`
`UNITED STATES DISTRICT COURT
`SOUTHERN DISTRICT OF CALIFORNIA
`BEFORE HONORABLE CATHY ANN BENCIVENGO, JUDGE PRESIDING
`
`BELL NORTHERN RESEARCH, LLC,,
`
`Plaintiff,
`
`vs.
`
`)
`)
`)
`)
`)
`)
`)
`
`COOLPAD TECHNOLOGIES, INC. AND
`)
`YULONG COMPUTER COMMUNICATIONS, )
`)
`Defendants.
`)
`------------------------------)
`BELL NORTHERN RESEARCH, LLC,
`)
`)
`
`Plaintiff,
`
`vs.
`
`HUAWEI TECHNOLOGIES Co., LTD.,
`HUAWEI DEVICE (HONG KONG) CO.,
`LTD., and HUAWEI DEVICE USA,
`INC.,
`
`)
`)
`
`)
`)
`)
`)
`)
`)
`
`Defendants.
`
`Plaintiff,
`
`)
`--------------------------------)
`BELL NORTHERN RESEARCH, LLC.,
`)
`)
`)
`)
`)
`
`vs.
`
`)
`KYOCERA CORPORATION and KYOCERA )
`INTERNATIONAL INC.,
`)
`)
`
`Defendants. )
`--------------------------------)
`
`CASE NO. 18CV1783-CAB-BLM
`
`SAN DIEGO, CALIFORNIA
`
`THURSDAY, JUNE 20, 2019
`
`CASE NO. 18CV1784-CAB-BLM
`
`CASE NO. 18CV1785-CAB-BLM
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`
`
`2
`
`CASE NO. 18CV1786-CAB-BLM
`
`CASE NO. 18CV2864-CAB- BLM
`
`Plaintiff,
`
`vs.
`
`ZTE CORPORATION, ZTE
`ZTE
`(TX)
`INC.
`
`(USA)
`
`INC.
`
`------------------------------ )
`BELL NORTHERN RESEARCH, LLC.,
`)
`)
`)
`)
`)
`)
`)
`)
`Defendants.)
`----------------------- ------- )
`BELL NORTHERN RESEARCH, LLC,,
`)
`)
`)
`)
`)
`)
`)
`)
`)
`)
`)
`)
`
`Plaintiff,
`
`vs.
`
`LG ELECTRONICS, INC., LG
`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
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`
`
`3
`
`COUNSEL APPEARI NG:
`For The Plaintiff:
`
`For The Defend ant s
`Coolpad and Yulong:
`
`for the Defendant s
`Huawei entities:
`
`For The Defend ant s
`ZTE entities:
`
`ALSO PRESENT:
`
`Sadaf Ra j a Abdul lah, Esq.
`Steven W. Hartsel l , Esq.
`Paul J. Skiermont, Esq.
`SKI ERMONT DERBY LLP
`Thanksgiving Tower
`160 1 Elm St reet, Suite 4400
`Da l las, Texas 75201
`
`Thomas Nathan Millikan, Esq.
`James Young Hurt, Esq.
`PERKI NS COIE, LLP
`11988 El Camino Real, Suite 350
`San Diego, California 92130
`
`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
`
`Amal Ajay Pari kh, Esq.
`Thomas DaMario, Esq.
`McDERMOTT WILL & EMERY LLP
`4 44 West Lake street, Suite 4000
`Chi cago, Il l inois 60606
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`
`
`4
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`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
`
`'842 patent was developed by engineers at Broadcom and filed in
`
`
`
`5
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`January of 2010. The
`
`'842 patent is a continuation of U.S.
`
`Patent Number 7,646,703 which claims pr i ority to at least
`
`July 2004. The
`
`'842 patent is directed to long train ing
`
`sequences with minimum peak-to-average power ratios, and today
`
`I wou ld l i ke 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 Wi Fi 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 s t a n ds for
`
`orthogonal frequency-division multiplexing, t o increase data
`
`t h roughput.
`
`I'm going to show you what that means on slide 5.
`
`At the t op, you can see thi s is how data was transmitted OFDM.
`
`Basically we have single carriers that are separated. When
`
`OFDM is i mp l ement ed, the carrier waves are essentially smu shed
`
`togethe r allowing you to send more data foun d 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 sl i de 6, each colored peak is a subcarrier which
`
`
`
`6
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`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 i nterfering with which other.
`
`Now as with many things while OFDM provides throughput
`
`improvement s and other advantages, it also brings certain
`
`disadvantages. And one of t he 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 p r esence of large numbers of
`
`independent l y 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 consumpt i on battery drai n .
`
`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 ampl ifier efficiency and
`
`higher overall power consumption.
`
`Another concept that may come up today is BPSK. BPSK
`
`stands f o r binary phase sh i ft keying which is a digital
`
`modulation process by changing or modulating a phase of a
`
`
`
`7
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`1 3
`
`14
`
`1 5
`
`1 6
`
`1 7
`
`18
`
`1 9
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`constant frequency reference signal. The patent explains at
`
`column 2, l i nes 29 to 34 that in the 802.l l a and 802.llg,
`
`versions of the standard when data packets are inserted, they
`include a preamble, and that preamble cont ains a short training
`
`sequence fo l lowed by a long training sequence which are used to
`
`synchronize -- which are used for synchroni zation between the
`
`sender and receiver devices.
`
`Now the long training sequence uses BPSK and,
`
`therefore, each subcarrier i n the train i ng 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 i n the packet is accurately
`
`read and and interpreted.
`
`In slide 10, this is a three-di mensional
`
`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 squ ares 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 overl apping data occurring
`
`at the same time.
`
`Now with higher data throughputs, the patentees
`
`recognized the need to creat e longer t r aini ng sequences to
`
`ensure proper synchronizat i on between sending and receiving
`
`
`
`8
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`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 minimi ze PAPR. You can see in the last slide essential l y
`
`t h ey took the existing long training sequences and they add
`
`subcarriers to either side. And there's a couple of exampl es
`
`in the patent. And they select these subcarriers such that the
`
`PAPR is min i mal.
`
`And as we saw on the previous slide, these preambles
`
`are sent with every data packet so they're constantly being
`
`sent, so i t 's desirable to minimize the PAPR as much as
`
`possible.
`And unless the court has any questions, I would hand
`
`it over t o 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 go i ng to be wireless basics, then switching to
`
`frequency and time domain, then talk a
`
`l ittle bit abou t
`
`orth ogonal frequency-division multiplexing or OFDM.
`
`(Court reporter interruption)
`
`
`
`9
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`MR. HURT: Oh, I'm sorry.
`
`And then we'll ta l k 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 cal l ed
`
`"the channe l ." What is the channel?
`
`I like to think of the
`
`channel l i ke 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.lln, we're u sing a 20 megahertz channe l . 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 channe l , the wireless
`
`channel bandwidth affected the more data you can get through.
`
`But to ge t that information through, you must pass through that
`
`chann el and that channel impairs and degrades the signal .
`
`So let's look at a typical WiFi environment. Here
`
`assuming your home o f fice, you have a transmitter device called
`
`an AP go i ng t o your client. The signals are going to travel
`
`through that space. You mi ght have a direct line path that
`
`
`
`10
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`1 3
`
`1 4
`
`15
`
`16
`
`1 7
`
`18
`
`1 9
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`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
`It is
`the receiver . This is known as a multipath environment.
`
`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 dif f erent
`
`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 fo l l ow that straight line path. Where
`does that come from? That comes from what 's called small scal e
`interference . This possible small scale int erference is a
`result of the multipath environment, the signals bouncing
`
`across the different objects in the envi ronment and then
`
`combining at the receiver e i ther constructi vely or
`
`destructively .
`
`Channel estimation. This is an important concept
`particularly to the ' 842 patent. For a receiver to actually
`receive the information from the transmi tter, it needs to know
`what the channel did to the signal. To do so, t he receiver
`needs to know in advance what the transmitter is actually going
`
`
`
`1 1
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`to be transmi t ting. 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 t he 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 s i gnal 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 doma i n. These are the
`
`frequenc i es that you want to hear. But you don't actu ally hear
`
`those. What you hear is the time domain sequence or the sound.
`
`Something in between was transformed, the frequ ency into t i me.
`
`What does t hat?
`
`In this case, it's the piano.
`
`It's the
`
`tran sformer. It's the device that converts frequency, notes,
`
`to sound, time .
`
`Here's a visual demonstration. You can see as I ta ke
`
`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
`
`
`
`12
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`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 t i me, 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 I nverse Fourier
`
`Transformer, t his signal, you end up with an actual cosign wave
`
`in the time domain at 128 hert z. Similarly you take a 256
`
`hertz cosign wave. You have a single sampl e 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 l eft
`
`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 comb i ned 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
`
`
`
`13
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`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.
`
`
`
`14
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`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
`
`
`
`15
`
`'842 is
`
`to be orthogonal in the context of the '842. The
`saying use all these differen t signals, use them in a
`non-interfering way to bring the data across all subcarriers.
`The subcarrie r s spacing is an important feature in
`OFDM. They need to be s p aced at certain regular spacing so
`In this case, we call that Delta F.
`they mainta i n ort hogonal.
`And the Kor t he index value is just a number how far away from
`
`the center .
`So in 802.lla, there are 52 subcarriers. They range
`In 802. l ln, the technology used today, we go
`-26 to +26.
`from
`from -28 to +28. You've added four subcarriers that we're
`using. But to be clea r , those subcarriers were already there.
`There are 64 defined subcarriers in the system. The question
`is not were they added: Were they u sed. That's the primary
`difference bet ween 802.lla and 802.lln.
`The patentee did not invent subcarriers. They were
`In fact, none of
`
`present. They were simply not used before.
`the stuff I discussed today so far was invented by the
`patentee. All this was known technology, known techniques.
`So moving quickly i n to the 802.11 family of standards.
`I just want to point out a couple
`I know t his is a busy slide.
`In 1 9 99, 8 02.l l a was introduced, using 20 megahertz
`It's max data
`
`of things.
`
`It was based on OFDM.
`of bandwi dth channel.
`rate was 54 megabits per second. Ten years later in 2009,
`It also has an option or capability to
`
`802.lln was introdu ced.
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`
`
`16
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`1 1
`
`1 2
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`u se 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 techno l ogy used and the max data rate that
`
`i t supports.
`The key thing about 8 02.11 though was is i t was
`designed to be backwards compatible. That meant the older
`devices and newer devices need to be abl e interoperate
`together, but more fundamentally, i t put constraints on newer
`standards. The standards cannot go and change t hings that the
`older devices are expecting to see. So during training
`sequences, the values that the receiver is going to use to
`determine what the channel did to its signal already defined
`the value. The BPSK value for that subcarrier, i t cannot be
`
`changed.
`
`'842 patent was about determining those four
`The
`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 i t is. This is the actual 802.lla training
`sequence. Again, 64 subcarriers already there, existed the
`entire time. Only 52 were active and i t has a -26 to a 26 with
`de with a zero index not being used. This training sequence
`
`was already defined in 802.11.
`
`
`
`17
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`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 +l 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
`I t 's used so the receiver can figure out
`
`training sequence.
`
`what did the channel do to my signal.
`So 802.lln came along. What do we want? What do we
`always want? We want better, faster, cheaper. 802.lln
`increased the data rate from a 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.lla.
`802.lln, 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 actua l patent itself was a patent app l ica t ion
`fi l ed by Broadcom dur i ng the 802.lln standardization process.
`
`
`
`18
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`The speci fica t ion 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 sequ ence.
`It was a l ready there for them. Here i t is again, the 802. l la
`
`training sequence.
`Now I want to talk a little bit about peak-to-average
`power ratios as cou nsel 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,
`i t 's going to matter how much variability you have going into
`your power amplifier. The more variability, the more back-off
`So he's right, minimizing peak-to-average power
`
`you need.
`
`ratio is an important aspect of OFDM's system.
`But one thing I want to know, if you look at the left
`sequence, the one in frequency domain, i t consists of only +ls
`and -ls. If you take the power of that sequence, its peak
`power and its average power are identical. They are both 1.
`Because when you take a 1 or a -1 and you square i t or multiply
`So i t ' s the peak in the
`
`i t by itself, 1 x 1, 1, -1 x -1, 1.
`average in the frequency domain where a sequence is defined by
`
`
`
`19
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`1 0
`
`11
`
`12
`
`1 3
`
`14
`
`1 5
`
`16
`
`1 7
`
`1 8
`
`1 9
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`It has no peak-to-average because they"re exactly
`BPSK is 1 .
`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. l ln t raining sequence.
`The four red dots, that's the supposedly inventive sequence of
`the '842 patent. Again, those subcarriers already existed.
`They were already there. What the patentee had to figure out
`was what do I want to put on t hese four subcarriers? Do I want
`to put a +l or do I want to put a -1 because there were only
`four addi tional subcarriers and we were restricted to +ls and
`-ls, there are only 16 possibilities the patentee could have
`It turns out that this se l ection, 1 out of 16,
`chosen from.
`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 f r equency
`domain sequence that you selected. Had you changed any one of
`these red dots from a +l to a -1 or take a -1 to a +l, the
`corresponding peak-to-ave r age power ratio will go up.
`I take the Inverse Fourier
`Let's go ahead and do that.
`Transformer, the extended l ong trai n ing sequence defined in the
`'842 patent and, again, we get to the r i ght a power domain
`sequence. And you'll notice the peak-to-average power ratio
`
`
`
`20
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`1 0
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`from 8011a to 802.lln went up just a little bit. It went from
`3.2 dB to 3.6 dB. dB is a relative scale that engin eers like to
`From .2 to .6 i s
`use. Approximately 3 dB is a factor of 2.
`just a smidgen more. Not a big deal. But the patentee and BNR
`are correct, you do want to try to minimize t his. Bu t you only
`had four values to mess with to figure o u t 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?
`I understood that the defen dants would
`MR. HARTSELL:
`be presenting first since they're the ones who put this term up
`
`for construction.
`
`MR. HURT:
`
`I'm happy to present f i rst, your Honor.
`
`THE COURT: Okay. Go ahead.
`MR. HURT: Do you mind if I do a have a q u ick swig of
`
`water?
`
`THE COURT: No, go right ahead.
`MR. HURT: Al l right. So we' r e here back to t a lk
`about the proper construction of the Inverse Fourier
`Tran sformer. Here is the claim language:
`Where i n the I nve r se Fourier Tr ansfo rmer processes the
`extended long tra i ning sequence, which we've d iscussed quite a
`bit before, from the signa l generator and provid es what? An
`optimal extended long training sequence with a minima l
`
`peak-to-average ratio.
`
`
`
`21
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`on the bottom left, I've shown again, here is the
`
`extended long trai n ing sequence of the '842. Take the Inverse
`
`Fourier Transform, you end up with this.
`
`I t has
`
`peak-to-average power ratio of 3.6 dB.
`
`Let's look at the proposed constructions. De fendan t s
`
`propose: A circuit and/or software that performs a defined
`
`mathematica l function that t ransforms a series of values from
`
`the frequency domain i n to the time domain.
`
`BNR proposes: Plain and ordinary meaning, or
`
`alternative l y, 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 I n verse Fourier Transformer?
`
`MR. HURT:
`
`So the transform is the actual defined
`
`mathematica l formula or the function. The transformer is just
`
`something tha t
`
`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 ini t ial 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 thi s,
`
`and I th i nk in both the briefing, it was recognized that this
`
`
`
`22
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`is a mathematical function that an electrical engineer is going
`to recognize. They're going t o know what this is.
`MR. HURT: Yes. Absolutely agree with your Honor.
`So why do I have to do anything more with
`THE COURT:
`i t when i t says that i t ' s there, i t ' s operatively coupled with
`the signal generator, and i t ' s going to p rocess this extended
`long tra i ning sequence from the signal generator and provide an
`optimal extended long training sequence with a minimal
`Isn ' t i t just doing what the formula
`
`peak-to-average ratio?
`
`does?
`
`MR. HURT: Yes, your Honor. But two reasons why you
`should construe this term. One is to provide -- resolve the
`dispute between t he parties as to t h e exact scope of this claim
`term. Second, provide clarity and guidance to the finder of
`fact. Going back to the f i rst, we have a fun d amental dispute
`with respect to what defendan t s 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-dimens i onal, can operate
`between multiple domains. Defendants d o not dispu te 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
`
`
`
`23
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`J
`
`Fourier Transformer, the one that takes frequency domain
`signals into t i me domain. BNR will be talking about this
`amorphous transform that can -- according to them, can do
`I t can take any number of dimensions, go anywhere to
`anything.
`any space to any other space. Yet the '8 42 patent never tal k s
`about anything else other than frequency in time.
`i t's not this
`
`THE COURT:
`mathematical functionality but rather that this claim is
`directed as a wireless communication device tha t comprises th i s
`
`So fundame