`
`IN THE UNITED STATES PATENT AND TRADEMARK OFFICE
`
`PATENT:
`
`6,424,625
`
`INVENTOR:
`
`Larsson, P. et al.
`
`TITLE:
`
`METHOD AND APPARATUS FOR DISCARDING PACKETS
`
`IN A DATA NETWORK HAVING AUTOMATIC REPEAT REQUEST
`
`Mail Stop PATENT BOARD
`Patent Trial and Appeal Board
`U.S. Patent & Trademark Office
`
`PO. Box 1450
`
`Alexandria, VA 22313-1450
`
`DECLARATION OF HARRY BIMS, PH.D.
`
`1, Harry Bims, declare as follows:
`
`General Background
`
`1.
`
`My name is Harry Bims.
`
`I have been asked to offer opinions
`
`regarding whether the claims of U.S. Patent No. 6,424,625 (the ‘625 patent) are
`
`anticipated or would have been obvious in View of the prior art, and to review a
`
`Petition for Inter Partes Review of the ‘625 patent.
`
`2.
`
`I received my B.S. in computer and systems engineering from
`
`Rensselaer Polytechnic Institute in 1985, my M.S. in electrical engineering from
`
`Stanford University in 1988, and my Ph.D. in electrical engineering from Stanford
`
`University in 1993. Since receiving my doctorate, I have worked on a number of
`
`ActiveUS 1 1654-8286v.l
`
`1
`
`BROADCOM 1006
`
`
`
`wireless and mobile technologies, including wireless pagers, wireless home LAN
`
`protocols, cellular products including 2.5G and 3G products, wireless network
`
`infrastructures based on the 802.11 wireless specification, and wireless networks in
`
`the 4G technology known as WiMAX, an implementation of 802.16.
`
`3.
`
`I have been actively involved in the development of the 802.16
`
`standards, which is a series of wireless broadband standards written by the Institute
`
`of Electrical and Electronics Engineers (IEEE), including as a vice-chair of the
`
`802.16 working group, and chair of two task groups. Previously, I was the vice-
`
`chair and secretary of the IEEE 802.16h License Exempt Task Group.
`
`4.
`
`I am currently working as both a technology consultant in the industry
`
`and an expert consultant for litigation matters.
`
`5.
`
`I began my technical career in 1992 just before completing my Ph.D.
`
`as one of the first employees at Glenayre Technologies, where I worked until 1998.
`
`While at Glenayre, I designed and built a 4—channel wireless pager demonstration
`
`based on the ReFLEX wireless protocol developed by Motorola, which led to an
`
`award for Narrowband Personal Communications Service (PCS) development.
`
`I
`
`invented, designed, and built a two-way pager test system for the ReFLEX
`
`protocol that was deployed around the country for testing pagers. Additionally, I
`
`co-developed a wireless application protocol for sending and receiving encrypted
`
`ActiveUS l 1654-8286v.l
`
`
`
`email messages over the paging channel, which was ultimately deployed for
`
`government agencies.
`
`6.
`
`In 1999, I was a member of the technical staff at T—SPAN Systems
`
`Corporation LLC, where I designed a wireless home LAN protocol. In 1999 I also
`
`served as a technical leader to Gigabit Wireless, Inc., where I lead the Wireless
`
`Media Access Control (MAC) design group. My work at Gigabit Wireless
`
`involved analyzing competing wireless MAC protocol standards, creation of a
`
`proprietary MAC protocol specification document, simulation of the protocol, and
`
`ultimate implementation of the protocol in a prototype.
`
`I also participated in
`
`meetings for the 802.16 standards starting at about that time.
`
`7.
`
`From 1999 to 2001, I served as the director of software architecture at
`
`Symmetry Communications Systems LLC, where I was responsible for the
`
`software architecture for their core products for the GPRS market.
`
`In 2001, I also
`
`worked as an entrepreneur in residence at the venture capital firm Bay Partners
`
`LLC, where I served as a technology expert to the partners of the firm on a range
`
`of wireless and networking subjects.
`
`8.
`
`From 2001 to 2004 I founded my own company, AirFlow Networks,
`
`Inc. LLC, where I served as CEO and CTO. AirFlow Networks was involved with
`
`a wireless network infrastructure based on the 802.11 wireless specification.
`
`ActiveUS I 1654-8286v.l
`
`
`
`9.
`
`From 2007 to 2009 I worked as a technology consultant to Apple,
`
`Inc., including participating in IEEE 802.16 standards meetings.
`
`10.
`
`I am a named inventor on nineteen U.S. Patents that involve various
`
`aspects of wireless and mobile communications. Examples of my patents include
`
`U.S. Patent No. 6,788,658 entitled “Wireless communication system architecture
`
`having split MAC layer,” which issued on September '7, 2004; and U.S. Patent No.
`
`6,557,134 entitled “ARQ method for wireless communication,” which issued on
`
`April 29, 2003; and most recently, U.S. Patent No. 8,468,426 entitled
`
`“Multimedia—aware quality-of-service and error correction provisioning,” which
`
`issued on June 18, 2013. Additionally, I have authored or co—authored a number of
`
`articles in the fields of electrical engineering and computer science.
`
`11.
`
`I have been a member or vice~chair of numerous associations,
`
`including the chair of the Silicon Valley Chapter of the IEEE Engineering
`
`Management Society, and vice-chair of the 802.16 Working Group of the IEEE
`
`802 Standards Development Committee.
`
`12.
`
`A copy of my latest curriculum vitae (CV) is attached as Appendix A.
`
`13.
`
`I am being compensated at my normal consulting rate for my work.
`
`My compensation is not dependent on and in no way affects the substance of my
`
`statements in this Declaration.
`
`ActiveUS 11654-8286v.l
`
`
`
`14.
`
`I have no financial interest in Petitioner.
`
`I have been informed that
`
`Ericsson purports to own the patent for which review is requested.
`
`1 have no
`
`financial interest in Ericsson.
`
`U.S. Patent No. 6,424,625
`
`15.
`
`I have reviewed and understand the specification, claims, and file
`
`history of the ‘625 patent.
`
`16.
`
`I have considered certain issues from the perspective of a person of
`
`ordinary skill in the art. In my opinion, a person of ordinary skill in the art for the
`
`‘625 patent would be a person with a bachelor's or graduate degree in a relevant
`
`field, such as electrical or computer engineering or computer science, with some
`
`amount of work experience in communications.
`
`Technical Background
`
`17.
`
`Automatic repeat request (“ARQ”) refers to a class of protocols for
`
`packet data networks that allows a transmitter to automatically retransmit packets
`
`(known also as protocol data units (“PDUS”), cells, frames, or fragments thereof)
`
`to a receiver when those packets are not received or are incorrectly received by the
`
`receiver (e.g., ‘625 patent, 1:34-37, Ex. 1001). A receiver can transmit a positive
`
`acknowledgement (“ACK” or “PACK”) that one or more data units have been
`
`received (and which can include a request for a data unit), and/or a negative
`
`acknowledgement (“NAK” or “NACK”) indicating that one or more packets have
`
`ActiveUS l 1654-8286v.l
`
`
`
`not been received. The transmitter can retransmit the packets if there is a NACK,
`
`or if there is no PACK (e.g., because a timer in the transmitter times out). (Id. at
`
`1:34-49 and 2:28-57, Ex. 1001).
`
`18.
`
`ARQ systems typically associate each packet with a sequence number
`
`(“SN”). The receiver uses the SN to track which packets have been sent,
`
`acknowledged, received, or not received, as well as to allow a receiver to
`
`reassemble a proper order of the packets if packets can be sent out of order.
`
`(See,
`
`e.g., id. at 1:50-2:2, Ex. 1001). The sequence numbers are used in a modulo
`
`manner, and thus can be represented as a circular buffer. That is, they are
`
`numbered, for example, 0, 1, 2, 30, 31, 0, 1, 2,
`
`Thus, the numbers should be
`
`viewed in the sense of earlier and later rather than higher and lower, although both
`
`terms are used.
`
`(See, id. at 7:10-16, Ex. 1001).
`
`19.
`
`There are a number of different types of ARQ systems, including
`
`systems known as “stop—and-wait,” Go-Back—N, and selective reject systems.
`
`Examples of such systems are identified in the Background of the ‘625 patent
`
`(‘625 patent, 1:13-3:40, Ex. 1001); other examples are also well-known. (See, e.g.,
`
`Bertsekas, pp. 58-73, Ex. 1012).
`
`In an example of a Go-Back~N implementation,
`
`the transmitter sends up to :1 packets, and receives acknowledgements with a
`
`request number (RN) requesting the next packet. The RN messages cumulatively
`
`acknowledge packets with SN<RN and request RN=SN. For example, the
`
`ActiveUS l 1654-8286v.l
`
`
`
`transmitter sends packets #1 through #18, while the receiver send RN #2 through
`
`#14 to acknowledge positively received packets #1 through #13.
`
`In this example,
`
`if the transmitter can send and buffer up to 7 packets without receiving an RN
`
`message, it will continue to transmit packets #14 through #20, filling the buffer.
`
`The transmitter then must “go back 7” and retransmit packets #14 through #20.
`
`(See, e.g., Bertsekas, pp. 63-66, Ex. 1012).
`
`20. Another ARQ protocol is known as a selective reject (“SR”) system.
`
`In a SR ARQ system, the transmitter has a transmit window with a range of
`
`sequence numbers it can send (based on the size of a transmit buffer), and the
`
`receiver has a receive window of sequence numbers it can receive (based on the
`
`size of a receive buffer). In a system that has k-bits for the sequence numbers and
`
`thus 21‘ sequence numbers, the window size is often 21“, or one-half the possible
`
`sequence numbers. (See ‘625 patent at 2:57-65, Ex. 1001). Thus, if there are 32
`
`possible sequence numbers (5 bits), the transmit and receive windows often will
`
`have 16 spaces.
`
`21.
`
`If a transmitter successfully transmits packets with SN #0 though #20,
`
`16-bit transmit and receive windows will move forward, e. g., from #0-#15, to #1-
`
`#16, to #2-#17, etc. The Garrabrant prior art, Figures 7A and 7B (Ex. 1002),
`
`shows how a “valid [receive] window” and a corresponding “rejection window”
`
`move in a circular manner:
`
`ActiveUS 1 1654-8286v.l
`
`
`
`
`
`22. According to the Background of the ’625 patent, where packets are
`
`time-sensitive, such as speech or streaming video, the additional delay experienced
`
`at the receiver while waiting for multiple rounds of NACKS and retransmissions
`
`may have a more negative impact on application performance than simply
`
`discarding a missed packet and moving on to receive other packets.
`
`(See, e.g. , id.
`
`at 3:46-67, Ex. 1001).
`
`23.
`
`The ‘625 patent asserts that “current ARQ methods do not recognize
`
`and allow for situations where data packets have a limited lifetime [such as time
`
`sensitive packets, like streaming video], and therefore fail to minimize bandwidth
`
`usage by not sending (or resending) significantly delayed or outdated data
`
`packets.” (Id. at 4:9-13, Ex. 1001).
`
`24.
`
`The ‘625 patent says that it seeks to remedy this situation through the
`
`use of an “enforcement bit” that, in effect, tells the receiver to move its receive
`
`ActiveUS I 1654-8286v.l
`
`
`
`window forward to a new sequence number, and therefore to “release any
`
`expectation of receiving outstanding packets” with lower sequence numbers than
`
`that new sequence number. (Id. at 10:20-21, Ex. 1001).
`
`25.
`
`In the example above where there was a NACK for a packet with
`
`SN#14, but packets with SN #15 through #20 were correctly received, assume the
`
`transmitter decides to discard the packet with SN #14 rather than retransmitting it.
`
`According to the ‘625 patent, the transmitter could transmit a packet with SN #15
`
`with an enforcement bit set to TRUE. Once the receiver receives an enforcement
`
`bit set to TRUE, the receiver can forward packets with SN #15 through #20 to
`
`higher layers for processing, and move the lower end of its window up to a packet
`
`with SN #15, thereby skipping the packet with SN #14. The receiver would then
`
`move the lower end of its window up to the data unit with SN #21 as data units
`
`with SN #15 through #20 are correctly received and forwarded up to higher layers
`
`for processing. (See, e.g., id. 8:24-30, Ex. 1001).
`
`26.
`
`This process is shown, for example, in Garrabrant, Ex. 1002, Figures
`
`8A and 8B. As described in more detail below, the transmitter and receiver start
`
`with a valid window of #2 through #17 (Figure 8A). The transmitter
`
`unsuccessfully transmits, and then decides to discard, packets with SN #2 though
`
`SN #6. The transmitter sends a control message (lost) and a packet with SN #7,
`
`and then the window moves to SN #7 and then to SN #8 (Figure 8B).
`
`ActiveUS 1 1654-8286v.l
`
`
`
`
`
`27.
`
`An example of a window of the ‘625 patent, along with different
`
`sequence number values that are maintained in relation to the window, is
`
`represented in Figure 10B:
`
`
`
`Actual buffer content
`
`Max buffer size
` ....................
`' PL+
`PL+
`PL+
`PL+
`
`
`PL+
`S PL+
`Comp.
`Comp.
`Comp.
`N
`
`
`
`Header
`Header
`Header
`
`
`
`PDUs which have
`not yet been
`ISN re_transmmed
`
`
`
`TSNMAX’
`TSN
`
`PSN
`
`2
`2
`2
`5
`s
`2
`_
`5
`5
`5
`5
`3
`3
`3........3 E........2 S........E
`
`Discarded cells
`DSN
`
`BSN
`
`Camp.
`Header
`
`Camp.
`Header
`
`Comp.
`Header
`
`SW
`PDUs which have
`been re-transmitted
`
`Comp.
`Hader
`
`Comp.
`Headeri
`
`
`
`
`
`PDUs pending
`for transmission
`
`
`
`
`28.
`
`The receive window (W) extends from sequence number values DSN
`
`to TSN,.m, while the “actual buffer” in the transmitter extends from BSN to PSN.
`
`The receive window is typically half of the sequence number space; that is, if the
`
`SN has 5 bits, and thus there are 32 possible sequence numbers, the receive
`
`AetiveUS ll6548286v.1
`
`-10-
`
`
`
`window would have 16 values. (Id. at 5:58-63, 6:20-50, 7:42-47, Ex. 1001). The
`
`DSN in Figure 10B designates the SN below which all packets have been
`
`acknowledged. BSN designates the SN of the lowest non—discarded packet. If
`
`there are no discarded packets, then DSN = BSN. If there are discarded PDUs but
`
`the receiver has not yet been notified, then DSN < BSN, as shown above where
`
`there are 3 discarded packets in window W. (Id. at Figure 10B, 5:65-6:9, Ex.
`
`1001).
`
`29.
`
`TSN,,,,,, designates the top of the window, and has a SN of DSN + W.
`
`(7:42-47). The packets with SNS above TSNM, are PDUs have been buffered, but
`
`are “not allowed to be transmitted.” (Id. at 6:56—60, Ex. 1001). The transmitter
`
`holds packets greater than TSNmax in a buffer because these packets would be
`
`outside the receive window, and therefore cannot be accepted at the receiver.
`
`(Id.
`
`at 7:26-33 and 11:38-48, EX. 1001).
`
`30.
`
`Because the receiver does not yet know that the packets between DSN
`
`and BSN will be discarded, the receiver holds spaces for them. Therefore, the
`
`receive window would extend from DSN to DSN + W. The receive window does
`
`not have room for the packets from TSN to PSN, and therefore these packets
`
`delayed until DSN and perhaps also BSN can move forward.
`
`(Id. at 6:51-65, Ex.
`
`1001).
`
`ActiveUS l 1654-8286v.l
`
`11
`
`
`
`31.
`
`The ‘625 patent identifies 6 situations or rules for operation, referred
`
`to as “cases” (id. at 824-9210, Ex. 1001), which are claimed in dependent claim 6.
`
`(Id. at 11:23-12:64). In a “third case” identified in the ‘625 patent, for example,
`
`the transmitter receives a NACK for an expected sequence number (ESN1) that is
`
`between DSN and BSN — i.e., a NACK is received for a packet that the transmitter
`
`has already decided to discard along with all other packets not acknowledged and
`
`below BSN. (See id. at FIG. 10B, “discarded cells”, Ex. 1001). As shown in FIG.
`
`10B and described in the ‘62S patent, “at least one packet is not yet discarded, and
`
`at least one outstanding packet exists that has a sequence number >= BSN.” In this
`
`case, “the first outstanding packet after BSN is retransmitted with [the Receiver
`
`Packet Enforcement Bit] RPEB = TRUE.” (Id. at 8:24-30, Ex. 1001).
`
`32.
`
`In this case, the first outstanding packet after BSN is retransmitted,
`
`even though there is no indication that it was specifically not acknowledged, but it
`
`is retransmitted anyway. Also, this retransmitted packet just after BSN is within
`
`the transmit and receive window.
`
`33.
`
`The purpose of RPEB = TRUE is to move the window forward and
`
`cause the receiver to not expect any packets below the new lower end of the
`
`window (Id. at 5:28-40, 8:24-30, Ex. 1001). If the sequence number, SN, of a
`
`packet sent with RPEB = TRUE is greater than the window length above the
`
`expected SN (i.e., ESN + W), then the receiver “perform[s] one of a) a restart
`
`ActiveUS l 1654-8286v.l
`
`12
`
`
`
`and b) reporting a failure.” (Id. 11:38-48, Ex. 1001). The packet sent with the
`
`enforcement bit RPEP must therefore be within the window. Similarly, in the Go-
`
`Back—N embodiment and dependent claim 4, the receiver accepts a packet within a
`
`window N(S) < ESN+2"", and rejects if N(S) >= EsN+2"".
`
`(Id. at 10:52-61, Ex.
`
`1001).
`
`Claim Construction
`
`34.
`
`I understand that the claims in an inter parres review should be given
`
`their “broadest reasonable construction in light of the specification” as commonly
`
`understood by a person of ordinary skill in the art.
`
`The term “commanding”
`
`35.
`
`“Commanding” as used in the ‘625 patent should be given a plain and
`
`ordinary meaning of sending a command. In my opinion, in a communications
`
`system, a definition is “an instruction represented in a control field to cause an
`
`addressed device to execute a specific control function.” This definition is
`
`consistent with the claims and specification of the ‘625 patent, where the primary
`
`example of a command is an “enforcement bit” that is in a control field and is
`
`designed to cause a specific control function (move the window).
`
`36.
`
`This construction is also consistent with a definition from the IEEE
`
`dictionary from 2000, which reads “command (1) (logical link control). In data
`
`communications, an instruction represented in the control field of a protocol data
`
`ActiveUS 1 1654-8286v.l
`
`13
`
`
`
`unit (PDU) and transmitted by a logical link control (LL-C).
`
`It causes the addressed
`
`LLC(s) to execute a specific data link control function.” (IEEE Dictionary at 214-
`
`215, Ex. 1011).
`
`I understand that in the Texas Litigation, the Patent Owner took a
`
`view of the meaning of “command” that I believe is too broad. For example,
`
`Patent Owner asserted that the mere act of receiving a data packet is a command
`
`because the receiver knows to receive it:
`
`A. Right. The —— the —— the —— receiving that packet is a command
`
`to receive out of order.
`
`Q. And explain one more time, why is receiving the packet a
`
`command to receive out of order?
`
`A. Because since this particular packet
`
`is out of order and
`
`you're not allowed to not receive it, you're required to receive it,
`
`it's a command. (Trial Transcript at p. 84:19-85:20, Ex. 1014)
`
`37.
`
`This broad reading is inconsistent with the plain and ordinary meaning
`
`of the term as used in the ‘625 specification itself. For example, the ‘625 patent
`
`states that “[i]f a packet is expected which has already been discarded, then the
`
`transmitter resynchronizes by renumbering data packets or by commanding the
`
`receiver to accept an arbitrarily chosen sequence number.” (Id. at 4:39-42, Ex.
`
`1001).
`
`I interpret this to teach that a sequence number is not a command because
`
`A(:tiveUS il6548286v.l
`
`-14-
`
`
`
`the ‘625 patent uses the term “command” separately from a “sequence number” to
`
`refer to different methods that the transmitter can use to resynchronize.
`
`38.
`
`The Patent Owner’s assertions regarding the meaning of
`
`“commanding” are also inconsistent with how one of ordinary skill in the art would
`
`understand this term. The packet and its sequence number are simply data that the
`
`receiver uses; the packet is not an instruction to perform any control function.
`
`Moreover, one of ordinary skill would not view the packet as a command to
`
`receive simply because the receiver receives the packet. Instead, as explained
`
`above, a “command” in the context of the ‘625 patent is “an instruction represented
`
`in a control field to cause an addressed device to execute a specific control
`
`function.” (Id.)
`
`The phrase “commanding. . .to receive... and [to] release any expectations”
`
`39.
`
`The claimed commanding steps are about the releasing of
`
`expectations, also referred to as the “transmitter resynchroniz[ing].” (Id. at 4:37-
`
`42, Ex. 1001). Moving the window forward is the function that the RBEP
`
`enforcement bit has in the selective reject embodiments. The examples in the ‘625
`
`patent teach that the transmitter is commanding the receiver to release expectations
`
`such that the receiver releases any expectation of receiving packets with a
`
`sequence number lower than a certain number.
`
`ActiveUS l 1654-8286v.l
`
`15
`
`
`
`40.
`
`In an expert report submitted by the Patent Owner, it was suggested
`
`that “commanding ...to receive one or more packets” requires that the receiver
`
`must receive packets outside a reception window. (Ex. 1010, discussed infra.) I
`
`disagree.
`
`41.
`
`This interpretation is inconsistent with the ‘625 patent specification
`
`because the transmitter commands the receiver to “expect” a sequence number of a
`
`next packet to be sent from the transmitter, such as by “commanding the receiver
`
`to accept an arbitrarily chosen sequence number.” (‘625 patent at 4:41-42, Ex.
`
`1001). The command does not force a requirement onto the receiver that it “must”
`
`do anything other than “the receiver can be commanded to skip or overlook the
`
`packets which have been discarded.” (Id. at 5:22-23, Ex. 1001).
`
`42.
`
`In the Texas Litigation, the Patent Owner asserted that the claimed
`
`“commanding .
`
`.
`
`. to receive one or more packets” requires that the receiver must
`
`receive packets under circumstances where it otherwise would not do so. The
`
`Patent Owner argued that the asserted prior art did not disclose this limitation
`
`because the prior art “discloses a system in which the receiver may reject an out-
`
`of—sequence packet sent by the transmitter if the packet is outside of the reception
`
`window.” (See e.g., Report at 1H[ 242, 250, 285, etc., Ex. 1010).
`
`I disagree for
`
`several reasons.
`
`43.
`
`The Patent Owner’s assertions are inconsistent with Claim 1. If the
`
`ActiveUS l 1654-8286v.l
`
`16
`
`
`
`“at least one packet” of Claim 1 has an SN that is greater than the window (e.g.,
`
`SN #18, where the window ranges from SN #1 through SN #16), a receiver could
`
`move its window to #3 through #18, but then the receiver would not be “releas[ing]
`
`any expectation of receiving outstanding packets having sequence numbers prior to
`
`the at least one packet,” because that would require moving the window to #18
`
`through #2.
`
`44.
`
`The Patent Owner’s assertions are also inconsistent with the
`
`specification of the ‘625 patent. In the ‘625 patent, the transmitter commands the
`
`receiver to “expect” a sequence number of a next packet to be sent from the
`
`transmitter, such as by “commanding the receiver to accept an arbitrarily chosen
`
`sequence number.” (‘625 patent at. 4:41-42, Ex. 1001). The command does not,
`
`however, require the receiver to receive a packet it would otherwise reject. Indeed,
`
`the claimed receiver is not required to do anything other than “skip or overlook the
`
`packets which have been discarded.” (Id. at 5:22-23, Ex. 1001).
`
`45.
`
`The Patent Owner’s assertions are also inconsistent with dependent
`
`Claims 4 and 6, each of which limits packet acceptance based on the sequence
`
`number, and indicates that the receiver would reject packets received outside the
`
`window.
`
`ActiveUS l 1654-8286v.1
`
`17
`
`
`
`The Challenged Claim Is Anticipated by Garrabrant (Ex. 1002)
`
`46.
`
`I have reviewed and understand Garrabrant, U.S. Patent
`
`No. 5,610,595 (“Garrabrant”).
`
`In my opinion, a person of ordinary skill in the art
`
`would find Garrabrant to be an enabling disclosure of the subject matter it
`
`discusses.
`
`47. Garrabrant is a patent, assigned on its face to Intermec Corp., relating
`
`to a Packet Radio Communication System Protocol for mobile devices, a base
`
`radio unit (BRU), and repeaters. (See Garrabrant, Figures 5 and 6, EX. 1002).
`
`48.
`
`Garrabrant uses a repeat counter to determine how many times to
`
`repeat sending a packet. (Garrabrant, Abstract, Ex. 1002). The repeat counter is
`
`decremented each time the packet is transmitted, and when the counter reaches
`
`zero, the packet is discarded. (Id. at 8:45—-57, Ex. 1002). Garrabrant also discloses
`
`sending a “lost” message to instruct the receiver to move its valid/rej ection
`
`window forward and skip over (release expectations of receiving) the lost packets.
`
`(Id. at 10:14-42, Ex. 1002). Garrabrant discloses the subject matter of claim 1 in
`
`two different ways -— both in recovery from lost packets and in recovery from a
`
`failure state.
`
`49.
`
`The transmitter and receiver have respective windows, including a
`
`valid reception window space that is one—half the Sequence Number (SN) space;
`
`i.e., there are 16 usable SNs in a space of 32 SNs. (Garrabrant, Figures 7a and 7b;
`
`ActiveUS l16548286v.l
`
`
`
`9:9-10:13, Ex. 1002). This is a common approach that is also primarily used in the
`
`‘625 patent. (‘625 patent, 2:58-65, 6:36-50, 7:26-33, 11:44-48, Figures 7A—7C and
`
`12, Ex. 1001).
`
`50.
`
`Errors in transmission can occur, as shown in Garrabrant’s Figures 8A
`
`and 8B. A valid window ranges from SN #2 through #17, and the rejection
`
`window includes SN #18 through #1. (Garrabrant, Figure 8A, 10: 14-23, Ex.
`
`1002). In this example, “packets 2 through 6 were lost. As a consequence, the
`
`‘valid’ Window 164 did not advance further.” (Id. at 10:29-31, Ex. 1002). The
`
`“rejection window is updated in response to the receipt of a ‘lost’ message” and
`
`packet #7 (Id. at 10:23-24, 10:37-38, Ex. 1002). When packet #7 arrives, it is
`
`“accepted by the destination unit”, and then the lower sequence count is set to #8
`
`as shown in Figure 8B. The “lost” message is a command that commands the
`
`receiver that upon receipt of the next received packet (which is non-consecutive
`
`with previously received packet #1), it should move its “rejection” window
`
`forward, and not expect to receive packets #2 through #6.
`
`51.
`
`Claim 1 of the ‘625 patent recites:
`
`1.
`
`[pre] A method for discarding packets in a data network
`
`employing a packet transfer protocol including an automatic
`
`repeat request scheme, comprising the steps of:
`
`[a]
`
`a transmitter in the data network commanding a receiver
`
`in the data network to a) receive at least one packet having a
`
`ActiveUS l 1654-8286v.l
`
`1
`
`2-;
`
`
`
`sequence number that is not consecutive with a sequence
`
`number of a previously received packet and b) release any
`
`expectation of receiving outstanding packets having sequence
`
`numbers prior to the at least one packet; and
`
`[b]
`
`the transmitter discarding all packets for which
`
`acknowledgment has not been received, and which have
`
`sequence numbers prior to the at least one packet.
`
`52.
`
`To the extent the preamble is a limitation of claim 1, Garrabrant
`
`discloses “[a] method for discarding packets in a data network employing a packet
`
`transfer protocol including an automatic repeat request.” (See Garrabrant, 2:1—12,
`
`10:13-15, Ex. 1002).
`
`53.
`
`Garrabrant discloses a system with sending and receiving windows,
`
`and where the receiving unit acknowledges transmission or else the transmission is
`
`repeated if there is a time—out.
`
`(See, id. at Figures 7A, 7B, 8A, 8B, and 12; 2:1-2,
`
`and 9:9-10:41, Ex. I002).
`
`54.
`
`Garrabrant teaches “[a] a transmitter in the data network commanding
`
`a receiver in the data network to a) receive at least one packet having a sequence
`
`number that is not consecutive with a sequence number of a previously received
`
`packet and b) release any expectation of receiving outstanding packets having
`
`sequence numbers prior to the at least one packet.”
`
`ActiveUS 1 1654-8286v.l
`
`20
`
`
`
`55.
`
`Garrabrant discloses a receiver updating its window in response to the
`
`receipt of a “lost” message.
`
`(Id. at 10:21-23, Ex. 1002). Upon receipt, the receiver
`
`receives packet #7, which is non—consecutive with previously received packet #1,
`
`moves the window forward to packet #7, and is then prepared to receive packet #8.
`
`(See id. at Figures 8A and 8B; 10:14-41, Ex. 1002). In Figures 8A and 8B,
`
`“packets 2 through 6 were lost. As a consequence, the ‘valid’ Window 164 did not
`
`advance fiirther.” (Id. at 10:29-31, Ex. 1002) The “rejection window is updated in
`
`response to the receipt of a ‘lost’ message” (Id. at 10:23-24, Ex. 1002); when
`
`packet #7 arrives, it is “accepted by the destination unit” (Id. at 10:37-38, Ex.
`
`1002), and then the lower sequence count is set to #8 as shown in Figure 8B.
`
`56. While I do not believe that the command to receive requires that the
`
`receiver receive a packet it otherwise would not receive, this lost command and the
`
`sending of packet #7 causes the receiver to advance the window to receive packets
`
`that include packets it otherwise would not have been able to receive prior to the
`
`lost command and transmission of packet 7.
`
`57. Garrabrant discloses “[b] the transmitter discarding all packets for
`
`which acknowledgment has not been received, and which have sequence numbers
`
`prior to the at least one packet.” (Garrabrant, 10:23-25, Ex. 1002). Lost packets 2
`
`through 6 would be discarded by having the transmitter move its window forward.
`
`ActiveUS l 1654-8286v.l
`
`21
`
`
`
`(See, e.g., id. at 10:16, Ex. 1002). As also indicated in Figure 12 and at 12:43-
`
`13:3, once the counter reaches zero, the packet is discarded.
`
`Patent Owner’s Response
`
`58.
`
`I understand that in a litigation the Patent Owner submitted a Report
`
`(“Report”, Ex. 1010 in redacted form) in which the Patent Owner disputed that
`
`Garrabrant taught a “command” as claimed in ‘the ‘625 patent.
`
`59.
`
`I disagree. Garrabrant discloses a “command.” Garrabrant discloses a
`
`“lost” message that causes the receiver to move its window forward, and thus
`
`release expectations of receiving certain packets. (Id. at 2:62 — 3:5; 4:6-9; and 10:
`
`17-42, Ex. 1002). The Report never explains why the lost message described in
`
`Garrabrant is not a “command.”
`
`60.
`
`The Report argued that Garrabrant “repeatedly states that if a packet is
`
`received outside of the reception window, the packet will be rejected.” (Report at
`
`p. 123, Ex. 1010).
`
`I believe it is irrelevant whether or not packets with sequence
`
`numbers outside a receive window are rejected. Claim 1 does not require that a
`
`receiver receive packets outside the reception window. As noted above, the
`
`specification and claims of the ‘625 patent indicate that the ‘625 patent method
`
`also rejects packets that are outside a reception window, and thus this is no
`
`difference from the claims.
`
`ActiveUS 1 1654-8286v.l
`
`22
`
`
`
`61.
`
`The Report also asserts that Garrabrant “does not indicate whether a
`
`transmitter discards packets 2 through 6.” (Id. at p. 123, Ex. 1010). But, the
`
`transmitter moves its window forward past the “lost” packets, and this process
`
`discards the packets. Further, Figure 12 shows that when a repeat count = 0 (box
`
`214), a packet is discarded (box 218).
`
`62.
`
`The Report states that Garrabrant “fails to enable one skilled in the art
`
`to implement a discard notification scheme which anticipates the ‘625 patent.”
`
`(Report at p. 12], Ex. 1010). But, the Report provides no basis for this argument,
`
`and never explains what teaching are purportedly absent from Garrabrant. In fact,
`
`Garrabrant does enable a discard notification scheme.
`
`Garrabrant Failure Recovery
`
`63.
`
`Garrabrant also discloses failure modes and recovery that cause the
`
`windows to advance to new values, and cause packets “below” the window to be
`
`discarded at the transmitter and not to be expected at the receiver. (Garrabrant,
`
`10:55-12:4, EX. 1002).
`
`64.
`
`The mobile devices in Garrabrant can move within and outside of a
`
`coverage area. (Garrabrant, 11:1-4, 11:11-13, Ex. 1002). The mobile device and
`
`base radio unit (BRU) (base station) can each be a transmitter and a receiver. (Id.
`
`at 4:46-50; FIGS. 1-2, EX. 1002). In some instances, such as a fail state, a number
`
`of packets are lost and synchronization is lost. (See id. at 10:42-60, Ex. 1002). In
`
`ActiveUS l 1654-8286v.l
`
`23
`
`
`
`such a case, there would be packets that the transmitter sent, but that were not
`
`received when synchronization was lost. (Id. at 9:59-10:13, Ex. 1002).
`
`65.
`
`Column 11 describes a series of steps whereby the mobile sends “an
`
`SABM command” to establish communication, along with a sequence number to
`
`synchronize windows. (Garrabrant, l1:1—12:4, Ex. 1002). There can be multiple
`
`rounds of messages, described in Garrabrant as “events”; eventually, the command
`
`to establish synchronization is received and accepted. (Id. at 11:58—67, Ex. 1002).
`
`66.
`
`To the extent the preamble is a limitation of claim 1, Garrabrant
`
`discloses “[a] method for discarding packets in a data network employing a packet
`
`transfer protocol including an automatic repeat request.” (See Garrabrant, 2: 1-12,
`
`10: 13-15, Ex. 1002). Garrabrant discloses a system with sending and receiving
`
`windows, and where the receiving unit acknowledges transmission or else the
`
`transmission is repeated if there is a time—out.
`
`(See, id.at Figures 7A, 713, 8A, 8B,
`
`and 12; 2:1-2, an