[00:00:07] Speaker 02: Okay, Mr. Grant. [00:00:10] Speaker 00: Good morning, Your Honors. [00:00:11] Speaker 00: May it please the Court. [00:00:13] Speaker 00: The Court should reverse the Board's final written decisions for three reasons. [00:00:17] Speaker 00: First, the Court should reverse the Board's final written decisions as to claims 1 through 40 because the Board erred in construing the term groups of symbols and accordingly erred in finding that Wallace discloses groups of symbols. [00:00:32] Speaker 00: Second, the court should reverse the board's final written decision as to claims 1 through 44, which is all claims of the 569 patent, because the board erred in finding that Wallace discloses a time frequency domain that has a first part and a second part. [00:00:47] Speaker 00: third and finally the court should reverse the board's final written decisions as to claims 7 17 27 and 37 because Wallace does not disclose at least a part of the encoded second beta is not continuously mapped to each symbol within each of the groups of symbols I'll start with what was the third in [00:01:12] Speaker 01: domain issue or is that the non-continuous mapping? [00:01:16] Speaker 00: The third issue is the non-continuous mapping issue, Your Honor. [00:01:21] Speaker 00: I'll start with the first reason, which is the construction of groups of symbols. [00:01:26] Speaker 00: So the board erred by construing groups of symbols such that the groups of symbols do not need to be defined prior to mapping. [00:01:34] Speaker 00: That construction is erroneous because it is inconsistent with the claim language and is inconsistent with the specification. [00:01:42] Speaker 02: I don't understand why you can't have mapping to groups of symbols if it's dynamic and iterative, particularly when the patent is directed to having feedback. [00:01:55] Speaker 02: frequency conditions, et cetera. [00:01:59] Speaker 02: So why does the map we have to be doing a predetermined group of symbols? [00:02:04] Speaker 02: Why can't the group of symbols be determined as part of the dynamic process of mapping? [00:02:10] Speaker 00: Well, there's two reasons, Your Honor. [00:02:12] Speaker 00: First, the claims reside both mapped to symbols and mapped to groups of symbols. [00:02:17] Speaker 00: And so under the case law, those two terms are presumed to mean something different. [00:02:22] Speaker 00: To make them mean something different, the lists of data need to be [00:02:25] Speaker 00: need to exist before mapping. [00:02:27] Speaker 02: That doesn't seem very convincing. [00:02:28] Speaker 00: What's the other one? [00:02:30] Speaker 00: The other one is the specification. [00:02:32] Speaker 00: If you look, it has four embodiments. [00:02:34] Speaker 00: In every single one of those embodiments, the groups of data are formed prior to mapping. [00:02:39] Speaker 00: So what happens in three of these embodiments? [00:02:42] Speaker 02: Well, that happens all the time. [00:02:46] Speaker 00: In this case your honor I don't think it's broader than the specification because I think you have the claim language itself that supports that groups of data need to be defined prior to mapping. [00:02:57] Speaker 02: Why is the claim language inconsistent with dynamic mapping? [00:03:03] Speaker 00: Well one for the the reason that we need to differentiate between mapped to symbols and mapped to groups of symbols. [00:03:09] Speaker 00: But second, I think the claim language maps to groups of symbols presumes that the groups of symbols already exist. [00:03:15] Speaker 00: You can't map to something that does not exist. [00:03:18] Speaker 00: For example, an example we gave in our brief, you can't assign people to teams if those teams do not exist. [00:03:25] Speaker 00: For the same reason, you cannot map to groups of symbols that do not already exist. [00:03:30] Speaker 02: Yeah, you could have a situation where teams are selected as part of a dynamic process. [00:03:35] Speaker 02: They don't have to pre-exist. [00:03:38] Speaker 02: I think in that... Kids divide themselves into two teams. [00:03:44] Speaker 02: It's a dynamic process. [00:03:46] Speaker 00: But in that case, Your Honor, you just said divide yourselves into two teams. [00:03:50] Speaker 00: So in that case, the teams do already exist. [00:03:52] Speaker 00: They may not have kids on the teams yet. [00:03:54] Speaker 02: No, the teams are determined by the process of selecting, and it's a dynamic process. [00:03:59] Speaker 00: I think in that case, the kids on the teams are determined by this dynamic process. [00:04:05] Speaker 00: But the existence of the teams is determined before the kids are assigned to the teams. [00:04:15] Speaker 00: But I think it is even true under your hypothetical, Your Honor. [00:04:19] Speaker 00: If you say divide kids into teams, by saying divide yourselves into teams, you are assuming that these teams exist. [00:04:26] Speaker 00: They don't have kids on them yet, but the teams exist. [00:04:28] Speaker 00: And I think that's what the 569 patent describes. [00:04:31] Speaker 00: And even if there is some ambiguity in this language, Your Honor, I think the specification fully supports that the groups of symbols need to be defined prior to mapping. [00:04:43] Speaker 00: And I do think the difference between mapped assembles and mapped groups of symbols is a strong reason that we need to construe groups of data to mean. [00:04:53] Speaker 01: Under Judge Dyke's example, if you're going to have, you're going to require groups to be defined before mapping. [00:05:02] Speaker 01: Why isn't, you can't say divide yourselves into team A and team B. Is that any different than just saying divide yourself into teams? [00:05:13] Speaker 01: Is there further defining of the groups when you do that? [00:05:20] Speaker 00: I don't think so, your honor. [00:05:22] Speaker 00: I think those would be the same. [00:05:25] Speaker 00: I think I would call that assigning two teams. [00:05:29] Speaker 00: That's part of the process of assigning two teams, whereas the creation of the teams comes before that. [00:05:35] Speaker 00: So for example, in the 569 patent, [00:05:39] Speaker 00: Deciding that you're going to map data one to these groups of symbols that's part of mapping But the groups of symbols to fall within the scope of the five six nine patent would need to be determined Determined in advance. [00:05:51] Speaker 00: That's how the the specification describes the five six nine patent every single embodiment describes it that way and I think that's consistent with the claim language and I think it's it's [00:06:03] Speaker 00: I don't know if it's undisputed, but Wallace does not disclose groups of data that are created prior to mapping. [00:06:12] Speaker 00: Ballway below and the board relied on Wallace figure 2. [00:06:16] Speaker 00: All Wallace figure 2 shows is data that has already been mapped to a time frequency domain. [00:06:23] Speaker 00: It doesn't show forming any groups of symbols prior to mapping to the time frequency domain. [00:06:28] Speaker 00: The board essentially agreed with that. [00:06:31] Speaker 00: The board said that Wallace does not disclose how figure two came to be. [00:06:36] Speaker 00: The board instead found that Wallace suggests mapping to data to groups of symbols. [00:06:42] Speaker 00: But the board's analysis suffers from a number of errors. [00:06:45] Speaker 00: One, it succumbed to hindsight bias. [00:06:48] Speaker 00: For example, the board explicitly compared Wallace Figure 2 to the figures of the 569 patent, and essentially found because they look similar, then Wallace suggests mapping data to groups of symbols. [00:07:01] Speaker 00: In our view that is classic hindsight bias. [00:07:04] Speaker 00: It is using the teachings of the 569 patent To to interpret the prior art It's viewing Wallace through the lens of the 569 patent instead of determining what would have been obvious at the time of the 569 patent The board also conflated [00:07:22] Speaker 00: grouping data that needs to be transmitted with grouping symbols prior to mapping the data. [00:07:30] Speaker 00: So for example, the board found that because data one through data six are going to get transmitted to different receivers, you would need to group the symbols that need to be transmitted beforehand. [00:07:42] Speaker 00: At best, that disclosure will ball us, I think, [00:07:48] Speaker 00: It doesn't suggest grouping data that needs to be transmitted. [00:07:51] Speaker 00: It doesn't suggest grouping symbols in the time frequency domain. [00:07:55] Speaker 00: So for that reason, it's our position that the court should reverse the board's final written decisions as to claims 1 through 40. [00:08:05] Speaker 00: Even if the court disagrees with that, the court should reverse the board's final written decisions as to claims 1 through 44, because Wallace does not disclose a first part and a second part of a domain. [00:08:17] Speaker 00: Under the claim language, the time frequency domain needs to be separated or divided into a first part and a second part. [00:08:25] Speaker 00: The claims recite first data and second data that are mapped to a first part of the domain and a second part of the domain. [00:08:33] Speaker 00: If the claims are not construed such that the time frequency domain must be split into a first part and second part, there's no need for the first part of the domain and second part of the domain claim language. [00:08:46] Speaker 00: The claims could instead just talk about the first data and the second data and how that data is mapped in the time frequency domain. [00:08:53] Speaker 00: I think it's undisputed that Wallace Figure 2 does not disclose a time frequency domain that is split into a first part and second part. [00:09:01] Speaker 00: It's a unitary time frequency domain that has data mapped to it. [00:09:06] Speaker 00: Finally, even if the court disagrees with those first two arguments, it should still reverse the board's final written decisions as to claims 7, 17, 27, and 37 because the board erred in finding that Wallace discloses data [00:09:24] Speaker 00: that is not continuously mapped to each symbol within each of the groups of symbols. [00:09:29] Speaker 00: So this claim language is a little bit complex, but the important thing is that it requires data that is not continuously mapped to each symbol within each of the groups of symbols. [00:09:39] Speaker 01: Within what? [00:09:40] Speaker 00: Within each of the groups of symbols. [00:09:43] Speaker 00: The board relied on Wallace's figure two disclosure of data one that is mapped in time slot two and time slot seven and essentially found that's one group of symbols and there's a discontinuity because that group does not include time slots three, four, five, and six. [00:10:04] Speaker 00: Incorrect for two reasons. [00:10:05] Speaker 00: One, incorrectly assumes that time slot two and time slot seven form a single group. [00:10:11] Speaker 00: There is no disclosure whatsoever in Wallace that would support treating those as a single group. [00:10:17] Speaker 00: Even if you could treat that as a single group, there is data in Wallace figure two mapped to every single one of those symbols. [00:10:25] Speaker 00: All of the data one is mapped in time slot two and time slot seven. [00:10:30] Speaker 01: What about the claim language that requires [00:10:33] Speaker 01: that there be some sort of alignment along the frequency index. [00:10:40] Speaker 00: So it does, it requires that the symbols be aligned along the frequency index, which was treated for the board as basically meaning that the mapping, you map in an increasing order on the frequency index. [00:10:57] Speaker 00: So if you have frequency index one, the next mapping would be frequency index two and then frequency index three. [00:11:02] Speaker 00: That's how the parties in the board treat it along that. [00:11:06] Speaker 01: I don't understand your argument. [00:11:07] Speaker 01: Are you saying that we should ignore that limitation? [00:11:11] Speaker 00: No, I'm not I'm not saying you should ignore that limitation that limitation is not really part of our argument Our argument is simply that the data is continuously mapped in the group. [00:11:23] Speaker 00: So if the group is defined as the board has said as Data 1 and time slot 2 and data 1 and time slot 7 There is data mapped continuously within that group. [00:11:34] Speaker 00: None of the symbols are skipped during mapping and [00:11:38] Speaker 00: So if the court has no further questions, I'll reserve the remainder of my time for rebuttals. [00:11:42] Speaker 01: Why do you ignore that part of the claim language? [00:11:45] Speaker 01: It says that if you're going to examine how the data is continuously mapped, it says align along, this is the claim language, the frequency index. [00:11:56] Speaker 01: And correct? [00:11:57] Speaker 01: That's what the claim says. [00:11:58] Speaker 00: That's correct. [00:11:59] Speaker 01: And are you saying that that doesn't matter? [00:12:02] Speaker 01: You just don't look at that? [00:12:05] Speaker 01: You're not addressing that. [00:12:08] Speaker 00: No, we're not saying that doesn't matter, but I think the symbols in Wallace Figure 2 are aligned along the frequency index. [00:12:16] Speaker 00: That's why we did not address it, because the data is mapped in an increasing order along the frequency index. [00:12:22] Speaker 00: That's why our argument focuses on the discontinuous nature and not the aligned along the frequency index. [00:12:29] Speaker 01: So it does matter. [00:12:31] Speaker 00: It does matter to the claim. [00:12:34] Speaker 00: Yes, it does matter. [00:12:35] Speaker 00: It has meaning in the claim, Your Honor. [00:12:36] Speaker 00: We just think that Wallace does disclose that part. [00:12:39] Speaker 00: It does not disclose the not continuously mapped to each symbol within each of the groups of symbols part. [00:12:45] Speaker 01: I don't want to beat a dead horse, but it does matter to the claim, but yet it doesn't matter to the degree where you think it's important to address [00:12:58] Speaker 00: I just think Wallace does disclose a line along the frequency index, so we haven't disputed that. [00:13:03] Speaker 00: I just don't think we have an argument there. [00:13:05] Speaker 00: I do think we have an argument that the second data is not continuously mapped to each symbol within each of the groups of symbols, which is why we focused on that. [00:13:14] Speaker 00: I'll reserve the remainder of my time. [00:13:27] Speaker 03: Good morning. [00:13:27] Speaker 03: May it please the court. [00:13:29] Speaker 03: Mike Tyler from the United States Patent and Trademark Office. [00:13:32] Speaker 03: In the IPR subject to these appeals, the board found all challenged claims unpatentable over the combination of Wallace and Ishikawa. [00:13:41] Speaker 03: Specifically, the board found that Wallace teaches each and every limitation of the claims except for any limitation requiring [00:13:48] Speaker 03: that mapping be done in an increasing order according to either the frequency or time index which Ishikawa teaches. [00:13:56] Speaker 03: Notably, Optus has not appealed the board's finding a motivation of these references or any of the teachings in the Ishikawa. [00:14:04] Speaker 03: As we just heard, their arguments are primarily about three of the teachings of Wallace, one of which they argue is dependent on a claim construction, that being the groups of symbols construction. [00:14:16] Speaker 03: Starting with that, the board appropriately viewed this as a BRI case and looked first at the language of the claims themselves and found that these are very broad claims. [00:14:28] Speaker 03: And in doing so, they noticed that there is nothing in the claims that describe the manner or method in which the mapping must take place. [00:14:38] Speaker 03: And you can kind of tell from the briefing, the party's dispute is whether this concept of predetermination needs to be read into the claims. [00:14:47] Speaker 03: And I will note that the patentee, if we look at defendant claim four, [00:14:51] Speaker 03: knew how to use the word predetermined when writing claim language. [00:14:55] Speaker 03: And if we look at claim 41, which is an independent claim, there is a little additional requirements of how mapping needs to be done. [00:15:05] Speaker 03: But something like claim one, it's completely broad and open-ended. [00:15:08] Speaker 03: And if the data has been mapped, regardless of how it got there, the board found that that would meet the limitation of the claim. [00:15:19] Speaker 03: Should this panel determine that that claim construction was an error, the board did do an extensive finding that even under Optus' proposed construction, it found that Wallace did teach that groups of symbols as well, that it does teach predetermining the group before mapping, or at least suggests it. [00:15:42] Speaker 03: Does the panel have any questions about the claim construction issue for me? [00:15:48] Speaker 03: Okay, so the next issue is this issue of whether or not Wallace teaches a first part of the data and a second part. [00:15:58] Speaker 03: And I think it's important to first note that the only time those terms show up in this patent are in the claims. [00:16:05] Speaker 03: So there's no discussion in the specification anywhere about dividing the time frequency grid or a first part or a second part. [00:16:14] Speaker 03: And much has been made about what has been characterized as a comparison of figures from the patent to the prior art. [00:16:22] Speaker 03: And I think what's missing there is that in both the board and hopefully in our brief, we made it clear that we were using the claims as the touchstone for both of those arguments. [00:16:33] Speaker 03: So in this case, the board looked and said, well, what is a first and second part of the domain according to the patent? [00:16:41] Speaker 03: We have nothing in the specification, but what we do have are these figures and some comments made in prosecution history about these figures and how they support this part of the claim. [00:16:54] Speaker 03: After doing that analysis, the board determined sort of what's good for the goose is good for the gander, that if figures four are enough to support the claim language, then one of skill in the art looking at figure two of the prior art reference, Wallace, [00:17:10] Speaker 03: would be able to also fill in the gap and come to that same, the claim language, and find that it meets those limitations. [00:17:19] Speaker 01: So in your view, figure 4D is an adequate example of a first part domain, second part. [00:17:28] Speaker 03: Yes, your honor. [00:17:29] Speaker 03: Figure 4D would be, in fact, I think any of the figures fours would satisfy this. [00:17:36] Speaker 03: 4D becomes a little more relevant in the non-contiguous argument, which I was going to get to next. [00:17:42] Speaker 03: But yes, your honor, you could find a first and second domain in figures 4D as well. [00:17:52] Speaker 03: And I guess we can move on to the noncontiguous argument. [00:17:56] Speaker 03: And again, this applies to four of the dependent claims. [00:18:00] Speaker 03: And again, we're in a situation where this concept of noncontiguous [00:18:05] Speaker 03: was not discussed in the specification. [00:18:07] Speaker 03: And this is again why we see the board looking at figure 4D and Dr. Wells provided expert testimony about one of skill in the art would look at figure 4D and determine that this group has some non-contiguous parts. [00:18:23] Speaker 03: In fact, there's one in the frequency domain and two in the time domain when we look at figure 4D. [00:18:34] Speaker 03: And I think there's some disagreement about, I wanted to cut actually on this point a couple of things. [00:18:44] Speaker 03: I believe I heard my colleagues say that Wallace nowhere discloses groups of data. [00:18:51] Speaker 03: And if I could just, if we go to appendix 1508, [00:18:57] Speaker 03: at column 20, lines 10 through 19, Wallace does have a discussion of using OFDM modulation in which it says, in one implementation of OFDM modulation, the data in each channel data stream is grouped to blocks, with each block having a particular number of data bits. [00:19:16] Speaker 03: So again, I think Wallace does have that concept of grouping. [00:19:20] Speaker 03: And as well, I think I heard my colleagues say that there was no disclosure [00:19:25] Speaker 03: Treating data one as a single group in Wallace and for that if we could just turn to appendix 1504 at line 24 through 25 It clearly states a data transmission eg data one may also occur over non-contiguous time slots So it's referring to it as a single group data one and that it's a [00:19:54] Speaker 03: In the prior art, it's saying here we are in non-contiguous time slots. [00:20:01] Speaker 03: And unless your honors have any other questions, I'll give back the remainder of my time. [00:20:09] Speaker 02: Thank you. [00:20:09] Speaker 02: Anything to do? [00:20:23] Speaker 00: Thank you, Your Honor. [00:20:24] Speaker 00: Just a few points briefly. [00:20:26] Speaker 00: I think Mr. Tyler said that there is nothing in the 569 patent about a first part and a second part of a domain. [00:20:34] Speaker 00: While the 569 patent may not use those terms, if you look at, for example, figures specifically for C, you can very easily see that there's two parts of a domain in which data is mapped in different ways. [00:20:49] Speaker 00: On the left side of the domain, it's mapped in these [00:20:52] Speaker 00: five by two blocks and then the right side it's mapped in these two by five blocks and then what is that a three by three by ten block so the patent does support a first part and a second part of a domain with respect to our third argument this argument about discontinuity and mapping [00:21:13] Speaker 00: I think Mr. Tyler fell into the exact same trap that the board fell into, explicitly comparing Wallace Figure 2 to the figures of the 569 patent and ignoring the claim language. [00:21:24] Speaker 00: The claim language specifically requires data that is not continuously mapped to each symbol within each of the group of symbols. [00:21:31] Speaker 00: If we compare the claim language to Wallace, it simply doesn't show it. [00:21:36] Speaker 00: To respond to two of his specific citations, he referred to the disclosure in Wallace on, I'm missing a number, but it's the book where he said that data one is mapped to time slots two and time slots seven, and therefore data one is a group of symbols. [00:21:54] Speaker 00: Data one is data that has been mapped to the domain. [00:21:57] Speaker 00: It is not the groups of symbols that have been formed prior to mapping. [00:22:03] Speaker 00: Unless the court has any further questions, we ask that the court reverse the board's final decisions.