[00:00:00] Speaker 04: Okay the next case is number 16, 1955. [00:00:29] Speaker 04: Intellectual Ventures II against LLC, against Erickson, Incorporated. [00:00:35] Speaker 04: Mr. Jaffer. [00:00:36] Speaker 00: Judge Newman, may it please the court. [00:00:39] Speaker 00: I thought it would be useful to first step back and explain how the pilot and training symbols are actually used, which might have been lost in the briefing. [00:00:48] Speaker 00: This might provide some context for why they're very different in the 127 patent. [00:00:52] Speaker 00: So the communication channel between the receiver and the transmitter [00:00:56] Speaker 00: is affected by noise, and so it changes really quickly. [00:01:00] Speaker 00: And so you frequently need to fine-tune the synchronization between the receiver and the transmitter. [00:01:06] Speaker 00: And that's accomplished by using these short pilot symbols in the data that do quick little course corrections. [00:01:13] Speaker 02: The training blocks, on the other hand... But that could still be done if it's put in between the blocks rather than within a block, right? [00:01:21] Speaker 00: Well, the difference, Your Honor, is that the training blocks are... No, but answer my question. [00:01:27] Speaker 02: Your honor, that is not how... That function can still... Well, maybe not, but the function you described can still be performed if the pilot symbols are put in between the data blocks, right? [00:01:39] Speaker 00: If the pilot symbols are placed between the data blocks, there can be cross correction, your honor, but there would be a problem. [00:01:45] Speaker 00: And the problem is the following. [00:01:47] Speaker 00: If you were to put the pilot symbols between the data blocks, the only way to do that, given the way the system works, [00:01:55] Speaker 00: is to have a string of pilot symbols together. [00:01:59] Speaker 00: The string of pilot symbols, and the reason you need to do this in order to have them between the data set blocks, is because of the way the Fourier transform works in this invention. [00:02:08] Speaker 00: So what you have originally is you have the data, which is represented in frequency space. [00:02:14] Speaker 00: So the y-axis would be the data, and the x-axis would be the frequency. [00:02:18] Speaker 00: So what's the point? [00:02:19] Speaker 00: And so the point being is if you add two things together in the frequency domain, [00:02:24] Speaker 00: and then take the inverse Fourier transform. [00:02:26] Speaker 00: The mathematical result of that is a convolution, so they get mixed together. [00:02:31] Speaker 02: So where does that say that in the record? [00:02:32] Speaker 00: So that is said in the record. [00:02:34] Speaker 00: It's shown on column seven. [00:02:37] Speaker 00: It explains how the pilot symbols are added. [00:02:42] Speaker 03: What lines are you looking at in column seven? [00:02:44] Speaker 00: So it's explaining figure two, which is talking about the pilot training symbol in server 46. [00:02:53] Speaker 00: And it says that the pilot training symbol asserted on column 7, lines 23 through 25, the pilot training symbol assertor typically provide pilot blocks and training blocks that are asserted into the data blocks by the adders 44. [00:03:11] Speaker 00: So what it's doing there is it's taking the pilot, they're making pilot symbols, and they're adding them to the data blocks using this adder in frequency space. [00:03:22] Speaker 00: What later happens, and you'll see on column 8 at the top, that the modulator now includes an inverse discrete Fourier transform. [00:03:30] Speaker 02: How is this telling me that it can't, that pilot symbols can't be inserted into individual data box? [00:03:37] Speaker 02: As between individual data box? [00:03:42] Speaker 00: The reason it's telling you that, Your Honor, is because the only way that you would be able to insert it in between is if you had a continuous string of pilot symbols. [00:03:52] Speaker 00: If you had that, that would be totally inefficient because you'd be doing this incredibly long amount of course correction, a huge amount of data being used to do this small course correction. [00:04:02] Speaker 02: Who said that? [00:04:04] Speaker 02: Where's the evidence in the record that that's true? [00:04:06] Speaker 02: You're saying that you'd have to have multiple pilot symbols if you put them in between the data blocks, and that would be inefficient. [00:04:14] Speaker 02: Where does the record tell us that? [00:04:16] Speaker 00: The way that this is described, Your Honor, in the record by IB's [00:04:22] Speaker 00: expert is can be found on appendix 1352, which is describing how the pilot symbols are inserted into the data structure. [00:04:38] Speaker 00: But where does he say that would be inefficient? [00:04:42] Speaker 00: He says that about efficiency on 1338, he talks about the efficiency, the [00:04:51] Speaker 00: At the time of the invention on A1338, at the time of the invention, the excess amount of transmission time required for training reduced the overall efficiency of communication. [00:05:03] Speaker 00: This is paragraph 26, Your Honor. [00:05:06] Speaker 00: This excess stemmed from the provision of different training sequences for different purposes. [00:05:11] Speaker 02: But that's not saying that you need multiple pilot symbols if you put them in between the blocks and that's inefficient. [00:05:18] Speaker 02: Where does somebody say that? [00:05:21] Speaker 00: place that someone says that it would be inefficient to place multiple symbols together? [00:05:26] Speaker 02: Well, first, that you'd have to put multiple pilot symbols in, and second, that that would be inefficient. [00:05:31] Speaker 02: Where is that? [00:05:31] Speaker 00: Well, I believe that's been agreed to by the parties. [00:05:34] Speaker 00: If you look on their brief, the red brief, page 20, you will see a diagram explaining this. [00:05:44] Speaker 02: But no, I'm just a... No, I'm saying that's why. [00:05:47] Speaker 02: I can't read a diagram. [00:05:48] Speaker 02: I want to see [00:05:49] Speaker 02: Testimony, where is there testimony saying what you've been saying? [00:05:54] Speaker 00: The testimony that I've been saying appears on 1352 talking about the insertion of the pilot blocks. [00:06:06] Speaker 00: The description by the expert Dr. Hartog describing the differences between pilot symbols and training symbols that I've been discussing right now [00:06:17] Speaker 02: I understand the difference between pilot symbols and training system symbols. [00:06:21] Speaker 02: What I'm asking is your argument is if you put the pilot symbols in between the data blocks, you have to have multiple pilot symbols and that's inefficient. [00:06:32] Speaker 02: Where does somebody say that? [00:06:36] Speaker 00: That is found, Your Honor. [00:06:49] Speaker 00: So that is found in Dr. Herzog's testimony, where he talks about the multiple. [00:06:55] Speaker 00: What page every time? [00:07:08] Speaker 00: I'm getting to your own. [00:07:10] Speaker 00: It's 1084. [00:07:10] Speaker 00: 1084? [00:07:10] Speaker 00: Yes. [00:07:11] Speaker 00: You will see the page 120 of his testimony, line 17. [00:07:26] Speaker 00: And he's asking, question, does 127 pattern describe or disclose anywhere how the pilot symbols are buried or spread within rectangle N? [00:07:38] Speaker 00: Answer, they are spread because they go through the FAT and everything. [00:07:41] Speaker 00: And every number, every value that occupies just one frequency spot has changes the value at every single one of the time spots input. [00:07:49] Speaker 00: And then he describes the frequency, how that all works so that they all get mixed together. [00:07:53] Speaker 02: But that's not saying what you're saying. [00:07:56] Speaker 02: Is that the best you have? [00:07:59] Speaker 00: So that is saying how the pilot symbols are mixed in with the data and why that needs to be the case through the FFT. [00:08:08] Speaker 00: What I'm saying is the point of the invention, as shown on, for example, column 3, lines 10 through 12, is this whole point of the patent [00:08:26] Speaker 00: And the discussion of the patent is that there is a problem which reduces system throughput and hence the system efficiency, column three, lines 11 through 13. [00:08:38] Speaker 00: And what I'm saying to you is the description I've just given of how the pilot symbols are mixed in with the data symbols, that occurs due to how the Fourier transform works. [00:08:52] Speaker 00: And you want to not have a large string of training symbols [00:08:57] Speaker 00: every time you want a course correct, there's a separate thing that does that. [00:09:01] Speaker 00: It's the training block, which is the Galacti block. [00:09:03] Speaker 00: We're not talking about training symbols. [00:09:04] Speaker 00: We're talking about pilot symbols. [00:09:07] Speaker 00: Right. [00:09:08] Speaker 00: And that's very important. [00:09:09] Speaker 00: They're very distinct, how those two things work. [00:09:13] Speaker 00: The training blocks are allowed to be a long string of continuous symbols, as the patent says. [00:09:19] Speaker 00: And the reason why they're allowed to be long is because they only work once. [00:09:22] Speaker 00: Whereas if you have a pilot symbol and you want to be quick. [00:09:25] Speaker 02: I understand. [00:09:27] Speaker 02: What I'm puzzled as to where this patent tells you that you can't have the pilot symbols between the data box. [00:09:38] Speaker 00: The reason, Your Honor, is twofold. [00:09:40] Speaker 00: The first is the way the claim is written. [00:09:42] Speaker 00: The claim includes two different things. [00:09:45] Speaker 00: One is it talks about pilot symbols and training symbols. [00:09:47] Speaker 00: And you can't conflate the two. [00:09:49] Speaker 00: The second way is that the claim talks about how you insert pilot symbols into data blocks. [00:09:56] Speaker 00: And that aligns with the natural language of putting many things into many other things. [00:10:01] Speaker 03: If I were to tell you... What about the fact that the specification says that the pilot symbols can be inserted anywhere in the data blocks? [00:10:11] Speaker 03: I mean, that's a quote from the specification. [00:10:13] Speaker 03: That's right. [00:10:14] Speaker 00: And they can be inserted anywhere in the data blocks in frequency space, and then they're Fourier transformed. [00:10:20] Speaker 03: But I think the question would be, because it says data blocks, [00:10:24] Speaker 03: Why would it be between data blocks or maybe at the beginning of a block or at the end of a block, which would consequently be between blocks? [00:10:32] Speaker 03: And so I think that the district hurts construction included not just inserting them within a given block, but also at the beginning or end, or if you say, it would say between data blocks. [00:10:44] Speaker 00: Right. [00:10:44] Speaker 00: But let me go back to the natural language. [00:10:46] Speaker 00: If I'm going to say, I want to insert many things into many other things. [00:10:50] Speaker 00: For example, I want to say, [00:10:51] Speaker 00: I'm going to put people into hotel rooms. [00:10:53] Speaker 03: Let's say into, because that's what the claim says. [00:10:56] Speaker 00: Let's say I'm going to put people into hotel rooms. [00:10:59] Speaker 00: That means that I have any number of people into any number of hotel rooms, many different, maybe one hotel room, maybe many. [00:11:09] Speaker 00: What I would not understand that to mean is that you put people into the corridor between the hotel rooms. [00:11:14] Speaker 00: That's not what the language says. [00:11:17] Speaker 00: And furthermore, the language, we could have said other things. [00:11:19] Speaker 00: We could have talked about, for example, would be claimed 20, where it actually talks about putting things between different blocks. [00:11:29] Speaker 02: Or you could have also said put it within an individual block, and it doesn't say that. [00:11:34] Speaker 00: You could have, but the problem, of course, is that we don't necessarily want it in just one block. [00:11:38] Speaker 00: We want it in multiple different blocks. [00:11:40] Speaker 00: And we might want different numbers in each of the multiple different blocks. [00:11:43] Speaker 00: And so the way to express that [00:11:45] Speaker 00: would be exactly what we said. [00:11:47] Speaker 00: We could have said, for example, another example, could we insert the pilot symbols into data structures? [00:11:55] Speaker 00: Then you don't know where in the structure it could be. [00:11:56] Speaker 00: It could be between the bit of blocks. [00:11:58] Speaker 00: It could be at the ending. [00:11:58] Speaker 00: It could be at the end. [00:11:59] Speaker 00: It didn't say that. [00:12:00] Speaker 00: It said, into the data blocks. [00:12:03] Speaker 00: And so therefore, because of the way that, that's the best way to express exactly what our construction would have. [00:12:10] Speaker 00: The board below, [00:12:12] Speaker 00: The sole reason the board came to a contrary position was simply based upon an assumption that inserted into must mean you could put it anywhere in between the blocks or in the blocks or any other place. [00:12:24] Speaker 00: But if you're going to do that, then you're going to run into the problem with the Fourier transform that I was talking about earlier, where you need a bunch of pilot symbols together. [00:12:33] Speaker 00: And if you do that, then you essentially have what's shown to be a training symbol, a training block. [00:12:39] Speaker 00: The training block is a bunch of pilot symbols all jammed together. [00:12:42] Speaker 00: And then that's where you transform it into a block. [00:12:46] Speaker 00: That's very different than just saying pilot symbols are being inserted into the data box. [00:12:51] Speaker 03: Am I correct in understanding that this particular element of the claim was in the prior art? [00:12:58] Speaker 03: Because I see in column two, which is describing background of the invention, it says instead of being placed as a prefix to the data structure, the pilot structures are periodically arranged within groups of data symbols. [00:13:12] Speaker 03: That's at column two, lines 19 through 22. [00:13:16] Speaker 00: Right, Your Honor. [00:13:16] Speaker 00: And that is describing in that system right there a system that exists solely in time space. [00:13:22] Speaker 00: There's no frequency dimension. [00:13:23] Speaker 00: And you can tell that by looking at column two, lines 30, where it starts talking about frequency domain. [00:13:29] Speaker 00: So beforehand, what you had, the problem was you'd stick your pilot symbols, pilot what they call pilot structures, in order to provide some sort of synchronization. [00:13:37] Speaker 00: People knew how to do synchronization before this patent. [00:13:40] Speaker 00: But because you were only acting in time domain, you weren't able to combine in the frequency domain, you'd have to do this. [00:13:48] Speaker 00: You'd have to stick these pilot structures between the data blocks. [00:13:53] Speaker 03: But it says arranged within. [00:13:55] Speaker 03: It uses the language arranged within. [00:13:57] Speaker 03: You're saying that means between? [00:14:00] Speaker 00: Well, it says that they are [00:14:05] Speaker 00: They're periodically regranged within groups of, which is different than inserted into. [00:14:11] Speaker 00: So they're placed within groups of the data blocks. [00:14:15] Speaker 00: That's how the old systems work. [00:14:17] Speaker 00: And that's how the prior art that the board relied upon worked. [00:14:20] Speaker 00: But what is shown here, especially in column seven, is the addition of the pilot symbols and frequency domain that then gets transformed. [00:14:27] Speaker 00: And so the data and the pilot symbols get mixed together. [00:14:31] Speaker 03: OK. [00:14:31] Speaker 03: I have another question for you, which is in column seven. [00:14:34] Speaker 00: Yep. [00:14:34] Speaker 03: line 24 or 23, it says that pilot blocks and training blocks are inserted into, paren, or combined with the data blocks. [00:14:47] Speaker 03: So is that equating asserted into and combined with? [00:14:50] Speaker 03: Do those mean the same thing? [00:14:51] Speaker 00: I don't believe so. [00:14:52] Speaker 00: And you can see that same language reflected in claim one, where you also see that you have inserting the pilot symbols into the data blocks. [00:15:02] Speaker 00: And then you also combine the data blocks [00:15:04] Speaker 00: data symbols with the data blocks. [00:15:06] Speaker 00: And so that is yet again another, because they're using different words, the presumption is they're going to mean different things. [00:15:13] Speaker 00: In fact, they do mean different things. [00:15:14] Speaker 00: One is you took the pile symbol and you use this Fourier transform to insert it into. [00:15:19] Speaker 00: The other is that you actually took the training block and jammed it to the beginning of the entire data structure, which is shown in Figure 6, for example. [00:15:31] Speaker 04: OK. [00:15:31] Speaker 04: Well, let's hear from the other side. [00:15:33] Speaker 04: We will save you some rebuttal time. [00:15:35] Speaker 04: Thank you. [00:15:45] Speaker 04: Mr. Lowes. [00:15:46] Speaker 01: Good morning. [00:15:47] Speaker 01: Andrew Lowes for Appellee Erickson. [00:15:51] Speaker 01: Judge Stoll, I think you got it right when you said, aren't we talking about the pilot symbols that are in the prior art? [00:15:55] Speaker 01: That's exactly what we're talking about. [00:15:57] Speaker 01: Those pilot symbols exist in the time domain, which is what the counselor said. [00:16:01] Speaker 01: And that's exactly the priority that the board applied. [00:16:04] Speaker 01: First, this patent is really about training symbols and the enhancement to those training symbols. [00:16:09] Speaker 01: And that's what the patent describes as the novelty or the new invention. [00:16:13] Speaker 01: And the board found that was disclosed by the Schmidl reference. [00:16:16] Speaker 01: And then we put forward the Arsalan reference showing that pilots were known to insert in the data structure in the time domain. [00:16:23] Speaker 01: And then we looked to the claim. [00:16:25] Speaker 01: And the claim says, insert pilot symbols into data blocks. [00:16:30] Speaker 01: And that's understood [00:16:31] Speaker 01: I think the board got it correct that it's understood to be within or between one or more data blocks. [00:16:38] Speaker 01: And that's a common usage. [00:16:40] Speaker 01: Council use the analogy of hotel rooms and that you get put into that. [00:16:44] Speaker 01: Other analogies exist as well. [00:16:46] Speaker 01: If you think of kids at a playground, you've got the ball pits full of balls. [00:16:50] Speaker 01: The kids jump into those balls. [00:16:52] Speaker 01: Are they jumping inside of those balls? [00:16:53] Speaker 01: No. [00:16:54] Speaker 01: They're jumping into a grouping. [00:16:56] Speaker 01: The same if you break up a pile of leaves and the kids jump into those leaves. [00:16:59] Speaker 01: They're going into the grouping. [00:17:01] Speaker 01: And that's the way this claim term is understood. [00:17:04] Speaker 01: And so when you put the pilots in, in the frequency domain, into the grouping of the data blocks, but then as it's transformed, you're going to end up with pilot blocks and ultimately pilot symbols in the time domain. [00:17:19] Speaker 01: And if we look to what's in the specification, what's disclosed, look at figure six, the description of that says that the pilot symbols are omitted for simplicity. [00:17:29] Speaker 01: But that is a time domain diagram. [00:17:31] Speaker 01: All the experts agree that's time domain. [00:17:34] Speaker 01: And these pilot symbols have been omitted. [00:17:36] Speaker 01: And Dr. Haas, I think on page 20 of our brief, Dr. Haas put forward the diagram that Council was mentioning, which is, what would it look like if you put those pilot symbols in? [00:17:46] Speaker 01: And he disclosed that and described this is exactly how a person with skill in the art would understand if they hadn't been omitted for simplicity. [00:17:54] Speaker 01: And that's also consistent with the specification that says, [00:18:00] Speaker 01: on column seven, line 42, that the pilot blocks themselves are typically transmitted with data blocks to calibrate the system. [00:18:13] Speaker 01: Well, under their construction, the pilot blocks disappear in the time domain. [00:18:18] Speaker 01: And the only transmission is in the time domain. [00:18:20] Speaker 01: So pilot blocks must exist in the time domain to be transmitted. [00:18:24] Speaker 01: And that's exactly our position. [00:18:26] Speaker 01: They exist in the time domain. [00:18:28] Speaker 01: And you look to the background, which we were talking about. [00:18:33] Speaker 01: And there, again, we're talking about what's in the time domain, which is what we're both talking about. [00:18:38] Speaker 01: Pilots exist in the time domain. [00:18:39] Speaker 01: That's what was disclosed. [00:18:41] Speaker 01: And there's nothing in the specification that says, no, no, no, we're totally changing that kind of prior art. [00:18:46] Speaker 01: And we're going to only put in these pilot symbols. [00:18:50] Speaker 01: And when they get transformed, they're going to disappear into data blocks. [00:18:53] Speaker 03: You're saying that maybe the specification doesn't explain why you would want to have it [00:18:57] Speaker 03: in the data structures as opposed to between the data blocks, as opposed to between the data blocks. [00:19:03] Speaker 01: That's absolutely right, Your Honor. [00:19:05] Speaker 01: And really, what we're doing here is they're trying to create an embodiment and then read limitations from that embodiment into the claims. [00:19:12] Speaker 01: That embodiment doesn't exist where the pilot symbol is inserted inside of a data block. [00:19:21] Speaker 01: So the priority of the patent shows that the pilot symbols exist in the time domain. [00:19:28] Speaker 01: And that's exactly what the expert looking at the disclosure illustrated. [00:19:36] Speaker 01: And it's consistent with the figure six. [00:19:38] Speaker 01: Figure six says in the time domain, pilot symbols would be there if they're just omitted for simplicity. [00:19:43] Speaker 01: It's also consistent with the language that pilot blocks themselves are typically transmitted with data blocks. [00:19:51] Speaker 01: Again, the only time that transmissions occur are in the time domain. [00:19:55] Speaker 01: So they must exist in the time domain. [00:19:58] Speaker 01: There is a third area that we addressed in the briefing, and that is that Pat D chose to incorporate by reference one of their earlier patent applications. [00:20:08] Speaker 01: And there again, that's consistent with our understanding and the board's construction of this term. [00:20:13] Speaker 02: I don't understand that. [00:20:15] Speaker 02: I looked at the provisional application. [00:20:17] Speaker 02: I don't see where that gets you. [00:20:20] Speaker 01: Well, Your Honor, when it talks about transmission of a known OFDM symbol, that is referring to the time domain. [00:20:27] Speaker 02: But it doesn't tell you anything about whether the symbol is within an individual block or between blocks, right? [00:20:34] Speaker 01: Well, Your Honor, it does in the terms of when they are placed between blocks in the frequency domain, when pilot symbols are between blocks in the frequency domain, when they get converted into the time domain. [00:20:45] Speaker 02: But answer my question. [00:20:46] Speaker 02: Where does the provisional tell us anything about that? [00:20:49] Speaker 01: It doesn't expressly tell us about that. [00:20:51] Speaker 01: That's the problem. [00:20:52] Speaker 01: Right. [00:20:52] Speaker 01: But, Your Honor, understood by a person of skill in the art, as Dr. Haas testified, [00:20:57] Speaker 01: it leads one to know that that's how they're created. [00:20:59] Speaker 01: They're first generated in the frequency domain and then converted into the time domain so that there's a known OFDM symbol. [00:21:06] Speaker 01: If there was data within that symbol, then it would no longer be known. [00:21:11] Speaker 01: So it has to be just pilots that have come through into the time domain. [00:21:22] Speaker 01: The other concept that council raised was that [00:21:27] Speaker 01: Perhaps they could have changed their claims and addressed it by instead of saying insert into data blocks, a group of data blocks, they could have said a data structure. [00:21:36] Speaker 01: However, in the patent, in the claim, data structure is in the time domain. [00:21:41] Speaker 01: It's not in the frequency domain. [00:21:43] Speaker 01: So if you look at the claim when it talks about the output of the modulator, they're talking about that data structure being in the time domain. [00:21:50] Speaker 01: So again, some of this is a little bit confusing. [00:21:55] Speaker 01: Everything starts in the frequency domain, it's put together, run through the modulator, and it comes out in the time domain. [00:22:01] Speaker 01: And then under our construction, the pilot symbols still exist in the time domain, just like in the prior arc, just like in the provisional, as well as described in the rest of the spec. [00:22:12] Speaker 01: Under their construction, the pilot symbols cease to exist in the time domain because they're all mixed up with data structures. [00:22:23] Speaker 01: The other aspect they raised was the inefficiency. [00:22:27] Speaker 01: Again, I think you're correct. [00:22:29] Speaker 01: That's nowhere in the specification. [00:22:31] Speaker 01: And in fact, it's contrary to the background, which uses exactly the structure we're talking about, which are time domain pilots, as well as it's contrary to the provisional, which uses the time domain OFDM pilot symbol, which is, again, it's a full pilot symbol. [00:22:49] Speaker 01: It's not a mix of data and pilots when it comes through the time domain. [00:23:07] Speaker 01: Another aspect that we've looked at with respect to the background is that, let me point to that, it's in column two, column two, line 17, that pilot symbols, they have the same structure as training symbols. [00:23:31] Speaker 01: And again, even in figure six, which is the embodiments [00:23:36] Speaker 01: description of the time domain, those training symbols appear in the time domain. [00:23:41] Speaker 01: And so if you have the same structure of the training symbols and they appear in the time domain, the same would happen for the pilot symbols. [00:23:48] Speaker 01: They would also appear in the time domain. [00:23:51] Speaker 01: So all of these things, the specification itself, describing the embodiments, the background, all of that supports that the pilots inserted into the groups of data blocks [00:24:03] Speaker 01: can be between those data blocks and ultimately result in pilot symbols in the time domain, just as the prior shows. [00:24:13] Speaker 01: Are there any questions? [00:24:16] Speaker 04: Any questions? [00:24:25] Speaker 00: Good. [00:24:27] Speaker 00: Thank you, Council. [00:24:31] Speaker 00: The first point is, [00:24:33] Speaker 00: Everyone acknowledges that pilot symbols can actually live within a data block. [00:24:39] Speaker 00: And their argument is solely they could also live between the data blocks. [00:24:44] Speaker 00: That makes the natural language very different than saying children can jump into a pile of leaves. [00:24:50] Speaker 00: No child jumps into an individual leaf. [00:24:54] Speaker 00: So that the natural language knowledge doesn't work as much. [00:24:58] Speaker 02: The president said a child jumps into leaves. [00:25:02] Speaker 02: You wouldn't think it goes into the leaf, right? [00:25:05] Speaker 00: Well, that's exactly the point. [00:25:07] Speaker 00: But the invention and both parties admit that the pilot symbol can actually live inside the data block. [00:25:16] Speaker 00: Now, they say it could also live between them, but they admit that it could live inside the block. [00:25:20] Speaker 00: So that's why it's very different. [00:25:22] Speaker 00: You can't have a child jump into a leaf. [00:25:27] Speaker 03: The second point goes to... I think the problem we might have is the specification nowhere says why it would be advantageous, why anyone would want to have it within the data blocks, as opposed to... I mean, within a particular data block, as opposed to into the data blocks, which might have multiple, which would include between data blocks. [00:25:48] Speaker 03: I think that's the problem. [00:25:50] Speaker 00: I understand the problem the court's having. [00:25:52] Speaker 00: Again, the thing that I go back to is the... [00:25:57] Speaker 00: point being is that if you're going to take just the way you're going to add a pilot block and a data symbol together and then add them together in the Fourier Transform, that's the natural, the thing that's going to result is that they're all mixed together. [00:26:10] Speaker 00: So the point being is, you're right, there's not a, and there's a good reason for that because of inefficiency, you don't want a large string of pilot symbols. [00:26:18] Speaker 02: Except you don't have any testimony that says that. [00:26:21] Speaker 00: Right, all I have is testimony explaining why it is the difference between [00:26:26] Speaker 00: a training symbol and a pilot symbol, which are quite different, and we need to keep that line separate. [00:26:32] Speaker 00: And the training symbols are a continuous set, whereas a pilot symbol is not. [00:26:37] Speaker 00: And so that difference, it can be explained through the efficiency that the entire patent is trying to achieve. [00:26:45] Speaker 00: Finally, there's the figure, that figure six, that they quite often rely upon, where they look at column 11 and say, [00:26:57] Speaker 00: at lines 43 for 47. [00:26:59] Speaker 00: It says, although admitted from figure 6 for simplicity, pilot symbols may also be intermittently inserted into the data symbols 80 by the pilot training inserter. [00:27:10] Speaker 00: So they keep saying, well, that shows that the simplicity just means they didn't want to draw two vertical lines together to show in between. [00:27:16] Speaker 00: But look what it says. [00:27:18] Speaker 00: It says that the pilot symbols are inserted into data symbol 80. [00:27:24] Speaker 00: The data symbol is, there's only one, there's a data symbol that's labeled 80. [00:27:28] Speaker 00: They could have said it's inserted into data structure 72. [00:27:33] Speaker 00: They didn't say that. [00:27:34] Speaker 00: They said it's inserted into data symbol 80. [00:27:36] Speaker 03: But nobody's disputing that the claim includes inserting it into an individual data symbol. [00:27:45] Speaker 03: It's the question is whether the language is broad enough to also include inserting it between data symbols, right? [00:27:51] Speaker 00: Right, and that's going back to the natural language point of people inside hotel rooms, whereas no person could be inside a leaf. [00:27:58] Speaker 00: That doesn't make any sense. [00:28:00] Speaker 00: And similarly, the point of the invention was to get the efficiency that I've been discussing. [00:28:08] Speaker 03: The difference is that in the prior art, if the prior art was inserted between data symbols, [00:28:13] Speaker 03: I mean, it's very different than a hypothetical about hotel rooms, because nobody says that anyone's going to sleep in the hallway when you're sleeping in hotel rooms. [00:28:21] Speaker 03: But the background of this patent is that the pilot symbols in the prior art were inserted between the data structures. [00:28:30] Speaker 00: Right. [00:28:30] Speaker 00: And the reason they had to do that is because they didn't have the ability to insert in frequency space. [00:28:35] Speaker 00: This was all in the time space. [00:28:37] Speaker 00: And similarly, the prior art was in the time space. [00:28:39] Speaker 00: But given now that you can do it in frequency space and add them that way, [00:28:43] Speaker 00: then the thing that's going to happen is the mixture. [00:28:46] Speaker 00: Thank you. [00:28:46] Speaker 04: Thank you. [00:28:47] Speaker 04: Thank you both. [00:28:48] Speaker 04: The case is taken on your submission.