`SOUTHERN DISTRICT OF CALIFORNIA
`BEFORE HONORABLE CATHY ANN BENCIVENGO,
`JUDGE PRESIDING
`
`BELL NORTHERN RESEARCH, LLC,,
`
`Plaintiff,
`
`CASE NO. 18CV1783-CAB-BLM
`
`vs.
`
`INC. AND
`COOLPAD TECHNOLOGIES,
`YULONG COMPUTER COMMUNICATIONS,
`
`SAN DIEGO, CALIFORNIA
`
`Defendants.
`
`THURSDAY,
`
`JUNE 20, 2019
`
`BELL NORTHERN RESEARCH, LLC,
`
`Plaintiff,
`
`CASE NO.
`
`18CV1784-CAB-BLM
`
`vs.
`
`HUAWEI TECHNOLOGIES Co., LTD.,
`HUAWEI DEVICE (HONG KONG) CO.,
`LTD., and HUAWEI DEVICE USA,
`
`Defendants.
`
`+ N
`
`h w a a
`
`o
`
`oY
`
`~J
`
`ao
`
`Oo
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`meeeaee
`
`BELL NORTHERN RESEARCH, LLC.,
`
`Plaintiff,
`
`vs.
`
`KYOCERA CORPORATION and KYOCERA
`
`INTERNATIONAL INC.,
`
`Defendants.
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`meeeeee
`
`CASE NO.
`
`18CV1785-CAB-BLM
`
`OnePlus Ex. 1020.0001
`IPR2022-00048
`
`OnePlus Ex. 1020.0001
`IPR2022-00048
`
`
`
`BELL NORTHERN RESEARCH, LLC.,
`
`vs.
`
`Plaintiff,
`
`CASE NO. 18CV1786-CAB-BLM
`
`ZTE CORPORATION, ZTE (USA)
`ZTE (TX)
`INC.
`
`INC.
`
`BELL NORTHERN RESEARCH, LLC,,
`
`Defendants.
`
`—
`
`Plaintiff,
`
`CASE NO.
`
`18CV2864—-CAB-BILM
`
`VS.
`
`INC., LG
`LG ELECTRONICS,
`ELECTRONICS U.S.A.
`INC., and
`LG ELECTRONICS MOBILE RESEARCH
`U.S.A., LLC,
`
`Defendants.
`
`
`REPORTER'S TRANSCRIPT OF PROCEEDINGS
`CLAIMS CONSTRUCTION HEARING
`DAY TWO, VOLUME TWO, PAGES 1-122
`
`Proceedings reported by stenography, transcript produced by
`computer assisted software
`
`Mauralee Ramirez, RPR, CSR No. 11674
`Federal Official Court Reporter
`ordertranscript@gmail.com
`
`10
`
`11
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`23
`
`24
`
`Zo)
`
`OnePlus Ex. 1020.0002
`IPR2022-00048
`
`OnePlus Ex. 1020.0002
`IPR2022-00048
`
`
`
`COUNSEL APPEARING:
`For The Plaintiff:
`
`Sadaf Raja Abdullah, Esq.
`Steven W. Hartsell, Esq.
`Paul J. Skiermont, Esq.
`SKIERMONT DERBY LLP
`Thanksgiving Tower
`1601 Elm Street, Suite 4400
`Dallas, Texas 75201
`
`For The Defendants
`Coolpad and Yulong:
`
`Thomas Nathan Millikan, Esq.
`James Young Hurt, Esq.
`PERKINS COIE, LLP
`11988 El Camino Real, Suite 350
`San Diego, California 92130
`
`for the Defendants
`Huawei entities:
`
`For The Defendants
`ZTE entities:
`
`ALSO PRESENT:
`
`Joanna M. Fuller, Esq.
`Jason W. Wolff, Esq.
`FISH & RICHARDSON P.C.
`12390 El Camino Real
`San Diego, California 92130
`
`Ethan J. Rubin, Esq.
`FISH & RICHARDSON, P.C.
`One Marina Park Drive
`Boston, MA 02210
`
`Jiaxiao Zhang, Esq.
`McDERMOTT WILL & EMERY LLP
`18565 Jamboree Road, Suite 250
`Irvine, California 92612
`
`Amol Ajay Parikh, Esq.
`Thomas DaMario, Esq.
`McDERMOTT WILL & BMERY LLP
`444 West Lake Street, Suite 4000
`Chicago, Illinois 60606
`
`10
`
`11
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`23
`
`24
`
`25
`
`OnePlus Ex. 1020.0003
`IPR2022-00048
`
`OnePlus Ex. 1020.0003
`IPR2022-00048
`
`
`
`San Diego, California; Thursday, June 20, 2019; 9:00 a.m.
`
`(Cases called)
`
`MS. ABDULLAH:
`
`Sadaf Abdullah on behalf of plaintiff,
`
`Bell Northern Research.
`
`MR. HARTSELL:
`
`Steven Hartsell on behalf of Bell
`
`Northern Research.
`
`THE COURT:
`
`Thank you.
`
`MR. SKIERMONT: Good morning your Honor.
`
`Paul
`
`Skiermont on behalf of Bell Northern Research.
`
`MS. ZHANG: Good morning, your Honor.
`
`Jiaxiao Zhang
`
`from McDermott Will & Emery on behalf of ZTE. With me is Amol
`
`Parikh and Thomas DaMario.
`
`MS. FULLER: Good morning.
`
`Joanna Fuller on behalf of
`
`Huawei with Fish & Richardson, and with me is Jason Wolff and
`
`Ethan Rubin.
`
`MR. MILLIKEN: Good morning, your Honor.
`
`Tom Milliken
`
`from Perkins Coie on behalf of Coolpad and Yulong. With me is
`
`James Hurt.
`
`THE COURT:
`
`Thank you. All right. We're back.
`
`So
`
`let's get started on the '842.
`
`MR. HARTSELL: Your Honor, may I approach?
`
`THE COURT: Yes.
`
`Go ahead.
`
`MR. HARTSELL: Good morning, your Honor. Again this
`
`is Steven Hartsell on behalf of Bell Northern Research.
`
`The
`
`'g42 patent was developed by engineers at Broadcom and filed in
`
`te
`
`SO
`
`ww
`
`aw
`
`on
`
`n ~
`
`l
`
`ow
`
`Ne}
`
`10
`
`11
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`24
`
`25
`
`OnePlus Ex. 1020.0004
`IPR2022-00048
`
`OnePlus Ex. 1020.0004
`IPR2022-00048
`
`
`
`January of 2010.
`
`The '842 patent is a continuation of U.S.
`
`Patent Number 7,646,703 which claims priority to at least
`
`July 2004.
`
`The '842 patent is directed to long training
`
`sequences with minimum peak-to-average power ratios, and today
`
`I would like to provide some background and a few common steps
`
`that I hope the Court would find useful in today's discussion.
`
`The '842 patent is taught against the backdrop of the
`
`802.11 WiFi standard which is promulgated by IEEE, which is the
`
`Institute for Electrical and Electronic Engineers. This
`
`standard governs how different wireless devices are designed
`
`and how they communicate with one another.
`
`Now as technology
`
`evolves,
`
`the 802.11 standard has been amended periodically to
`
`add additional capabilities, usually resulting in faster speeds
`
`and better coverage.
`
`As you can see on our slide,
`
`in 1999,
`
`the 802.11
`
`standard was amended to implement OFDM, which stands for
`
`orthogonal frequency-division multiplexing,
`
`to increase data
`
`throughput.
`
`I'm going to show you what that means on slide 5.
`
`At the top, you can see this is how data was transmitted OFDM.
`
`Basically we have single carriers that are separated. When
`
`OFDM is implemented,
`
`the carrier waves are essentially smushed
`
`together allowing you to send more data found within the given
`
`bandwidth. As you can see on the OFDM,
`
`there's an overlap in
`
`the subcarriers which is necessary to achieve high data rates.
`
`In slide 6, each colored peak is a subcarrier which
`
`10
`
`11
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`23
`
`24
`
`25
`
`OnePlus Ex. 1020.0005
`IPR2022-00048
`
`OnePlus Ex. 1020.0005
`IPR2022-00048
`
`
`
`carries data essentially, for example,
`
`the data you Might need
`
`to load your website.
`
`The carriers are designed to be
`
`orthogonal which allows them to occupy the same bandwidth
`
`without interfering with which other.
`
`Now as with many things while OFDM provides throughput
`
`improvements and other advantages, it also brings certain
`
`disadvantages. And one of the disadvantages to using OFDM
`
`systems is they are known to have high peak-to0average power
`ratio,
`in other words, PAPR, when compared to single carrier
`
`systems.
`
`PAPR is the ratio of peak power to the average power
`
`signal.
`
`Now due to the presence of large numbers of
`
`independently modulated subcarriers in an OFDM system,
`
`the peak
`
`value of a system can be very high as compared to the average
`
`of the system as a whole. This is a problem -- PAPR is a
`
`problem because it reduces the power efficiency of radio
`frequency amplifiers, and this results essentially in high
`
`power consumption battery drain.
`
`Therefore,
`
`the RF amplifiers are operated usually with
`
`a certain safety margin called a power back-off.
`
`Increasing
`
`the power back-off can result in lower amplifier efficiency and
`
`higher overall power consumption.
`
`Another concept that may come up today is BPSK.
`
`BPSK
`
`stands for binary phase shift keying which is a digital
`
`modulation process by changing or modulating a phase of a
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`23
`
`24
`
`Za)
`
`OnePlus Ex. 1020.0006
`IPR2022-00048
`
`OnePlus Ex. 1020.0006
`IPR2022-00048
`
`
`
`constant frequency reference signal.
`
`The patent explains at
`
`column 2,
`
`lines 29 to 34 that in the 802.1la and 802.119,
`
`versions of the standard when data packets are inserted,
`
`they
`
`include a preamble, and that preamble contains a short training
`
`sequence followed by a long training sequence which are used to
`
`synchronize -- which are used for synchronization between the
`
`sender and receiver devices.
`
`Now the long training sequence uses BPSK and,
`
`therefore, each subcarrier in the training sequence consists of
`
`either a +l or a -1. That's just an artifact of using BPSK.
`
`So there are very few symbols that are actually available
`
`behind using BPSK coding, making it very important to be able
`
`fine tune the timing so that data in the packet is accurately
`
`read and and interpreted.
`
`In slide 10,
`
`this is a three-dimensional
`
`representation of an OFDM channel. At the top left in the kind
`greenish-gray area, you can see these are the short training
`
`fields.
`
`To the right,
`
`the blue squares represent the long
`
`training fields, and the gray blocks further to the right
`
`represent the data that is actually being transmitted. And as
`
`you can see in OFDM,
`
`there a lot of overlapping data occurring
`
`at the same time.
`
`Now with higher data throughputs,
`
`the patentees
`
`recognized the need to create longer training sequences to
`
`ensure proper synchronization between sending and receiving
`
`10
`
`11
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`OnePlus Ex. 1020.0007
`IPR2022-00048
`
`OnePlus Ex. 1020.0007
`IPR2022-00048
`
`
`
`wireless devices especially since we were going to start
`
`compacting more data than we were before.
`
`The solution that
`
`the inventors devised built upon the existing training
`
`sequences by adding subcarriers which are selected in a manner
`
`to minimize PAPR. You can see in the last slide essentially
`
`they took the existing long training sequences and they add
`
`subcarriers to either side. And there's a couple of examples
`
`in the patent. And they select these subcarriers such that the
`
`PAPR is minimal.
`
`And as we saw on the previous slide,
`
`these preambles
`
`are sent with every data packet so they're constantly being
`
`sent, so it's desirable to minimize the PAPR as much as
`
`possible.
`
`And unless the Court has any questions,
`
`I would hand
`
`it over to defendants’ counsel.
`
`THE COURT:
`
`I'm sure I will, but go ahead.
`
`MR. HURT: Good morning, your Honor.
`
`James Hurt from
`
`Perkins Coie on behalf of the defendants.
`
`THE COURT:
`
`Thank you.
`
`MR. HURT:
`
`So today for you,
`
`I am going to present a
`
`tutorial.
`
`The roadmap,
`
`I have four basic modules.
`
`Those four
`
`modules are going to be wireless basics,
`
`then switching to
`
`frequency and time domain,
`
`then talk a little bit about
`
`orthogonal frequency-division multiplexing or OFDM.
`
`(Court reporter interruption)
`
`10
`
`11
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`OnePlus Ex. 1020.0008
`IPR2022-00048
`
`OnePlus Ex. 1020.0008
`IPR2022-00048
`
`
`
`MR. HURT: Oh,
`
`I'm sorry.
`
`And then we'll talk a little about the 802.11
`
`standards themselves.
`
`So what is wireless digital
`
`communications? Fundamentally this is getting bits from the
`
`transmitting apparatus to the receiving apparatus.
`
`It involves
`
`the movement of information from the transmitter to the
`
`receiver. All it is moving,
`
`information from point A to point
`
`B.
`
`But that information needs to go to something called
`
`"the channel." What is the channel?
`
`I like to think of the
`
`channel like a hose.
`
`It's just a pipe that connects the
`
`transmitting device to the receiving device.
`
`The more
`
`bandwidth you use,
`
`the fatter the hose is going to be.
`
`So in
`
`802.11n, we're using a 20 megahertz channel. There are other
`
`technologies out there such as like CDMA that only use the 1.25
`
`megahertz channel.
`
`(Court reporter interruption)
`
`MR. HURT:
`
`I'm sorry.
`
`The channel,
`
`the wireless
`
`channel bandwidth affected the more data you can get through.
`
`But to get that information through, you must pass through that
`
`channel and that channel impairs and degrades the signal.
`
`So let's look at a typical WiFi environment. Here
`
`assuming your home office, you have a transmitter device called
`
`an AP going to your client.
`
`The signals are going to travel
`
`through that space. You might have a direct line path that
`
`Rr
`
`No
`
`w a a
`
`on
`
`OO
`
`~l
`
`ao
`
`©
`
`10
`
`11
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`23
`
`24
`
`25
`
`OnePlus Ex. 1020.0009
`IPR2022-00048
`
`OnePlus Ex. 1020.0009
`IPR2022-00048
`
`
`
`10
`
`goes from the transmitting device to the receiving device.
`
`You
`
`may have a path that bounces off the wall or you may have a
`
`path that bounces off your couch.
`
`Those three paths combine at
`
`the receiver. This is known as a multipath environment.
`
`It is
`
`the multipath environment that is one source of channel
`
`degradation.
`
`The signals bounce around the environment,
`
`they
`
`arrive at the receiver with different replicas at different
`
`times.
`
`Another impairment is what's known as signal fading or
`
`variation in received signal power. You can see, as you might
`
`expect,
`
`the further away you move from the transmitting device,
`
`your received signal gets lower. Here we have an example of
`
`the actual received signal. You see that the signal is moving
`
`up and down and doesn't foliow that straight line path. Where
`
`does that come from? That comes from what's called small scale
`
`interference. This possible small scale interference is a
`
`result of the multipath environment,
`
`the signals bouncing
`
`across the different objects in the environment and then
`
`combining at the receiver either constructively or
`
`destructively.
`
`Channel estimation. This is an important concept
`
`particularly to the '842 patent. Fora receiver to actually
`
`receive the information from the transmitter, it needs to know
`
`what the channel did to the signal.
`
`To do so,
`
`the receiver
`
`needs to know in advance what the transmitter is actually going
`
`10
`
`11
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`23
`
`24
`
`25
`
`10
`
`OnePlus Ex. 1020.0010
`IPR2022-00048
`
`OnePlus Ex. 1020.0010
`IPR2022-00048
`
`
`
`11
`
`to be transmitting.
`
`The '842 patent is about training
`
`sequences. As an example, we take the transmitter.
`
`It sends
`
`the signal through the channel. We see that the channel
`
`degrades the signal.
`
`It does something to it.
`
`The receiver
`
`gets that signal and it needs to look at it.
`
`It says hey, what
`
`did I receive from the channel? Oh.
`
`I see something that's
`
`distorted from the known signal that I'm expecting to receive.
`
`Once it sees that signal, it can correct for it.
`
`It makes that
`
`correction and says okay.
`
`Now I know what the channel is going
`
`to do to my signal.
`
`So moving on to the second part of the tutorial,
`
`frequency and time domain.
`
`I like to use an analogy for
`
`frequency and time domain. Here on the left, you see music
`
`notes on a scale.
`
`To the right, you see a speaker.
`
`The notes
`
`on a staff represent the frequency domain. These are the
`
`frequencies that you want to hear. But you don't actually hear
`
`those. What you hear is the time domain sequence or the sound.
`
`Something in between was transformed,
`
`the frequency into time.
`
`What does that?
`
`In this case, it's the piano.
`
`It's the
`
`transformer.
`
`It's the device that converts frequency, notes,
`
`to sound,
`
`time.
`
`Here's a visual demonstration. You can see as I take
`
`the frequency to the left,
`
`the period of the wave form
`
`increases. This corresponds to the low note on the scale. As
`
`we increase the frequency,
`
`the period of the wave form
`
`10
`
`11
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`23
`
`24
`
`25
`
`11
`
`OnePlus Ex. 1020.0011
`IPR2022-00048
`
`OnePlus Ex. 1020.0011
`IPR2022-00048
`
`
`
`12
`
`decreases. That will correspond to the high note on the scale.
`
`A signal can be described in both time or the frequency domain.
`
`They're effectively equivalent representations of the same
`
`signal, but they're described differently. One is saying here
`
`is what you look like in time,
`
`the other is saying here's what
`
`you look like in frequency.
`
`So here's an example. Here's a cosign of 128 hertz.
`
`This is saying I'm a cosign and the one sample is 128, that's
`
`what I want to transmit.
`
`You take the Inverse Fourier
`
`Transformer,
`
`this signal, you end up with an actual cosign wave
`
`in the time domain at 128 hertz. Similarly you take a 256
`
`hertz cosign wave. You have a single sample saying,
`
`I want
`
`256.
`
`Take the Inverse Fourier Transform of that, you end up
`
`with a cosign 256 hertz.
`
`So you might ask yourself, what happens if I combine
`
`them? What is this going to look like?
`
`So we put on the left
`
`both 128 and 256,
`
`take that Inverse Fourier Transformer. What
`
`do we have? We have something that doesn't look like a cosign
`
`wave anymore because the signals have combined and now we have
`
`the combined representation of both 128 and 256. We know that
`
`the time domain signal on the right was synthesized or created
`
`from the frequency domain signal on the left.
`
`Moving briefly into OFDM or Orghogonal Frequency
`
`Division Multiplexing.
`
`I want to explain exactly what OFDM is
`
`compared to some other techniques and talk a little bit more
`
`10
`
`11
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`12
`
`OnePlus Ex. 1020.0012
`IPR2022-00048
`
`OnePlus Ex. 1020.0012
`IPR2022-00048
`
`
`
`13
`
`about subcarriers.
`
`So before we get specifically into OFDM,
`
`I
`
`want to talk a little bit more about wireless spectrum.
`
`I'm sure, as your Honor knows, back in the 80's, back
`
`in the 90's, we had radio bands. Oftentimes we would have to
`
`scan our FM radios to figure out what music channel we wanted
`
`to listen to. Here you can see as we scan the FM radio band,
`
`the frequency or peak of what channel we want to tune to
`
`increases. Once we see that specific channel, we go ahead and
`
`tune back to there. And we see, boom, here's the signal that
`
`we want, here's the frequency at which it was present.
`
`The point being here is, a signal may be transmitted
`
`at different frequencies, as if using different channels of a
`
`FM radio without changing the information content. What this
`
`means is that you can have the same song playing on 88.3 as
`
`91.1. They're on two different frequency channels, but it's
`
`the same information content.
`
`It's the ability to send that
`
`information on separate frequencies at the same time. That's
`
`the basis of OFDM.
`
`So going back, what is OFDM? Here is an analogy I
`
`like to think of. Going back to the hose or that fat pipe, you
`
`have a single fat pipe of water. That's your bandwidth in a
`
`single carrier system. We're going to take that pipe of water
`
`and we're going to divide into multiple independent parallel
`
`streams,
`
`like from a showerhead. That’s the picture to your
`
`right.
`
`RB
`
`IN
`
`ud
`
`Ay
`
`ao
`
`ov
`
`~ o e
`
`o
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`13
`
`OnePlus Ex. 1020.0013
`IPR2022-00048
`
`OnePlus Ex. 1020.0013
`IPR2022-00048
`
`
`
`14
`
`So why would we use OFDM?
`
`OFDM is more efficient.
`
`Here is a spectrum comparison for the same data rate
`
`transmission, if we use multi-carrier, multiple faucets or like
`
`an FM radio we have to have guard bands in between each
`
`station, but we're able to go ahead and use every channel on
`
`that FM radio band to transmit data, or we can decide to try to
`
`glop all that data together and do something called "single
`
`carrier."
`
`Now single carrier when you spread the data rate, it
`
`causes bandwidth to expand. That's the basis for a technology
`
`called CDMA, which was actually invented here in San Diego by a
`
`company called Qualcomm.
`
`A similar technology called frequency
`
`hopping spread spectrum was actually invented in the '40s by
`
`Austrian-born actress Hedy Lamarr.
`
`She, during the 1940s,
`
`worked with our allies to help the Allies defeat the Germans by
`
`coming up with a system that would hop frequencies to overcome
`
`the German jamming of the Allied torpedoes.
`
`Similarly though when you take away from a single
`
`carrier, we can crunch even more. We can get down to OFDM
`
`because we're able to overlap these subcarriers and these
`
`signals in a very special way. This is a very similar slide to
`
`what co-counsel has shown you before.
`
`I want to point out a couple key things about this
`
`one. When you look at the peak of the red signal, you'll see
`
`that all of the other colors go to zero. That's what it means
`
`he
`
`N W
`
`w
`
`fi
`
`uo
`
`oY
`
`~ a \
`
`o
`
`10
`
`11
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`24
`
`25
`
`14
`
`OnePlus Ex. 1020.0014
`IPR2022-00048
`
`OnePlus Ex. 1020.0014
`IPR2022-00048
`
`
`
`15
`
`The '842 is
`to be orthogonal in the context of the '842.
`saying use all these different signals, use them in a
`non-interfering way to bring the data across all subcarriers.
`The subcarriers spacing is an important feature in
`
`They need to be spaced at certain regular spacing so
`OFDM.
`they maintain orthogonal.
`In this case, we call that Delta F.
`And the K or the index value is just a number how far away from
`
`the center.
`
`So in 802.1la,
`
`there are 52 subcarriers.
`
`They range
`
`from -26 to +26.
`
`In 802.1ln,
`
`the technology used today, we go
`
`from -28 to +28. You've added four subcarriers that we're
`using. But to be clear,
`those subcarriers were already there.
`There are 64 defined subcarriers in the system.
`The question
`is not were they added: Were they used. That's the primary
`
`difference between 802.1la and 802.11n.
`
`They were
`The patentee did not invent subcarriers.
`They were simply not used before.
`In fact, none of
`present.
`the stuff I discussed today so far was invented by the
`patentee. All this was known technology, known techniques.
`So moving quickly into the 802.11 family of standards.
`
`I just want to point out a couple
`I know this is a busy slide.
`of things.
`In 1999, 802.1la was introduced, using 20 megahertz
`of bandwidth channel.
`It was based on OFDM.
`It's max data
`
`Ten years later in 2009,
`rate was 54 megabits per second.
`802.11n was introduced.
`It also has an option or capability to
`
`10
`
`11
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`15
`
`OnePlus Ex. 1020.0015
`IPR2022-00048
`
`OnePlus Ex. 1020.0015
`IPR2022-00048
`
`
`
`16
`
`use 20 megahertz channels, also based on OFDM technology, but
`its max data rate goes to 600 megabits per seconds. WiFi has
`
`evolved both in the technology used and the max data rate that
`
`it supports.
`The key thing about 802.11 though was is it was
`
`designed to be backwards compatible. That meant the older
`devices and newer devices need to be able interoperate
`
`together, but more fundamentally, it put constraints on newer
`standards.
`The standards cannot go and change things that the
`
`older devices are expecting to see.
`
`So during training
`
`the values that the receiver is going to use to
`sequences,
`determine what the channel did to its signal already defined
`
`the value.
`
`The BPSK value for that subcarrier, it cannot be
`
`changed.
`
`The '842 patent was about determining those four
`values that they're going to use on the two extra subcarriers
`on the left and the two extra subcarriers on the right. That's
`
`the invention. That's what they're claiming,
`this inventive
`sequence that's four defined values for subcarriers on the left
`
`and the right.
`
`Here it is. This is the actual 802.lla training
`
`sequence. Again, 64 subcarriers already there, existed the
`entire time. Only 52 were active and it has a -26 toa 26 with
`
`dc with a zero index not being used. This training sequence
`
`was already defined in 802.11.
`
`10
`
`il
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`23
`
`24
`
`25
`
`16
`
`OnePlus Ex. 1020.0016
`IPR2022-00048
`
`OnePlus Ex. 1020.0016
`IPR2022-00048
`
`
`
`17
`
`53 subcarriers
`The sequence itself is on the bottom.
`is OFDM training symbol, modulated by a sequence of L. Those L
`sequence values are all +1 or -ls BPSK. You can think of a
`training sequence just like the notes on a scale.
`The receiver
`knows what the transmitter is going to be sending during this
`training sequence.
`It's used so the receiver can figure out
`
`what did the channel do to my signal.
`
`So 802.11n came along. What do we want? What do we
`
`802.11n
`always want? We want better, faster, cheaper.
`increased the data rate froma 54 megabits per second to 600
`
`megabits per second. Many different ways for the system
`designers to achieve that goal. One of the ways they achieved
`that goal was to increase the used subcarriers.
`So again, only have 64. Using 52 in 802.1la.
`802.lin, all right, let's use four more. What enabled that was
`improved digital filtering technology. Technology not invented
`
`by the patentee here.
`So now instead of having six subcarriers on the left
`
`and five on the right, we can decrease it, use those extra
`subcarriers to carry more data.
`To do that, you have to define
`
`values for those subcarriers during the training sequence, so
`you can determine what did the channel do to that specific
`
`subcarrier.
`So the actual patent itself was a patent application
`filed by Broadcom during the 802.11n standardization process.
`
`10
`
`11
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`7
`
`OnePlus Ex. 1020.0017
`IPR2022-00048
`
`OnePlus Ex. 1020.0017
`IPR2022-00048
`
`
`
`18
`
`The specification disclosed the exact same training sequence as
`specified in the eventual standard. Again, you were required
`to start with 802.lla.
`They didn't invent the entire sequence.
`
`It was already there for them. Here jt is again,
`
`the 802.1lla
`
`training sequence.
`
`Now I want to talk a little bit about peak-to-average
`
`power ratios as counsel discussed as well. Here's the sequence
`on the left. You take the Inverse Fourier Transform with this
`the sequence, you end up with this sequence on the right.
`The
`sequence to the right is the power sequence that you actually
`will get out when you take the Inverse Fourier Transform.
`I have shown the solid red line,
`the average value.
`
`And the dotted green line is the peak value. As counsel
`
`indicated, depending upon the ratio of the peak to the average,
`it's going to matter how much variability you have going into
`your power amplifier.
`The more variability,
`the more back-off
`you need.
`So he's right, Minimizing peak-to-average power
`
`ratio is an important aspect of OFDM's system.
`
`But one thing I want to know, if you look at the left
`the one in frequency domain, it consists of only +1s
`sequence,
`If you take the power of that sequence, its peak
`and -1s.
`power and its average power are identical.
`They are both 1.
`Because when you take a 1 or a -1l and you square it or multiply
`it by itself, 1x1, 1,
`-1x*% -1, 1.
`So it's the peak in the
`average in the frequency domain where a sequence is defined by
`
`10
`
`11
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`23
`
`24
`
`25
`
`18
`
`OnePlus Ex. 1020.0018
`IPR2022-00048
`
`OnePlus Ex. 1020.0018
`IPR2022-00048
`
`
`
`19
`
`BPSK is 1.
`
`It has no peak-to-average because they're exactly
`
`the same thing which means in the context of the '842 patent,
`peak-to-average power ratio is a time domain property. There
`is no peak-to-average power ratio for a frequency domain
`
`signal.
`Here is the '842 patent and the 802.11n training sequence.
`
`The four red dots, that's the supposedly inventive sequence of
`
`those subcarriers already existed.
`the '842 patent. Again,
`They were already there. What the patentee had to figure out
`was what do I want to put on these four subcarriers?
`Do I want
`
`to put a +1 or do I want to put a -1 because there were only
`four additional subcarriers and we were restricted to +1s and
`
`there are only 16 possibilities the patentee could have
`-lIs,
`chosen from.
`It turns out that this selection,
`1 out of 16,
`
`this is the one that gives you the minimal peak-to-average
`
`power ratio when converted to time domain.
`That property of the peak-to-average power ratio in the
`
`time domain is an inherent characteristic of the frequency
`
`domain sequence that you selected. Had you changed any one of
`these red dots from a +l to a -l1 or take a -l1 to a +1,
`the
`
`corresponding peak-to-average power ratio will go up.
`Let's go ahead and do that.
`I take the Inverse Fourier
`Transformer,
`the extended long training sequence defined in the
`'g42 patent and, again, we get to the right a power domain
`sequence. And you'll notice the peak-to-average power ratio
`
`10
`
`11
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`19
`
`OnePlus Ex. 1020.0019
`IPR2022-00048
`
`OnePlus Ex. 1020.0019
`IPR2022-00048
`
`
`
`20
`
`from 80lla to 802.11n went up just a little bit.
`It went from
`3.2 dB to 3.6 dB. dB is a relative scale that engineers like to
`use. Approximately 3 dB is a factor of 2.
`From .2 to .6 is
`just a smidgen more. Not a big deal. But the patentee and BNR
`are correct, you do want to try to minimize this. But you only
`had four values to mess with to figure out how you wanted to do
`
`this.
`
`Thank you, your Honor.
`
`THE COURT: Okay. All right.
`
`Do you want to start
`
`with the first term that's at issue here?
`
`MR. HARTSELL:
`
`I understood that the defendants would
`
`be presenting first since they're the ones who put this term up
`
`for construction.
`
`MR. HURT:
`
`I'm happy to present first, your Honor.
`
`THE COURT: Okay.
`
`Go ahead.
`
`MR. HURT:
`
`Do you mind if I do a have a quick swig of
`
`water?
`
`THE COURT: No, go right ahead.
`
`MR. HURT: All right.
`
`So we're here back to talk
`
`about the proper construction of the Inverse Fourier
`
`Transformer. Here is the claim language:
`
`Wherein the Inverse Fourier Transformer processes the
`extended long training sequence, which we've discussed quite a
`bit before,
`from the signal generator and provides what? An
`optimal extended long training sequence with a minimal
`
`peak-to-average ratio.
`
`10
`
`il
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`21
`
`22
`
`23
`
`24
`
`25
`
`20
`
`OnePlus Ex. 1020.0020
`IPR2022-00048
`
`OnePlus Ex. 1020.0020
`IPR2022-00048
`
`
`
`21
`
`On the bottom left, I've shown again, here is the
`
`I
`
`extended long training sequence of the '842.
`
`Take the Inverse
`
`Fourier Transform, you end up with this.
`
`It has
`
`peak-to-average power ratio of 3.6 dB.
`
`Let's look at the proposed constructions. Defendants
`
`A circuit and/or software that performs a defined
`propose:
`mathematical function that transforms a series of values from
`
`the frequency domain into the time domain.
`
`BNR proposes: Plain and ordinary Meaning, or
`
`alternatively, circuit and/or software that at least performs
`
`Inverse Fourier Transform.
`
`Let's talk about the first, plain and ordinary
`
`meaning.
`
`THE COURT: This is a fundamentally, perhaps, stupid
`
`question, but why does it bounce back and forth from Inverse
`
`Fourier Transform to Inverse Fourier Transformer?
`
`MR. HURT:
`
`So the transform is the actual defined
`
`mathematical formula or the function.
`
`The transformer is just
`
`something that implements that function.
`
`THE COURT: And this is clearly something that people
`
`who do this stuff would recognize.
`I mean, it's written in the
`patent in initial caps.
`So while I'm not exactly sure what it
`is,
`I would suspect you certainly would know, people who would
`practice this sort of technology are going to recognize this,
`and I think in both the briefing, it was recognized that this
`
`10
`
`11
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`21
`
`OnePlus Ex. 1020.0021
`IPR2022-00048
`
`OnePlus Ex. 1020.0021
`IPR2022-00048
`
`
`
`22
`
`is a mathematical function that an electrical engineer is going
`
`to recognize. They're going to know what this is.
`MR. HURT: Yes. Absolutely agree with your Honor.
`
`THE COURT:
`So why do I have to do anything more with
`it when it says that it's there, it's operatively coupled with
`the signal generator, and it's going to process this extended
`long training sequence from the signal generator and provide an
`optimal extended long training sequence with a minimal
`peak-to-average ratio?
`Isn't it just doing what the formula
`
`does?
`
`MR. HURT: Yes, your Honor. But
`
`two reasons why you
`
`should construe this term. One is to provide -- resolve the
`dispute between the parties as to the exact scope of this claim
`term.
`Second, provide clarity and guidance to the finder of
`fact. Going back to the first, we have a fundamental dispute
`with respect to what defendants believe the Inverse Fourier
`Transformer of the '842 is doing relative to what BNR proposes
`
`that an Inverse Fourier Transformer in the abstract can do.
`
`BNR has already proposed and argued that the Inverse
`
`Fourier Transformer can be multi-dimensional, can operate
`
`between multiple domains. Defendants do not dispute that
`mathematical concept in the abstract. What defendants -- our
`concerns are is that even when we get to expert reports, if we
`
`have a fundamental dispute,
`
`the arguments are not going to be
`
`joined. We're going to be talking about the ‘842 Inverse
`
`10
`
`11
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`23
`
`24
`
`25
`
`22
`
`OnePlus Ex. 1020.0022
`IPR2022-00048
`
`OnePlus Ex. 1020.0022
`IPR2022-00048
`
`
`
`23
`
`Fourier Transformer,
`
`the one that takes frequency domain
`
`BNR will be talking about this
`signals into time domain.
`amorphous transform that can -— according to them, can do
`anything.
`It can take any number of dimensions, go anywhere to
`any space to any other space. Yet the '842 patent never talks
`about anything else other than frequency in time.
`
`THE COURT:
`
`So fundamentally, it's not this
`
`mathematical functionality but rather that this claim is
`
`directed as a wireless communication device that comprises this
`
`transformer?
`
`MR. HURT: Absolutely, your Honor.
`
`THE COURT:
`
`In the context of the claim language
`
`itself that says this is a transformer that is comprised in a
`
`wireless communications device, your argument is how it
`
`operates that mathematical principle is limited?
`
`MR. HURT: Absolutely.
`
`THE COURT: And limited to this frequency into time
`
`domain?
`
`MR. HURT: Yes, your Honor. That's exactly correct.
`
`THE COURT: Okay. And now help me find out other than
`
`beyond the fact that it says that it's a wireless
`communications device, why would someone recognize that?
`
`MR. HURT: Absolutely. Happy to do that.
`
`So
`
`plaintiffs and defendants agree that this can be a circuit
`
`and/or software.
`
`No dispute there.
`
`10
`
`il
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`23
`
`24
`
`25
`
`23
`
`OnePlus Ex. 1020.0023
`IPR2022-00048
`
`OnePlus Ex. 1020.0023
`IPR2022-00048
`
`
`
`24
`
`But let's lo