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
`
`UNITED STATES DISTRICT COURT
`SOUTHERN DISTRICT OF CALIFORNIA
`BEFORE HONORABLE CATHY ANN BENCIVENGO, JUDGE PRESIDING
`
`________________________________
` )
`BELL NORTHERN RESEARCH, LLC,, )
` )
` Plaintiff, ) CASE NO. 18CV1783-CAB-BLM
` )
` vs. )
` ) SAN DIEGO, CALIFORNIA
`COOLPAD TECHNOLOGIES, INC. AND )
`YULONG COMPUTER COMMUNICATIONS, )
` ) THURSDAY, JUNE 20, 2019
` Defendants. )
`--------------------------------)
`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, )
`INC., )
` Defendants. )
`--------------------------------)
`BELL NORTHERN RESEARCH, LLC., )
` )
` Plaintiff, ) CASE NO. 18CV1785-CAB-BLM
` )
` vs. )
` )
`KYOCERA CORPORATION and KYOCERA )
`INTERNATIONAL INC., )
` )
` Defendants. )
`--------------------------------)
`
`
`
`
`
`
`
`
`
`SAMSUNG EXHIBIT 1026
`
`Page 1 of 122
`
`
`
` 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
`
`UNITED STATES DISTRICT COURT
`SOUTHERN DISTRICT OF CALIFORNIA
`BEFORE HONORABLE CATHY ANN BENCIVENGO, JUDGE PRESIDING
`
`________________________________
` )
`BELL NORTHERN RESEARCH, LLC,, )
` )
` Plaintiff, ) CASE NO. 18CV1783-CAB-BLM
` )
` vs. )
` ) SAN DIEGO, CALIFORNIA
`COOLPAD TECHNOLOGIES, INC. AND )
`YULONG COMPUTER COMMUNICATIONS, )
` ) THURSDAY, JUNE 20, 2019
` Defendants. )
`--------------------------------)
`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, )
`INC., )
` Defendants. )
`--------------------------------)
`BELL NORTHERN RESEARCH, LLC., )
` )
` Plaintiff, ) CASE NO. 18CV1785-CAB-BLM
` )
` vs. )
` )
`KYOCERA CORPORATION and KYOCERA )
`INTERNATIONAL INC., )
` )
` Defendants. )
`--------------------------------)
`
`
`
`
`
`
`
`
`
`SAMSUNG EXHIBIT 1026
`
`Page 1 of 122
`
`
`
` 2
`
` 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
`
`--------------------------------)
`BELL NORTHERN RESEARCH, LLC., )
` )
` Plaintiff, ) CASE NO. 18CV1786-CAB-BLM
` vs. )
` )
`ZTE CORPORATION, ZTE (USA) INC. )
`ZTE (TX) INC. )
` )
` Defendants.)
`--------------------------------)
`BELL NORTHERN RESEARCH, LLC,, )
` )
` Plaintiff, ) CASE NO. 18CV2864-CAB-BLM
` )
` 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
`
`Page 2 of 122
`
`
`
` 3
`
` 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
`
`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 Thomas Nathan Millikan, Esq.
`Coolpad and Yulong: James Young Hurt, Esq.
` PERKINS COIE, LLP
` 11988 El Camino Real, Suite 350
` San Diego, California 92130
`
`for the Defendants Joanna M. Fuller, Esq.
`Huawei entities: 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
`
`For The Defendants Jiaxiao Zhang, Esq.
`ZTE entities: McDERMOTT WILL & EMERY LLP
` 18565 Jamboree Road, Suite 250
` Irvine, California 92612
`
` Amol Ajay Parikh, Esq.
` Thomas DaMario, Esq.
` McDERMOTT WILL & EMERY LLP
` 444 West Lake Street, Suite 4000
` Chicago, Illinois 60606
`
`ALSO PRESENT:
`
`
`
`
`
`
`
`Page 3 of 122
`
`
`
` 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
`
`Page 4 of 122
`
`
`
` 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 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
`
`Page 5 of 122
`
`
`
` 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 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
`
`Page 6 of 122
`
`
`
` 7
`
` 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
`
`constant frequency reference signal. The patent explains at
`column 2, lines 29 to 34 that in the 802.11a and 802.11g,
`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 +1 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
`
`Page 7 of 122
`
`
`
` 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 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)
`
`Page 8 of 122
`
`
`
` 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 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
`
`Page 9 of 122
`
`
`
` 10
`
` 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
`
`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 follow 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. For a 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
`
`Page 10 of 122
`
`
`
` 11
`
` 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 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
`
`Page 11 of 122
`
`
`
` 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 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
`
`Page 12 of 122
`
`
`
` 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.
`
`Page 13 of 122
`
`
`
` 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
`
`Page 14 of 122
`
`
`
` 15
`
` 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 orthogonal in the context of the '842. The '842 is
`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
`OFDM. They need to be spaced at certain regular spacing so
`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 802.11a, there are 52 subcarriers. They range
`from -26 to +26. In 802.11n, 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 802.11a and 802.11n.
`The patentee did not invent subcarriers. They were
`present. They were simply not used before. In fact, none of
`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 know this is a busy slide. I just want to point out a couple
`of things. In 1999, 802.11a was introduced, using 20 megahertz
`of bandwidth channel. It was based on OFDM. It's max data
`rate was 54 megabits per second. Ten years later in 2009,
`802.11n was introduced. It also has an option or capability to
`
`Page 15 of 122
`
`
`
` 16
`
` 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
`
`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
`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, 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 802.11a training
`sequence. Again, 64 subcarriers already there, existed the
`entire time. Only 52 were active and it has a -26 to a 26 with
`dc with a zero index not being used. This training sequence
`was already defined in 802.11.
`
`Page 16 of 122
`
`
`
` 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
`
`The sequence itself is on the bottom. 53 subcarriers
`is OFDM training symbol, modulated by a sequence of L. Those L
`sequence values are all +1 or -1s 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
`always want? We want better, faster, cheaper. 802.11n
`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 802.11a.
`802.11n, 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.
`
`Page 17 of 122
`
`
`
` 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 specification disclosed the exact same training sequence as
`specified in the eventual standard. Again, you were required
`to start with 802.11a. They didn't invent the entire sequence.
`It was already there for them. Here it is again, the 802.11a
`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
`sequence, the one in frequency domain, it consists of only +1s
`and -1s. 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 it or multiply
`it by itself, 1 x 1, 1, -1 x -1, 1. So it's the peak in the
`average in the frequency domain where a sequence is defined by
`
`Page 18 of 122
`
`
`
` 19
`
` 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
`
`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
`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 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
`-1s, there are only 16 possibilities the patentee could have
`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 +1 to a -1 or take a -1 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
`'842 patent and, again, we get to the right a power domain
`sequence. And you'll notice the peak-to-average power ratio
`
`Page 19 of 122
`
`
`
` 20
`
` 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
`
`from 8011a 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.
`
`Page 20 of 122
`
`
`
` 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 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
`propose: A circuit and/or software that performs a defined
`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
`
`Page 21 of 122
`
`
`
` 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 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
`
`Page 22 of 122
`
`
`
` 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
`
`Fourier Transformer, the one that takes frequency domain
`signals into time domain. BNR will be talking about this
`
Accessing this document will incur an additional charge of $.
After purchase, you can access this document again without charge.
Accept $ Charge
Still Working On It
This document is taking longer than usual to download. This can happen if we need to contact the court directly to obtain the document and their servers are running slowly.
Give it another minute or two to complete, and then try the refresh button.
A few More Minutes ... Still Working
It can take up to 5 minutes for us to download a document if the court servers are running slowly.
Thank you for your continued patience.
This document could not be displayed.
We could not find this document within its docket. Please go back to the docket page and check the link. If that does not work, go back to the docket and refresh it to pull the newest information.
Your account does not support viewing this document.
You need a Paid Account to view this document. Click here to change your account type.
Your account does not support viewing this document.
Set your membership
status to view this document.
With a Docket Alarm membership, you'll
get a whole lot more, including:
- Up-to-date information for this case.
- Email alerts whenever there is an update.
- Full text search for other cases.
- Get email alerts whenever a new case matches your search.
One Moment Please
The filing “” is large (MB) and is being downloaded.
Please refresh this page in a few minutes to see if the filing has been downloaded. The filing will also be emailed to you when the download completes.
Your document is on its way!
If you do not receive the document in five minutes, contact support at support@docketalarm.com.
Sealed Document
We are unable to display this document, it may be under a court ordered seal.
If you have proper credentials to access the file, you may proceed directly to the court's system using your government issued username and password.
Access Government Site
