`
`Guest Tek Interactive v. Nomadix, Inc.
`
`Peter Dordal, Ph.D.
`
`Page 1
`
` UNITED STATES PATENT AND TRADEMARK OFFICE
` BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
` Guest Tek Interactive Entertainment Ltd.,
`Petitioner,
`v.
`Nomadix, Inc.,
`Patent Owner.
`
`Case IPR2019-00253 Patent 8,626,922
`Case IPR2019-00211 Patent 7,953,857
`
` Deposition of DR. PETER DORDAL, taken at
` One North Wacker Drive, Chicago, Illinois,
` before Donna M. Kazaitis, IL-CSR, RPR, and
` CRR, commencing at the hour of 9:29 a.m. on
` Friday, September 13, 2019.
`
`________________________________________________
`DIGITAL EVIDENCE GROUP
`1730 M Street, NW, Suite 812
`Washington, D.C. 20036
`(202) 232-0646
`
`www.DigitalEvidenceGroup.com
`
`Digital Evidence Group C'rt 2019
`
`202-232-0646
`
`NOMADIX 2005
`Guest Tek v. Nomadix
`IPR2019-00253
`
`
`
`9/13/2019
`
`Guest Tek Interactive v. Nomadix, Inc.
`
`Peter Dordal, Ph.D.
`
`APPEARANCES:
`
`Page 2
`
`ON BEHALF OF THE PETITIONER:
` BAKER HOSTETLER LLP
` BY: JEFFREY W. LESOVITZ, ESQ.
` 2929 Arch Street
` 12th Floor
` Philadelphia, Pennsylvania 19104-2891
` 215.564.2406
` jlesovitz@bakerlaw.com
`
`ON BEHALF OF THE PATENT OWNER:
` KNOBBE, MARTENS, OLSON & BEAR, LLP
` BY: KENDALL LOEBBAKA, ESQ.
` 2040 Main Street
` 14th Floor
` Irvine, California 92614
` 949.760.0404
` kendall.loebbaka@knobbe.com
`
`1
`
`2 3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`www.DigitalEvidenceGroup.com
`
`Digital Evidence Group C'rt 2019
`
`202-232-0646
`
`NOMADIX 2005
`Guest Tek v. Nomadix
`IPR2019-00253
`
`
`
`9/13/2019
`
`Guest Tek Interactive v. Nomadix, Inc.
`
`Peter Dordal, Ph.D.
`
`Page 3
`
` INDEX
` PAGE
`DR. PETER DORDAL
` Examination by Ms. Loebbaka 4
` Examination by Mr. Lesovitz 96
`
` EXHIBITS
`NUMBER PAGE
`Exhibit 1002 Corrected Declaration of 43
` Dr. Peter Dordal
` (IPR2019-00253)
`Exhibit 1002 Corrected Declaration of 37
` Dr. Peter Dordal
` (IPR2019-00211)
`Exhibit 1004 U.S. Patent 5,864,540 50
` Bonomi et al.
` (IPR2019-00211)
`Exhibit 1005 U.S. Patent 7,392,279 40
` Chandran et al.
` (IPR2019-00211)
`Exhibit 1006 U.S. Patent 6,587,433 89
` Borella et al.
` (IPR2019-00211)
`Exhibit 1007 Index: A platform for 26
` determining how people
` value the quality of their
` internet access,
` (IPR2019-00211)
`
`1
`
`2
`
`3
`
`4 5
`
`6
`
`7
`8
`
`9
`10
`
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`
`www.DigitalEvidenceGroup.com
`
`Digital Evidence Group C'rt 2019
`
`202-232-0646
`
`NOMADIX 2005
`Guest Tek v. Nomadix
`IPR2019-00253
`
`
`
`9/13/2019
`
`Guest Tek Interactive v. Nomadix, Inc.
`
`Peter Dordal, Ph.D.
`
`Page 4
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
` DR. PETER DORDAL,
`having been first duly sworn, was examined and
`testified as follows:
` EXAMINATION
`BY MS. LOEBBAKA:
` Q. Dr. Dordal, my name is Kendall
`Loebbaka. Thank you for coming in today.
` Have you been deposed before?
` A. No.
` Q. No, you have not, okay. So I'd like
`to go over a couple of ground rules since you
`haven't been deposed before. That way we make it
`as easy and simple as possible.
` So the court reporter is going to
`be recording everything that you and I say today
`and your counsel if he has objections. So try
`best not to talk over each other. Wait until I
`finish my question before you answer. That way
`she can record the questions and answers.
` You need to give verbal answers.
`So a "yes" or a "no," not nodding your head. That
`way she can record that.
`
`www.DigitalEvidenceGroup.com
`
`Digital Evidence Group C'rt 2019
`
`202-232-0646
`
`NOMADIX 2005
`Guest Tek v. Nomadix
`IPR2019-00253
`
`
`
`9/13/2019
`
`Guest Tek Interactive v. Nomadix, Inc.
`
`Peter Dordal, Ph.D.
`
`Page 5
`Your counsel may object from time
`to time. You still need to answer the question
`when he objects, unless he instructs you not to.
`Have you consumed any substance,
`drugs or alcohol, that would prevent you from
`giving your truthful testimony today?
`A.
`No.
`Q.
`And if at any time any of my questions
`are unclear, please ask me to clarify the
`question.
`
`Do you understand that the
`deposition today is to cover two IPRs for which
`you have given declarations?
`A.
`Yes.
`Q.
`And so in those two IPRs, do you
`understand that there are some shared exhibits
`between the IPRs.
`A.
`Yes. These are the '857 and '922
`Patents.
`Q.
`Patents.
`
`Yes, correct. Nomadix's '857 and '922
`
`So, for instance, there are some
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`www.DigitalEvidenceGroup.com
`
`Digital Evidence Group C'rt 2019
`
`202-232-0646
`
`NOMADIX 2005
`Guest Tek v. Nomadix
`IPR2019-00253
`
`
`
`9/13/2019
`
`Guest Tek Interactive v. Nomadix, Inc.
`
`Peter Dordal, Ph.D.
`
`Page 6
`exhibits for prior art that you have relied upon
`that are shared between the two IPRs, and I will
`introduce with agreement upon counsel just one of
`those exhibits to make it simpler for you.
`Dr. Dordal, could you give me an
`overview of your educational background after high
`school?
`I went to the University of Chicago as
`A.
`an undergraduate. I went to Harvard University
`and received a Ph.D. in mathematics in the field
`of mathematical logic.
`Q.
`And what did you do after you received
`your degrees?
`A.
`In the course of receiving my Ph.D., I
`learned a certain amount about programming. And
`when I applied for my first job at Loyola
`University, they needed someone to teach
`programming. So I got increasingly drawn into
`computer science.
`Q.
`And have you been a professor at
`Loyola ever since you graduated from Harvard?
`A.
`Since 1982, yes.
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`www.DigitalEvidenceGroup.com
`
`Digital Evidence Group C'rt 2019
`
`202-232-0646
`
`NOMADIX 2005
`Guest Tek v. Nomadix
`IPR2019-00253
`
`
`
`9/13/2019
`
`Guest Tek Interactive v. Nomadix, Inc.
`
`Peter Dordal, Ph.D.
`
`Page 7
` Q. What are the courses? Can you give me
`an overview of the courses that you teach at
`Loyola?
` A. I teach two sections of computer
`networks each year. I usually teach a course
`in -- I'm trying to recall the precise title.
`"Social, Ethical, and Legal Issues in Computing."
` I teach introductory courses of
`introductory programming at different levels. I
`teach a certain number of database courses. And
`that would be I think almost everything. There's
`isolated special cases.
` Q. And do any of those courses -- or, do
`you have experience with network bandwidth
`management?
` A. Yes.
` Q. Can you give me an overview of what
`that experience is?
` A. Normally in a network, for example, in
`the TCP/IP network, there's no built-in bandwidth
`management. So typically some router along the
`path is called into service to support some method
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`www.DigitalEvidenceGroup.com
`
`Digital Evidence Group C'rt 2019
`
`202-232-0646
`
`NOMADIX 2005
`Guest Tek v. Nomadix
`IPR2019-00253
`
`
`
`9/13/2019
`
`Guest Tek Interactive v. Nomadix, Inc.
`
`Peter Dordal, Ph.D.
`
`Page 8
`of managing traffic typically by verifying whether
`the traffic flow is in accord with a pre-arranged
`traffic contract so that users cannot use more
`bandwidth than they have paid for in effect.
` Most home ISPs impose a maximum
`bandwidth and then home subscribers can pay more
`for a higher service level. That happens for
`companies too at usually a somewhat higher level.
` Q. And what has -- you've kind of given
`me an overview of bandwidth management. What's
`your work experience with respect to bandwidth
`management?
` A. I used to run the department network
`up until -- the network was first installed in
`1986 and then through 2003. But during that
`period I never actually installed bandwidth
`management software. That was seldom done. But
`I've implemented experimental and demonstration
`systems.
` Q. Can you give me some examples of
`experimental and demonstration systems with
`bandwidth management?
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`www.DigitalEvidenceGroup.com
`
`Digital Evidence Group C'rt 2019
`
`202-232-0646
`
`NOMADIX 2005
`Guest Tek v. Nomadix
`IPR2019-00253
`
`
`
`9/13/2019
`
`Guest Tek Interactive v. Nomadix, Inc.
`
`Peter Dordal, Ph.D.
`
`Page 9
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
` A. I currently have a thesis student
`who's finishing up his thesis, and our work
`together involved using the mini-net network
`emulation test bed to implement a mechanism by
`which various TCP connections could have their
`flows, their bandwidths, either measured or
`constrained.
` Q. Okay. I understand that you submitted
`a corrected declaration in both of the IPRs that
`we're talking about today, the '857 Patent IPR and
`the '922 Patent IPR. Does that sound familiar?
` A. Yes.
` Q. What did you correct in those two
`declarations?
` A. The only change was an addition of a
`statement that everything was true and that I
`understood under subject of penalties of perjury.
`That's all I can recall in that wording. But if I
`were to look at the declaration, I would be able
`to read it back to you.
` Q. Okay. So it was just that last
`paragraph --
`
`www.DigitalEvidenceGroup.com
`
`Digital Evidence Group C'rt 2019
`
`202-232-0646
`
`NOMADIX 2005
`Guest Tek v. Nomadix
`IPR2019-00253
`
`
`
`9/13/2019
`
`Guest Tek Interactive v. Nomadix, Inc.
`
`Peter Dordal, Ph.D.
`
`Page 10
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
` A. That was the only change.
` Q. There was no substance change to the
`declaration?
` A. No.
` Q. Okay. Are you familiar with ATM
`networks?
` A. Yes.
` Q. What does "ATM" stand for?
` A. Asynchronous transfer mode.
` Q. Can you describe what an ATM network
`is for me?
` A. Essentially all networks involve
`transmission of so-called packets, small buffers
`of data. For ethernet, "small" means up to 1500
`bytes. And these packets contain a header that
`contains the information as to where the packet is
`to be delivered. For ATM, packets are rather
`small, 53 bytes total: 5 bytes of header and 48
`bytes of data.
` ATM was originally developed as a
`one-size-fits-all approach that would handle both
`data networking and voice, telephones.
`
`www.DigitalEvidenceGroup.com
`
`Digital Evidence Group C'rt 2019
`
`202-232-0646
`
`NOMADIX 2005
`Guest Tek v. Nomadix
`IPR2019-00253
`
`
`
`9/13/2019
`
`Guest Tek Interactive v. Nomadix, Inc.
`
`Peter Dordal, Ph.D.
`
`Page 11
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
` ATM uses the concept of virtual
`circuits where as part of setting up an ATM
`connection all the routers or switches -- I'll use
`the term interchangeably -- along the path of the
`connection negotiate among themselves, first to
`determine that path and then to determine that
`that connection is allowed to transmit.
` So then the packets are -- the
`address is the form of a circuit identifier. So
`that is recognized then by all the
`switches/routers in the path.
` Q. All right. Can you clarify what do
`you mean by a circuit identifier?
` A. It's an address field that indicates
`to which circuit this particular packet belongs.
`It indicates that to the routers involved and also
`to the end system.
` The routers may, and almost always
`do, reassign the circuit identifier in accordance
`with their own rules.
` So, for example, a packet coming in
`as part of a circuit with identifier 57 from the
`
`www.DigitalEvidenceGroup.com
`
`Digital Evidence Group C'rt 2019
`
`202-232-0646
`
`NOMADIX 2005
`Guest Tek v. Nomadix
`IPR2019-00253
`
`
`
`9/13/2019
`
`Guest Tek Interactive v. Nomadix, Inc.
`
`Peter Dordal, Ph.D.
`
`Page 12
`left, that identifier is updated to 118 as it goes
`out on the right in accordance with a precise
`understanding by all switches of the particular
`circuit identifier.
` The circuit identifier is unique
`only on each link, in other words, and it is
`updated as the packet moves from link to link.
`But at any given point it uniquely identifies the
`entire path that that packet has been assigned,
`that that connection has been assigned.
` Q. I was kind of going to ask you about
`that. So for an ATM network, is it assigned a
`particular path for each connection?
` A. Yes, yes.
` Q. And for that connection, do all of the
`packets travel that connection?
` A. Yes, for the duration of that
`connection.
` Q. When you're describing the circuit
`identifier, you threw out some numbers, 57 and
`118. I'm not quite sure what you were meaning by
`that.
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`www.DigitalEvidenceGroup.com
`
`Digital Evidence Group C'rt 2019
`
`202-232-0646
`
`NOMADIX 2005
`Guest Tek v. Nomadix
`IPR2019-00253
`
`
`
`9/13/2019
`
`Guest Tek Interactive v. Nomadix, Inc.
`
`Peter Dordal, Ph.D.
`
`Page 13
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
` A. ATM has some complexities, but the
`straight forward concept of the virtual circuit
`identifier is that it is a relatively small
`number. In ATM it varies from 16 bits to a total
`of I think 28 bits are available. So those
`numbers are in effect arbitrary. They are not
`like IP addresses that are permanently assigned to
`individual nodes.
` Q. Okay.
` A. And in that sense an ATM connection is
`maybe a little more akin to a telephone connection
`as opposed to a TCP/IP connection that a browser
`might manage to download a remote website.
` Q. Can you kind of elaborate? What do
`you mean by it's akin to a telephone connection?
` A. In that the connection has to be
`negotiated on startup and all the intermediate
`switches are involved in the decision as to
`whether the connection is allowed.
` And in ATM they must, because each
`pair of adjacent switches negotiates what
`identifier is to be used for that connection on
`
`www.DigitalEvidenceGroup.com
`
`Digital Evidence Group C'rt 2019
`
`202-232-0646
`
`NOMADIX 2005
`Guest Tek v. Nomadix
`IPR2019-00253
`
`
`
`9/13/2019
`
`Guest Tek Interactive v. Nomadix, Inc.
`
`Peter Dordal, Ph.D.
`
`Page 14
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`that particular link.
` Q. Okay. Earlier you mentioned that ATM
`has some complexities. Can you elaborate on some
`of those complexities?
` A. There are different options. There's
`a virtual channel identifier and something called
`a VPI -- which I don't remember. It's virtual
`something identifier. Different modes of usage
`can mean that different numbers of those bits are
`used for addressing.
` For example, within a site you
`would almost never need more than 12 bits for the
`address. 16 is plenty. For intercontinental
`traffic, the full address would be needed.
` Q. Do you have any other ideas that come
`to mind about other complexities for an ATM
`network?
` MR. LESOVITZ: Objection to form.
` THE WITNESS: I would not recall any
`at the moment.
`BY MS. LOEBBAKA:
` Q. Okay.
`
`www.DigitalEvidenceGroup.com
`
`Digital Evidence Group C'rt 2019
`
`202-232-0646
`
`NOMADIX 2005
`Guest Tek v. Nomadix
`IPR2019-00253
`
`
`
`9/13/2019
`
`Guest Tek Interactive v. Nomadix, Inc.
`
`Peter Dordal, Ph.D.
`
`Page 15
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
` A. I mean ATM -- the full ATM
`specification is enormously detailed.
` Q. Are you familiar, I assume, with an IP
`network?
` A. Yes.
` Q. Could you give me an overview of what
`an IP network is?
` A. The letters "IP" come from
`inter-network protocol. The idea is that every
`device is assigned an IP address and that routers
`are devices that have multiple connections and
`each router contains a table, a forwarding table.
`And when a packet arrives, the router examines the
`destination IP address and figures out the next
`hop. That is, the neighboring router or final
`destination to which the packet is forwarded.
` That's known as datagram
`forwarding. In distinction to virtual circuits
`commonly used by ATM, datagram forwarding does not
`require that routers along the path negotiate the
`connection before the connection can actually
`begin.
`
`www.DigitalEvidenceGroup.com
`
`Digital Evidence Group C'rt 2019
`
`202-232-0646
`
`NOMADIX 2005
`Guest Tek v. Nomadix
`IPR2019-00253
`
`
`
`9/13/2019
`
`Guest Tek Interactive v. Nomadix, Inc.
`
`Peter Dordal, Ph.D.
`
`Page 16
` Q. Can you elaborate on that? What does
`that result in? How is that different?
` MR. LESOVITZ: Objection to form.
` THE WITNESS: It's not all that
`different. One advantage of the ATM approach is
`that since all routers participate in the decision
`as to whether to allow a certain connection, they
`can also make service guarantees as to, say,
`bandwidth or, somewhat less commonly, delay.
` There have been -- there was a
`major initiative, the so-called integrative
`services suite of protocols, that added virtual
`circuits to the top of IP to achieve that same end
`of service guarantees, quality of service
`guarantees.
` Integrative services is seldom
`used. But certainly the general idea of making IP
`look a little bit more like a virtual circuit
`system was successful in principle.
` I'm not sure if I answered your
`original question.
` Q. Let me go back.
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`www.DigitalEvidenceGroup.com
`
`Digital Evidence Group C'rt 2019
`
`202-232-0646
`
`NOMADIX 2005
`Guest Tek v. Nomadix
`IPR2019-00253
`
`
`
`9/13/2019
`
`Guest Tek Interactive v. Nomadix, Inc.
`
`Peter Dordal, Ph.D.
`
`Page 17
` So you mentioned that IP networks,
`they added a virtual circuit system to the IP
`networks?
` A. That was the specific integrative
`services set of protocols. You may regard it as
`an experimental protocol because it did not
`receive widespread adoption.
` Q. Okay. So in general it's not widely
`used today?
` A. No.
` MR. LESOVITZ: Objection to form.
` THE WITNESS: That being integrative
`services is not widely used today, neither is ATM.
`BY MS. LOEBBAKA:
` Q. Would you consider the IP network a
`connectionless network?
` A. That word is used in the IP world to
`describe this fundamental idea that each router
`has this table of instructing it what to do for
`each destination, what is the next hop, what is
`the neighboring router to which the packet should
`be forwarded. Routers generally are not aware of
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`www.DigitalEvidenceGroup.com
`
`Digital Evidence Group C'rt 2019
`
`202-232-0646
`
`NOMADIX 2005
`Guest Tek v. Nomadix
`IPR2019-00253
`
`
`
`9/13/2019
`
`Guest Tek Interactive v. Nomadix, Inc.
`
`Peter Dordal, Ph.D.
`
`Page 18
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`any particular connection.
` So if I am downloading a large
`file, there might be a thousand packets passing
`through as part of that file transfer. One
`router -- and each packet has a destination looked
`up independently, and the router has no large
`scale view, has no view at all, of what number of
`connections are passing through.
` Some routers in the interests of
`bandwidth management, among other things, do try
`to record and track some of that information. But
`the basic principle of a router is that a packet
`comes in, its destination address is looked up in
`this table, and the packet is then forwarded, and
`the next single hop, next single step of the path,
`in accordance with that lookup.
` Q. So for the example that you gave of
`downloading a document, a thousand packets, is it
`possible that those packets go different routes to
`arrive at the destination?
` A. It is possible. It very seldom occurs
`in practice that a route changes on any kind of
`
`www.DigitalEvidenceGroup.com
`
`Digital Evidence Group C'rt 2019
`
`202-232-0646
`
`NOMADIX 2005
`Guest Tek v. Nomadix
`IPR2019-00253
`
`
`
`9/13/2019
`
`Guest Tek Interactive v. Nomadix, Inc.
`
`Peter Dordal, Ph.D.
`
`Page 19
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`short-term basis.
` An early theoretical advantage of
`this forwarding approach was that if a router does
`fail, traffic can be rerouted around it without
`involving any other than the immediately adjacent
`nodes.
` Whereas, it was long known that for
`circuit and virtual circuit forwarding mechanisms,
`such as used by ATM, the entire path would have to
`be renegotiated. And in practice the circuit
`would fail and then would have to be renegotiated
`from scratch.
` Q. Okay. You mentioned earlier that
`ethernet is I believe 1500 bits; is that correct?
` A. Bytes.
` Q. Bytes. Sorry.
` A. Maximum packet size.
` Q. Okay.
` A. And it might be 1502. There's some
`strangeness in the early ethernet specifications.
`There's a couple different versions.
` Q. And that is referring to the IP
`
`www.DigitalEvidenceGroup.com
`
`Digital Evidence Group C'rt 2019
`
`202-232-0646
`
`NOMADIX 2005
`Guest Tek v. Nomadix
`IPR2019-00253
`
`
`
`9/13/2019
`
`Guest Tek Interactive v. Nomadix, Inc.
`
`Peter Dordal, Ph.D.
`
`Page 20
`network that we were talking about just a minute
`ago?
` MR. LESOVITZ: Objection to form.
` THE WITNESS: Ethernet is commonly
`used with IP networks.
` In here Wi-Fi is used, which is --
`except it's transmitted over radio waves instead
`of wires, very similar to ethernet, and Wi-Fi
`packets can inter-operate with ethernet. They can
`travel partway through the radio waves and partway
`through the ethernet cables.
`BY MS. LOEBBAKA:
` Q. How many bytes are in an IP packet?
` A. The absolute minimum is the size of
`the IP header, which is 20. Normally there is a
`local area network header. An ethernet header is
`14 bytes. The IP header can be more than 20
`bytes. There's space for some options. And then
`if TCP is used to manage a specific connection,
`then that's another 20 bytes.
` So that brings us to 54 bytes.
`There is an official rule for ethernet that the
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`www.DigitalEvidenceGroup.com
`
`Digital Evidence Group C'rt 2019
`
`202-232-0646
`
`NOMADIX 2005
`Guest Tek v. Nomadix
`IPR2019-00253
`
`
`
`9/13/2019
`
`Guest Tek Interactive v. Nomadix, Inc.
`
`Peter Dordal, Ph.D.
`
`Page 21
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`minimum packet size must be at least a certain
`amount. The original ethernet, it was 64 bytes.
`Other ethernets have chosen somewhat different
`values. But there is sometimes some need to pad
`out the barebones IP packet.
` The most common barebones packets
`are TCP acknowledgments which contain nothing but
`the IP header and TCP header and the possible LAN
`header.
` Q. So from your explanation I understand
`that IP packets can very widely on size.
` A. They do --
` MR. LESOVITZ: Objection to form.
` I'm sorry. Is there a question?
`BY MS. LOEBBAKA:
` Q. Do IP packets vary in size?
` A. Yes.
` Q. Can you explain to me how they vary in
`size?
` A. They contain the amount of data that
`the sender chose to include. And the sender may
`be the sending application or it may be the TCP
`
`www.DigitalEvidenceGroup.com
`
`Digital Evidence Group C'rt 2019
`
`202-232-0646
`
`NOMADIX 2005
`Guest Tek v. Nomadix
`IPR2019-00253
`
`
`
`9/13/2019
`
`Guest Tek Interactive v. Nomadix, Inc.
`
`Peter Dordal, Ph.D.
`
`Page 22
`layer which generally makes its own decisions as
`to how packets, how data is to be packetized, how
`data is to be divided into packets.
` As a general rule, for long data
`transfers larger packet size is preferred. There
`are some drawbacks to going above 1500 bytes in
`some experiments in that direction but they've
`generally not met with widespread adoption.
`Whereas, interactive traffic or, say, voice
`telephony traffic tends to involve smaller
`packets.
` Then there are the acknowledgment
`packets to confirm receipt of data. Those are
`probably the smallest of all.
` Q. What is the packet size for ATM
`networks?
` A. 53 bytes.
` Q. And is that a fixed packet size?
` A. ATM specification spells out a fixed
`packet size, yes.
` Q. Are you familiar with an IP stack or
`IP layer?
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`www.DigitalEvidenceGroup.com
`
`Digital Evidence Group C'rt 2019
`
`202-232-0646
`
`NOMADIX 2005
`Guest Tek v. Nomadix
`IPR2019-00253
`
`
`
`9/13/2019
`
`Guest Tek Interactive v. Nomadix, Inc.
`
`Peter Dordal, Ph.D.
`
`Page 23
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
` A. Yes.
` Q. What is an IP stack or IP -- do you
`have a preference between the two?
` A. The stack refers somewhat informally
`to the software that the IP layer -- that is
`necessary to implement the IP layer.
` The IP layer is a conceptual layer
`in the system software that -- well, it attaches,
`for outgoing traffic, it attaches the IP header,
`makes sure that the destination IP address is
`appropriate, attaches the sender, the appropriate
`sender or source IP address, and determines where
`to send the packet as its own next hop.
` For machines in this room, assuming
`a packet is being addressed outside of the Baker
`Hostetler network, that would mean sending a
`packet to a local router, which somewhere is in
`the building but we don't know where.
` For incoming traffic for an
`endpoint, end user computer, the IP layer
`determines that the destination address is one
`that's acceptable for this computer and then
`
`www.DigitalEvidenceGroup.com
`
`Digital Evidence Group C'rt 2019
`
`202-232-0646
`
`NOMADIX 2005
`Guest Tek v. Nomadix
`IPR2019-00253
`
`
`
`9/13/2019
`
`Guest Tek Interactive v. Nomadix, Inc.
`
`Peter Dordal, Ph.D.
`
`Page 24
`delivers the packet up to the application, often
`through the TCP layer, which is reassembling
`packets.
` Probably the most important point
`for the IP layer is in routers where packets are
`forwarded on their next hop, according to this
`table of destinations and next hops, corresponding
`next hops. And that's done at each step of the
`way which typically for cross-continental traffic
`10 to 20 steps.
` Q. Okay. How many layers are there in an
`IP packet?
` A. In the packet, layers apply to the
`system software, not so much to the packet.
`Packets typically have a LAN header, an IP header,
`and almost always a transport header.
`Corresponding to the LAN layer, the IP layer and
`the transport layer.
` There is frequent separation within
`the LAN layer into the logical layer and physical
`layer. For Wi-Fi, for example, the physical layer
`requires that radio waves be used to transmit the
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`www.DigitalEvidenceGroup.com
`
`Digital Evidence Group C'rt 2019
`
`202-232-0646
`
`NOMADIX 2005
`Guest Tek v. Nomadix
`IPR2019-00253
`
`
`
`9/13/2019
`
`Guest Tek Interactive v. Nomadix, Inc.
`
`Peter Dordal, Ph.D.
`
`Page 25
`packet instead of wires. But the logical layer,
`the layout of bits and bytes of the packet is,
`well, not exactly the same as ethernet but quite
`similar.
` Q. Okay. What information is contained
`within the ATM header?
` A. Their primary piece of information is
`what I will refer to as the virtual circuit
`identifier. There's some ATM specific terms like
`virtual channel identifier that are related to
`that, that identifies the circuit that this packet
`is traveling on.
` The circuits are these logical
`constructs that are created by negotiation among
`all the routers, switches, along the path, and
`then each packet must be assigned one of those
`circuits to be sent at all.
` Other information in the ATM header
`includes an 8-bit checksum for error detection and
`some miscellaneous other management bits. Some of
`those bits are sometimes used for running the IP
`protocol on top of ATM.
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`www.DigitalEvidenceGroup.com
`
`Digital Evidence Group C'rt 2019
`
`202-232-0646
`
`NOMADIX 2005
`Guest Tek v. Nomadix
`IPR2019-00253
`
`
`
`9/13/2019
`
`Guest Tek Interactive v. Nomadix, Inc.
`
`Peter Dordal, Ph.D.
`
`Page 26
` The IP packet would then be divided
`into small chunks. And some bits in the ATM
`header might signal, for example, when that
`particular IP packet was starting and ending.
` Q. Now, earlier you mentioned that ATM
`networks are not used as frequently anymore. Do
`you have an understanding about why that is?
` A. Marketplace issues primarily. The
`telecommunications industry developed ATM in the
`1980s, which is about roughly the same time that
`IP networking was being developed. I don't think
`there's any compelling technical reason to choose
`one or the other, but IP networking won out in the
`marketplace.
` Q. Okay. Dr. Dordal, I'm going to hand
`you what's marked as Exhibit 1007. This is
`Exhibit 1007 in both IPRs. So the IPR for the
`'857 Patent and the '922 Patent. (Document
`tendered to the witness.)
` A. Okay.
` Q. So I refer to this reference as Rupp,
`but I'd like to refer to it as how you know this
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`www.DigitalEvidenceGroup.com
`
`Digital Evidence Group C'rt 2019
`
`202-232-0646
`
`NOMADIX 2005
`Guest Tek v. Nomadix
`IPR2019-00253
`
`
`
`9/13/2019
`
`Guest Tek Interactive v. Nomadix, Inc.
`
`Peter Dordal, Ph.D.
`
`Page 27
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`reference. Is there a name that --
` A. Rupp is fine.
` Q. Rupp, okay.
` If you could turn to Page 3 of Rupp
`that has Figure 1. Could you please explain to me
`how the Rupp system works with respect to Figure
`1?
` A. The primary network goal for
`established connections in the Rupp system was to
`limit users to a pre-selected bandwidth level. In
`fact, Figure 2 shows the interface by which users
`could choose their bandwidth level.
` So two things would happen. One is
`that users would be billed according to what their
`bandwidth choice was. The higher the bandwidth,
`the higher the permitted cost.
` And, two, the hardware involved --
`and the paper doesn't spell out specifically which
`hardware does what. But the hardware involved
`would enforce these bandwidth constraints, so that
`a user who signed up for 32 kbits per second would
`genuinely be limited to that. In attempting to
`
`www.DigitalEvidenceGroup.com
`
`Digital Evidence Group C'rt 2019
`
`202-232-0646
`
`NOMADIX 2005
`Guest Tek v. Nomadix
`IPR2019-00253
`
`
`
`9/13/2019
`
`Guest Tek Interactive v. Nomadix, Inc.
`
`Peter Dordal, Ph.D.
`
`Page 28
`download a very large file, they would not receive
`more bandwidth than that 32 kilobits per second
`limit.
` Q. So if a user in Figure 1 is trying to
`download a file and is sending that request, how
`does that request go through the system?
` A. The request to download a file?
` Q. Yes. I'm wondering how the packets
`travel through.
` A. Okay. The --
` MR. LESOVITZ: I'm sorry, Dr. Dordal,
`you have to wait for me to get in my objections.
`Objection to form.
` THE WITNESS: So I will imagine that
`the user is using a browser and is clicking on a
`file to be downloaded.
` So that travels -- the
`connection -- the request travels through the
`network to the internet, the cloud at the lower
`right, and a TCP connection, a transmission
`control protocol, is set up exactly as would be
`the case without an index scheme involved, and the
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`www.DigitalEvidenceGroup.com
`
`Digital Evidence Group C'rt 2019
`
`202-232-0646
`
`NOMADIX 2005
`Guest Tek v. Nomadix
`IPR2019-00253
`
`
`
`9/13/2019
`
`Guest Tek Interactive v. Nomadix, Inc.
`
`Peter Dordal