Hu Dec. ¶ 71.)
`
`Idle words are not continuously transmitted in Neve. (Hu Dec. ¶ 71.)
`
`DSS contorts the inclusion of Neve as somehow suggesting that Natarajan
`
`operates identically, which it does not. (Id. at 72.) In particular, DSS suggests that
`
`Natarajan transmits idle words (POR, pp. 32-33), which it does not. (Hu Dec. ¶
`
`72.) Indeed, neither Natarajan, nor Schwartz’s discussion of HDLC for that matter,
`
`- 16 -
`
`

`
`IPR2015-00369
`U.S. Pat. No. 6,128,290
`mention idle words. This is likely because using idle words in Natarajan would be
`
`pointless. (Id. at 73.) First, HDLC has a mechanism–start/stop flags–for synchroni-
`
`zation, so idle words would be redundant. (Id.) Second, if Natarajan used idle
`
`words, they would be transmitted “out into the air and there is nobody really pay-
`
`ing attention to it.” (Dezmelyk Depo., 84:22-85:1 (emphasis added).) This is illog-
`
`ical. (Hu Dec. ¶ 73.) Mr. Dezmelyk tries to reconcile, asserting that when receivers
`
`turned back on, they could synchronize onto the stream of idle words. (Dezmelyk
`
`Depo., 85:2-14.) But Natarajan discloses that “each receiving mobile unit can
`
`compute exactly when it should be ready to receive packets from the base sta-
`
`tion…to wake itself up at its designated time for receiving data.” (Natarajan, 4:67-
`
`5:3 (emphasis added).) Thus, a POSA would understand that Natarajan does not
`
`operate in the manner Mr. Dezmelyk suggests. (Hu Dec. ¶ 73.)
`
`Mr. Dezmelyk even acknowledges that Natarajan does not actually disclose
`
`continuous transmission: “…using the Natarajan example, there is a period of time
`
`for outbound data traffic. There will be effectively, if there is a lot of data sent,
`
`continuous transmission over across that interval.” (Dezmelyk Depo., 62:3-6 (em-
`
`phasis added).) And although he is factually inaccurate because the HDLC proto-
`
`col does not use idle words, Mr. Dezmelyk asserts that HDLC only “typically” us-
`
`es idle words. (Id. at 69:10-16.)
`
`- 17 -
`
`

`
`IPR2015-00369
`U.S. Pat. No. 6,128,290
`Moreover, Mr. Dezmelyk asserts that systems using idle words can have, for
`
`example, ten consecutive 1’s. (Dezmelyk Depo., 98:1-13.) Again, Mr. Dezmelyk is
`
`wrong. (Hu Dec. ¶ 74.) HDLC cannot have more than six consecutive 1’s, other-
`
`wise an error is declared. (Schwartz, p. 135-136; Hu Dec. ¶ 74.) That is why, ac-
`
`cording to the book cited by DSS, “[i]f any portion of the data in the frame con-
`
`tains more than five 1 bits, a zero-bit insertion technique inserts a 0 to ensure that
`
`data is not mistaken for a flag,” which is 01111110. (APL 1013, p. 582.)
`
`Nothing in Natarajan or the HDLC protocol suggests using idle words. (Hu
`
`Dec. ¶ 75.) And, in fact, a POSA would understand that HDLC does not use idle
`
`words. (Id.) DSS’s argument on this point is meritless.
`
`IV. DSS’s “low duty cycle” argument is meritless.
`A. Mr. Dezmelyk’s definition of “duty cycle” is nonsensical.
`Mr. Dezmelyk asserts in his declaration that “duty cycle” should be con-
`
`strued to mean “the ratio of the duration during which the server transmitter is en-
`
`ergized to the total duration designated for outbound transmissions–from the server
`
`unit to the peripheral units.” (Dezmelyk Dec. ¶ 23.) But in his deposition, he pro-
`
`vides two different, yet equally confounding, explanations for determining duty
`
`cycle. In each case, the ’290 patent would have a 100% duty cycle.
`
`First, in his redirect testimony, Mr. Dezmelyk asserts that “duty cycle is
`
`measured based