[00:00:03] Speaker 02:
Okay, the next argued case this morning is number 20.

[00:00:07] Speaker 02:
1790.

[00:00:08] Speaker 02:
HTC Corporation against InvenSBE LLC.

[00:00:14] Speaker 02:
Mr. Franklin.

[00:00:28] Speaker 01:
May it please the court.

[00:00:30] Speaker 01:
To succeed on this appeal, appellants need to show that the Patent Trial and Appeal Board erred in finding no motivation to combine the Lee, BJN, and Hashim references.

[00:00:42] Speaker 01:
There is no dispute that these three references disclose each and every element of claims one through eight of the 439 patent.

[00:00:52] Speaker 01:
The only issue is the motivation to combine them.

[00:00:56] Speaker 01:
Which we review under a substantial evidence standard, right?

[00:00:58] Speaker 01:
That's correct.

[00:01:01] Speaker 01:
And appellants presented substantial evidence of a motivation to combine in the form of a reduced signaling overhead when the references are combined.

[00:01:13] Speaker 01:
And this evidence came from the references themselves, from our experts' testimony, and from the background section of the 439 patent, which talked about how the subcarriers of the prior art were combined into subbands

[00:01:31] Speaker 01:
with a single modulation and coding parameter applied to each of the subbands and how that's saved on signaling overhead.

[00:01:39] Speaker 03:
Can you just clarify something for me?

[00:01:41] Speaker 03:
And I guess I probably should understand this.

[00:01:43] Speaker 03:
First of all, signaling overhead is or is not different from feedback overhead.

[00:01:48] Speaker 03:
And in any event, can you explain in concrete terms that I might understand, in really concrete terms, what those phrases refer to?

[00:01:57] Speaker 01:
Sure.

[00:01:58] Speaker 01:
Overhead and feedback relate to the direction by which the travels.

[00:02:03] Speaker 01:
And I forget which one relates to which direction, but one relates to the uplink to the base station.

[00:02:08] Speaker 03:
Oh, that's all that is.

[00:02:10] Speaker 01:
Okay.

[00:02:10] Speaker 01:
And the other relates to the downlink to the handset.

[00:02:13] Speaker 03:
Okay.

[00:02:14] Speaker 03:
And then overhead is just the, here the number of words or bits one needs to use to tell people where to look for things.

[00:02:23] Speaker 01:
That's right.

[00:02:24] Speaker 01:
It's not the actual data that the person is viewing on their cell phone.

[00:02:28] Speaker 01:
It's the data that tells you how the subbands are created and what modulation and coding is being applied to those subbands.

[00:02:36] Speaker 01:
Okay.

[00:02:36] Speaker 01:
So in the admitted prior art, where the subcarriers were combined into subbands and the single modulation coding scheme was applied to each of those subbands,

[00:02:46] Speaker 01:
It was recognized in the admitted prior art that that would save on signaling overhead because you have fewer parameters that need to be conveyed between the user equipment and the base station.

[00:02:58] Speaker 01:
All the 439 patent is further bundled those bundles of subcarriers into a subband group and then do the same thing, which is apply a single modulation and coding scheme over all of those subcarriers.

[00:03:14] Speaker 03:
Reference Lee the primary reference and it is that it does Is there evidence that there is a trade-off between saving on overhead and So-called packing efficiency or those work.

[00:03:33] Speaker 03:
Yes, those do work in opposite directions in some sense.

[00:03:36] Speaker 01:
That's right and that that trade-off is mentioned in VJ and I

[00:03:40] Speaker 01:
And that was the trade-off that the Patent Trial and Appeal Board relied upon.

[00:03:44] Speaker 03:
And what's the, again, the intuitive concrete explanation?

[00:03:47] Speaker 03:
If you need less overhead, why does that reduce?

[00:03:53] Speaker 01:
Packing efficiency.

[00:03:54] Speaker 03:
The amount of each packet that can consist of data, which I gather is what was meant, is that right?

[00:04:00] Speaker 01:
That's right.

[00:04:01] Speaker 01:
So as you combine all of these subcarriers into a big subband group, there's the possibility that you'll end up with more wasted space.

[00:04:11] Speaker 01:
Think about packing suitcases for a trip.

[00:04:16] Speaker 01:
If you use smaller suitcases, you're going to be more efficient in your packing because you'll have less wasted space.

[00:04:23] Speaker 01:
If you use bigger suitcases, you're more likely to have wasted space in your suitcase.

[00:04:27] Speaker 01:
And so that's what packing efficiency relates to at a very high level.

[00:04:33] Speaker 01:
Where the board went wrong was in relying upon the trade-off with packing efficiency to negate the motivation to combine in order to achieve a reduction in signaling overhead.

[00:04:48] Speaker 01:
And that error was both a legal error and a factual error.

[00:04:54] Speaker 01:
The legal error was in failing to recognize that when a reference discloses both advantages and disadvantages of approach, that those disadvantages don't necessarily negate the motivation to combine.

[00:05:09] Speaker 01:
And that law is discussed in the Medicam versus Rolabo case.

[00:05:14] Speaker 01:
In the Erickson versus TCL case which are cited in our briefs.

[00:05:18] Speaker 04:
I agree with you It doesn't necessarily negate it but at the same time wouldn't a complete 103 analysis require taking into account both the advantages and the disadvantages and then have some kind of rationale some kind of explanation why you would want to

[00:05:38] Speaker 04:
It would be okay.

[00:05:40] Speaker 04:
A skilled artisan would nevertheless be motivated by the advantages and would be willing to absorb the disadvantages in the context of making the modification.

[00:05:53] Speaker 01:
I have two responses to that.

[00:05:54] Speaker 01:
The first one is that the case law doesn't require us to address the disadvantages because whenever you take a course of action, there's going to be some disadvantages associated with that.

[00:06:06] Speaker 01:
It was enough for us to show that the prior art shows the advantage of applying a single modulation coding scheme over a group of subbands and that this would achieve signaling overhead.

[00:06:17] Speaker 01:
But here we get to the factual error in the board's decision.

[00:06:21] Speaker 01:
which is the combination of the primary reference Lee and the secondary reference.

[00:06:26] Speaker 04:
I guess I'm a little lost there where you're saying that there's no requirement to take into account the disadvantages.

[00:06:33] Speaker 04:
I mean, what if the prior art reference said you really can improve your signaling overhead, but with the trade-off of producing a lot of cancer-inducing radiation?

[00:06:46] Speaker 04:
I mean, I would think anybody trying to rely on that teaching to combine it with another reference would have to take into account that a skilled artist would understand that the reference is teaching you that if you go ahead and add this feature to anything, it's going to also include cancer-inducing radiation.

[00:07:05] Speaker 01:
Yes, and if VGM, the second reference, had said something like that, then you may have a good point there.

[00:07:11] Speaker 01:
But all VGM was saying is that there's a recognized trade-off, a known trade-off, when you use subband groups.

[00:07:17] Speaker 01:
You get the benefit of signaling overhead.

[00:07:19] Speaker 01:
But a person of ordinary skill in the art would also have to take into account

[00:07:23] Speaker 01:
the loss of the reduction in packing efficiency and that's just something that a person of ordinary skill in the art would know how to do.

[00:07:31] Speaker 01:
Just like there's always going to be some disadvantages when you take a certain course of action.

[00:07:37] Speaker 04:
But here we happen to have the disadvantage

[00:07:43] Speaker 04:
articulated and so therefore it's staring at the fact-finder in the face.

[00:07:49] Speaker 04:
I don't understand why you're saying that it's so easy to just dismiss and almost pretend it's not there because any course of action necessarily always has some kind of trade-off with it.

[00:08:05] Speaker 01:
Well, it's a known trade-off which means a person of ordinary skill in the art would know to take that into account when doing a

[00:08:13] Speaker 01:
we're doing a combination.

[00:08:14] Speaker 01:
And if you have a combination for which signaling overhead is a priority, as it is in Lee, which specifically has a concern with saving on signaling overhead, then you are going to, you know, be motivated to make that combination.

[00:08:30] Speaker 01:
If packing efficiency is more important, then that might be a disadvantage.

[00:08:33] Speaker 01:
But Lee specifically talks about, Lee is the primary reference, signaling overhead and saving on that is a priority.

[00:08:40] Speaker 01:
The factual problem with the board's analysis is that the combination we propose between Lee and VJN actually doesn't affect the packing efficiency.

[00:08:50] Speaker 01:
And that's because packing efficiency concerns how the sub-band groups are put together, how many sub-bands are in it.

[00:08:59] Speaker 01:
And all of this is set in the primary reference Lee.

[00:09:03] Speaker 01:
Lee teaches the subband groups.

[00:09:06] Speaker 01:
For example, in Figure 6 of Lee.

[00:09:09] Speaker 01:
So by the time we get done with the primary reference, Lee, the subband groups are already set, and plus the packing efficiencies are already set.

[00:09:18] Speaker 01:
All appellants relied upon BJ Anne for

[00:09:21] Speaker 01:
was the notion that once you've already got the subband group set in Lee and plus the packing efficiency, you can achieve some overhead signaling savings by applying joint modulation and coding parameters over all of the subbands in that subband group.

[00:09:39] Speaker 03:
That's the factual error.

[00:09:40] Speaker 03:
Can I return to what I asked you before?

[00:09:41] Speaker 03:
One of the problems with this case and this case is not even close to unique is that people keep using phrases

[00:09:50] Speaker 03:
that do not carry immediate concrete images of what's going on.

[00:09:55] Speaker 03:
Tell me more concretely than suitcases what's going on with packing efficiency.

[00:10:00] Speaker 03:
Because until I understand that, I'm just hearing words floating around, and I'm not getting an image of whether you're substantively onto something or not about saying, oh, there is no more problem of this sort, which I haven't explained yet, in moving from subbands to subband groups.

[00:10:20] Speaker 03:
Concretely, what is going on with packing efficiency?

[00:10:23] Speaker 01:
So with packing efficiency, you're

[00:10:26] Speaker 01:
You're taking different sub bands on the frequency spectrum and combining them into sub band groups So to go to the suitcase analogy You can have small suitcases and put your clothes in those small suitcases and you can make better use of the space potentially because you may have less wasted space within a suitcase where you can use bigger suitcases and

[00:10:51] Speaker 01:
which may take less information to describe because you've got fewer suitcases, but because the suitcases are bigger, you may have more wasted space in the suitcases.

[00:11:00] Speaker 01:
Is that concrete enough for you or?

[00:11:03] Speaker 03:
No, but that's okay.

[00:11:05] Speaker 01:
All right.

[00:11:05] Speaker 01:
Well, it sounds like this is an important point, so let me see if I can maybe do a better job of this.

[00:11:13] Speaker 03:
How about doing it without the analogy?

[00:11:15] Speaker 01:
Okay, without the analogy.

[00:11:18] Speaker 01:
On the frequency spectrum,

[00:11:21] Speaker 01:
You can put together the

[00:11:24] Speaker 01:
various sub bands.

[00:11:25] Speaker 01:
And if you take a look at, for example, at Figure 9 of VJN, you can see how they put together different sub band groups, which they call PLCs.

[00:11:33] Speaker 01:
If you make the PLCs big... That's a physical layer carrier?

[00:11:38] Speaker 01:
That's right, a physical layer channel.

[00:11:39] Speaker 01:
Channel, sorry.

[00:11:41] Speaker 01:
If you make those PLCs big, you're more likely to have wasted space, like shapes that you've got to put together like a puzzle form.

[00:11:50] Speaker 01:
And if the shapes are big, you're more likely to have wasted space between the shapes.

[00:11:55] Speaker 01:
If you make them smaller, more granular, you've got less wasted space between the shapes.

[00:11:59] Speaker 01:
But you have to have more overhead signaling to convey between the handset and the base station about how those shapes are set up and what modulation and coding parameters go along with them.

[00:12:15] Speaker 01:
So that has to do with the packing efficiency, but Lee, the primary reference, has already specified that you're going to have subband groups.

[00:12:23] Speaker 01:
It's already put the various subbands together into the subband groups.

[00:12:28] Speaker 01:
Lee doesn't have subband groups.

[00:12:31] Speaker 01:
Lee does have subband groups.

[00:12:34] Speaker 01:
And if you look at figure six of Lee.

[00:12:37] Speaker 04:
I thought you were relying on Vijayan for the disclosure of subband groups and you're modifying Lee to have subband groups in light of Vijayan.

[00:12:47] Speaker 01:
Am I misunderstanding it?

[00:12:49] Speaker 01:
That is a misunderstanding.

[00:12:51] Speaker 01:
Lee teaches subband groups and if you look at figure six of Lee, you'll see the subband groups Lee takes every fourth subband along the frequency axis and makes that into a subband group We're relying upon VG and not for the disclosure of a subband group But for the disclosure that once you have a subband group you can apply a single Modulation and coding scheme over the entire subband group

[00:13:20] Speaker 01:
And that's where we're relying upon the VJN for.

[00:13:26] Speaker 01:
That makes sense.

[00:13:28] Speaker 03:
What you do with the sub-band groups that you rely on VJN to add to sub-band groups that were present in Lee but not used this way?

[00:13:41] Speaker 01:
That's right.

[00:13:42] Speaker 01:
So in Lee, each sub-band within the sub-band group still gets its own modulation and coding scheme.

[00:13:50] Speaker 01:
In VJN, VJN teaches you can apply the same modulating coding scheme over all of the subbands in the subband group.

[00:13:59] Speaker 01:
And that's all we're relying upon VJN for.

[00:14:02] Speaker 01:
Lee has already created the various subband groups, decided which subbands go into it and how they're going to be configured.

[00:14:09] Speaker 01:
VJN says once you've done that, just apply a single modulation and coding scheme over that, and you can save on signaling overhead.

[00:14:17] Speaker 01:
The only other reference that we need to bring in is Hashem, because once you combine Lee and Vijayan, the question is where are those parameters decided?

[00:14:26] Speaker 01:
Are they decided at the handset or are they decided at the base station?

[00:14:32] Speaker 01:
And even putting Hashem aside, under KSR versus Teleflex, we know that when there's a finite number of known and predictable solutions to a problem, here there's only two.

[00:14:42] Speaker 01:
Either the parameters are decided at the handset

[00:14:44] Speaker 01:
or they're decided at the base station, it's obvious to choose one or both of those choices.

[00:14:52] Speaker 01:
And we could have stopped there, but we brought in Hashim, which makes clear that not only would a person of ordinary skill in the art known to have done the parameter selection at the handset, but that there was a reason to do so.

[00:15:04] Speaker 01:
Because Hashim says when you make the parameter decision at the handset, you can further save on signaling overhead.

[00:15:11] Speaker 01:
And it was air for the board, first not to recognize that there was only two choices that could have been done here, and then not to look at Hashim and see that Hashim explicitly states the motivation to do it that way.

[00:15:26] Speaker 01:
And I see that I'm done with my argument.

[00:15:28] Speaker 01:
Before I turn this over, does the Federal Circuit have any additional questions?

[00:15:35] Speaker 02:
OK.

[00:15:35] Speaker 02:
Let's hear from the other side, and we'll save you rebuttal time.

[00:15:38] Speaker 01:
Thank you.

[00:15:48] Speaker 00:
Morning your honors Your honors a panasonic invented adaptive modulation and coding or AMC on a subband group basis based on channel estimations made per Subband and made on the phone

[00:16:08] Speaker 00:
And the 439 patent did say that AMC based on a subcarrier by subcarrier basis and based on a subband basis was known, but not on a subband group basis based on channel estimations per subband.

[00:16:22] Speaker 00:
That's what the Patent Office granted a patent for.

[00:16:26] Speaker 00:
And substantial evidence certainly supports the board's finding that HTC failed to meet its burden

[00:16:32] Speaker 00:
to cobble together three different and very disparate references to manufacture the claim elements at issue.

[00:16:40] Speaker 00:
Just a step through the references to make sure we know what we're talking about.

[00:16:43] Speaker 03:
Well, just explain.

[00:16:46] Speaker 03:
There's a picture formed by the other side that

[00:16:52] Speaker 03:
It's kind of no big deal.

[00:16:54] Speaker 03:
It's pretty obvious.

[00:16:56] Speaker 03:
You can do something at the most granular level.

[00:16:59] Speaker 03:
You can do it at the next grouping level.

[00:17:01] Speaker 03:
You can do it at the next grouping level.

[00:17:04] Speaker 03:
That's all you've done is done the third grouping level.

[00:17:08] Speaker 03:
And, you know, there are tradeoffs in things with names here called packing efficiency and overhead.

[00:17:15] Speaker 03:
So, you know, obvious to at least try each of these choices.

[00:17:22] Speaker 03:
So I think what was missing... I assume it's more complicated than that, but that's kind of the picture they're presenting.

[00:17:28] Speaker 00:
Yes, Your Honor.

[00:17:28] Speaker 00:
I was going to say, I think we're missing technologically what's actually going on here.

[00:17:33] Speaker 00:
And I think it relates to this idea of the channel estimation.

[00:17:37] Speaker 00:
So each sub-carrier may have different channel conditions, poor quality, higher quality,

[00:17:43] Speaker 00:
And each subband is a bunch of subcarriers that may have different conditions.

[00:17:47] Speaker 00:
And then the different subbands themselves spaced out over the entire bandwidth may have different conditions.

[00:17:53] Speaker 00:
So this packing efficiency trade-off thing is if you measure each subcarrier and its conditions and give the modulation and coding parameters that are perfect for that subcarrier, well, then you're going to do a good job on packing

[00:18:09] Speaker 00:
Efficiency because you're going to get exactly what that subcarrier needs But you're going to have to report all that information on a subcarrier by subcarrier basis And you're going to have a lot more overhead signaling so that overhead signaling is the stuff you're reporting

[00:18:26] Speaker 00:
Yes, it's the stuff that's not the data that you're going to send.

[00:18:30] Speaker 00:
It's this signaling of here is what I need or here is what I want.

[00:18:36] Speaker 00:
And so the Lee reference looks at this and says, well, we're going to do this on a subband level basis.

[00:18:44] Speaker 00:
So for multiple subcarriers, we're going to report one modulation encoding.

[00:18:48] Speaker 00:
And that's what the 439 that I already said was known, was subband-based adaptive modulation and coding.

[00:18:57] Speaker 00:
Vagion is a completely, I think that's the third different pronunciation of Vagion, but that's how I've been saying it.

[00:19:03] Speaker 00:
It's completely different.

[00:19:05] Speaker 00:
So Vagion is entirely on the base station.

[00:19:08] Speaker 00:
The whole reference is just about base station allocation for the downlink channel to the phones where Vagion, of course, knows everything, knows all conditions, knows all the phones in its area.

[00:19:19] Speaker 00:
And there, it doesn't always do that.

[00:19:21] Speaker 00:
If you look at the best picture, if you want a picture, look at figure 7a of a giant, it'll show you the PLCs.

[00:19:29] Speaker 00:
One of them, it just does one subband and does modulation coding parameters for that.

[00:19:34] Speaker 00:
The next one, it chooses to put three together.

[00:19:37] Speaker 00:
And the next one, it chooses to just do two.

[00:19:40] Speaker 00:
Because it can look at that and say for contiguous sub bands that are next to each other either base station can decide well these are close enough I can do one set of modulation coding parameters for these two sub bands together or these three sub bands together

[00:19:58] Speaker 04:
Is everybody creating a subband group in that instance?

[00:20:02] Speaker 00:
Yes.

[00:20:02] Speaker 00:
And we haven't argued that the vagina doesn't have a subband group for the downlink.

[00:20:07] Speaker 00:
They are contiguous, next to each other, subbands that the vagina sort of puts together on the fly based on the conditions that it sees.

[00:20:17] Speaker 03:
Why is contiguity of the subbands relevant to the amount of overhead?

[00:20:23] Speaker 00:
Well, if you dig into the 103 analysis and what HTC would have to show here, their proposal is to take this notion of one modulation coding parameter for contiguous subbands in Vagina and just plug it into Lee.

[00:20:40] Speaker 00:
What Lee has is different modulation coding parameters for each subband.

[00:20:45] Speaker 00:
Separately, Lee suggests another thing we could do is have, they call them clusters, but it's the same thing as a subband.

[00:20:53] Speaker 00:
For a given wireless device, for a given phone, let's assign it subbands spaced out over the entire bandwidth.

[00:21:02] Speaker 00:
One here, one here, one here, one way over here.

[00:21:05] Speaker 00:
If we have those spaced out, then at least one or two of them should see good channel conditions.

[00:21:11] Speaker 00:
and that way that phone will get decent throughput because at least some of the conditions will be good.

[00:21:18] Speaker 00:
We are still going to do different modulation and coding parameters for each subband spaced out because they could have wildly, radically different channel conditions and need to have completely different modulation and coding parameters for each one of those.

[00:21:35] Speaker 00:
The proposal from HTC was to say

[00:21:38] Speaker 00:
Well, just take where Vagine has figured out for the downlink that two subbands next to each other could get the same parameters.

[00:21:46] Speaker 00:
Just apply that in LEED to its four subbands spaced out over the spectrum.

[00:21:52] Speaker 00:
Even though LEED doesn't say to do that.

[00:21:53] Speaker 00:
LEED says do each one separately because they could be totally different.

[00:21:57] Speaker 00:
Just apply that down to LEED.

[00:21:59] Speaker 00:
And this is where the tradeoff comes in.

[00:22:02] Speaker 00:
This is what the board found, that there wasn't a sufficient articulated, you know, rationale or reasoning to do that.

[00:22:10] Speaker 00:
Why would what Vagina figured out to do for a neighboring subband

[00:22:14] Speaker 00:
apply to these cluster groups that are spaced out?

[00:22:18] Speaker 00:
Why would you make that trade-off where you know if you apply one set of modulation and coding parameters to each of these spaced out clusters, you're going to have some that are not going to work?

[00:22:31] Speaker 00:
And Dr. Boychich, getting back to our substantial evidence standard, he goes through this in some detail, what the problem is.

[00:22:39] Speaker 00:
where if you use one set of modulation and coding parameters, if you use the minimum for the worst, you know, whatever cluster has the worst channel quality, if you use that minimum one, well, they're all going to work, but you're going to lose out on other clusters that have higher quality.

[00:22:57] Speaker 00:
You're not going to get as much throughput.

[00:22:58] Speaker 00:
You're not going to get as much data through on those clusters.

[00:23:01] Speaker 03:
Empty space in the larger suitcase.

[00:23:04] Speaker 00:
Yes, you could have.

[00:23:06] Speaker 00:
It's a little bit of a, I wouldn't go with that analogy, your honor.

[00:23:09] Speaker 00:
But yes, you're going to miss out on what you could do.

[00:23:13] Speaker 00:
And if you use the highest, as Dr. Vojtich explained, use the best cluster and that modulation coding parameter set, you have good throughput.

[00:23:22] Speaker 00:
But for lower quality clusters, they're not even going to be able to decode it.

[00:23:27] Speaker 00:
So it's not going to work.

[00:23:28] Speaker 00:
What about the average?

[00:23:31] Speaker 00:
What's that?

[00:23:31] Speaker 00:
What about the average?

[00:23:32] Speaker 00:
The average, Your Honor.

[00:23:34] Speaker 00:
The average comes in.

[00:23:35] Speaker 00:
This is something that really came up in reply.

[00:23:39] Speaker 00:
HTC had Dr. Ding put in about, I think, over a hundred page reply declaration that goes into this and says, well, let's do the average.

[00:23:48] Speaker 00:
If you do the average, you're just going to have some set of the same problems.

[00:23:53] Speaker 00:
It's still going to be the modulation coding parameters are going to be too high for some clusters or subbands and too low for others.

[00:24:02] Speaker 00:
And the board looked right at this issue, again, getting back to substantial evidence, and said there's not a good enough reason to say why would you use the average?

[00:24:11] Speaker 00:
The 439 patent, the Lee reference, when they talk about using the minimum,

[00:24:16] Speaker 00:
or there's a maximum, or the average, why would you do that?

[00:24:20] Speaker 00:
Why would you make that trade-off?

[00:24:22] Speaker 00:
Where again, I'm not even sure, but perhaps in Vagina with its subbands next to each other, maybe it's saying, well, these are pretty close, and if I use the average, that'll be good enough.

[00:24:34] Speaker 00:
And so I'll group these to what it calls PLCs for the downlink down to the phone.

[00:24:40] Speaker 00:
But why would you do that again in LEED?

[00:24:43] Speaker 00:
where Lee's clustered groups are sub bands spaced out over the bandwidth.

[00:24:50] Speaker 00:
And so why would a person with ordinary skill in the art think, oh, I'll just slap the average on there, and that'll be great?

[00:24:57] Speaker 00:
It may or may not be great based on the different conditions.

[00:25:00] Speaker 00:
And so the board required a better reasoning for why you would do that for that particular combination of Vagina and Lee.

[00:25:14] Speaker 00:
And I think you know the other point to make on that just again for appellate purposes if you look at the board's decision It's very clear, and then they go through dr. Voytage they go through dr. Ding they conclude for instance Appendix 48 that dr. Ding's testimony on that exact issue.

[00:25:32] Speaker 00:
I was just discussing was Conclusory and they accord it minimal weight and

[00:25:38] Speaker 00:
They separately look at Dr. Voychich and say that he provides credible explanation.

[00:25:45] Speaker 04:
And so that would be a basis right there for... I heard your opposing counsel this morning say that Lee teaches subband groups already.

[00:25:54] Speaker 04:
And by the way, whatever you... the teaching that you borrowed from the JN to Lee, it's not going to hurt packing efficiency at all.

[00:26:05] Speaker 00:
Yeah, I think that's just...

[00:26:08] Speaker 00:
technological nonsense your honor.

[00:26:10] Speaker 04:
What did the board say about that?

[00:26:12] Speaker 00:
So Lee of course doesn't do adaptive modulation and coding per subband group.

[00:26:19] Speaker 00:
They would rely on that.

[00:26:20] Speaker 00:
It does it per subband.

[00:26:22] Speaker 00:
Different modulation and coding parameters for each subband.

[00:26:28] Speaker 00:
The idea, which I know it's technologically difficult to picture this, but if you look at Lee's figure six again,

[00:26:37] Speaker 00:
What it suggests is putting different clusters together that are spaced out over the bandwidth, not for doing adaptive modulation and coding, not for providing one set of parameters, but instead to achieve frequency diversity for a given cell phone so that that cell phone will get

[00:26:57] Speaker 00:
Piece of the bandwidth over here over here over here over here so that for that cell phone It will get at least some clusters that are That are good, and this will maximize overall throughput.

[00:27:08] Speaker 00:
This is kind of the balance that that Lee strikes Vagina be like their proposal of HTC just apply an average Modulation and coding parameters use average Si and R and

[00:27:22] Speaker 00:
across all of those, that is for sure going to result in less packing efficiency.

[00:27:28] Speaker 00:
Because for leave, say cluster number one is fairly low channel quality.

[00:27:35] Speaker 00:
It then gets the modulation and coding parameters appropriate for that cluster.

[00:27:41] Speaker 00:
Cluster number two, spaced far apart on the spectrum, very high channel quality.

[00:27:47] Speaker 00:
And that gets much higher modulation and coding parameters.

[00:27:51] Speaker 00:
So that will get the most data through there, the most suitcases or full suitcases if we're going with that in the higher quality one.

[00:27:58] Speaker 00:
If you go and just take the average, now this high quality channel over here, high quality cluster is going to do less than it could.

[00:28:08] Speaker 00:
You're going to get less through.

[00:28:09] Speaker 00:
There's where you're sacrificing your packing efficiency.

[00:28:12] Speaker 00:
and use the average over on the low quality cluster, it might still work.

[00:28:17] Speaker 00:
If it's low enough, it won't be able to decode it, and it won't work at all.

[00:28:21] Speaker 00:
And this is what Dr. Voychich went through.

[00:28:24] Speaker 00:
That's the trade-off.

[00:28:26] Speaker 00:
And this is the trade-off that was mentioned all the way back in the petition.

[00:28:29] Speaker 00:
It's very clear they went through talking about this trade-off, said it would work.

[00:28:33] Speaker 00:
But when you dig into the details, Vagian and Lee are so different from one another.

[00:28:39] Speaker 00:
That it doesn't work and the board said you would need a much Much better reasoning for why a person of ordinary skill in the art would make that trade-off I Will mention Hesham of course this is a three reference combination.

[00:28:57] Speaker 00:
I'll take quick issue with my Colleagues saying that there's no dispute about whether all the elements are in the art As we put is our statement of issues number one

[00:29:09] Speaker 00:
Element 1B is not in the art.

[00:29:11] Speaker 00:
Element 1B is the parameter deciding section where you decide modulation and coding parameters for a subband group based on channel estimation per subband.

[00:29:23] Speaker 03:
I took their point to be not that each element in the sense of the paragraphing is in the prior art, but that every little specific thing that you need to find you can find so as to

[00:29:38] Speaker 00:
put together everything that would be found in 1B, if that's... Sure, but the issue is that element says, decide modulation and coding parameters for a subband group based on channel estimation per subband.

[00:29:55] Speaker 00:
Now, Lee, of course, does not decide modulation and coding parameters for a subband group.

[00:30:00] Speaker 00:
Vagyan does, but it never mentions channel estimation.

[00:30:04] Speaker 00:
So it doesn't say I'm making this decision based on channel estimation per subband.

[00:30:10] Speaker 00:
Hashem is that a completely different type of reference.

[00:30:15] Speaker 00:
In Hashem, you measure every single subcarrier.

[00:30:19] Speaker 00:
figure out which ones are acceptable or unacceptable, as far as whether they are above a threshold of, say, quality.

[00:30:26] Speaker 00:
If they're above, so then Hashem takes basically the good sub-carriers.

[00:30:32] Speaker 00:
gets rid of the bad ones, says just take the good ones.

[00:30:35] Speaker 00:
And now we can do this thing called link mode where they just pick modulation and coding parameters for the good subcarriers.

[00:30:42] Speaker 00:
And Hashem says, yes, you could do that at the base station or do that at the phone.

[00:30:46] Speaker 00:
But that is, again, a completely different system, a different situation.

[00:30:50] Speaker 00:
The board said, you would need a good reason to say why you would do that to move the kind of on-the-fly contiguous subband group modulation and coding up a giant.

[00:31:01] Speaker 00:
and move that into Lee and into its frequency diverse cluster groups.

[00:31:07] Speaker 00:
It's not so simple as all that because these are, again, three very different references.

[00:31:13] Speaker 00:
And so that's, of course, to win on a, you know, if either of those fails, either of those links in the chain of combining the three references fails, the decision should be affirmed.

[00:31:23] Speaker 00:
And I do think that should be the result here.

[00:31:27] Speaker 00:
And I think I can wrap it up unless the panel has any other specific questions.

[00:31:32] Speaker 02:
Thank you.

[00:31:34] Speaker 02:
Thank you counsel.

[00:31:36] Speaker 01:
Thank you your honor Thank you your honor

[00:31:56] Speaker 01:
The trade-off that my esteemed colleague was talking about, whereby if you use subband groups and having to use less than ideal parameters if you use a joint modulation coding scheme, that is a different trade-off than the packing efficiency, which is what the board based its decision on.

[00:32:17] Speaker 01:
Packing efficiency relates to how you put the subband groups together.

[00:32:22] Speaker 01:
So as we discussed in here's figure six,

[00:32:26] Speaker 01:
Lee Lee talks about creating sub band groups by taking every fourth Subband along the frequency spectrum and creating a sub band group out of it So sub band groups are certainly taught in Lee see figure six Here's VJ and in VJ and depending on the size of the sub band group

[00:32:49] Speaker 01:
And how it's configured, you could have losses in packing efficiency.

[00:32:52] Speaker 01:
Because you may not need a PLC that's this big.

[00:32:56] Speaker 01:
Or by creating PLCs in this shape, you may have some wasted space.

[00:33:01] Speaker 01:
But none of this is implicated by the combination of VJM and Lee, because the subband groups are already created in Lee, which is our primary reference.

[00:33:10] Speaker 01:
All we're relying upon VJM for is the idea that once you've created the subband groups, you can use a single modulation and coding scheme over that entire group.

[00:33:19] Speaker 01:
So that was an error in the board's decision.

[00:33:22] Speaker 01:
The trade-off that my esteemed colleague is talking about has to do with when you use a joint modulation and coding scheme over an entire subband group, some of the subcarriers in that subband group will get less than ideal parameters.

[00:33:35] Speaker 01:
But our expert addressed that in length in his expert declaration.

[00:33:40] Speaker 01:
You can look at paragraphs 177 to 186 of this declaration, which starts at around page 1921 of the Joint Appendix.

[00:33:51] Speaker 01:
Almost all of this evidence was not considered by the board in its final written decision.

[00:33:57] Speaker 03:
And this evidence... Cited but you think not considered, is that right?

[00:34:02] Speaker 01:
It's at one point the board's, you know, string site cites all of the paragraphs, but it's,

[00:34:09] Speaker 01:
There's no substantive discussion of the fact that the Lee reference teaches using an average, that the Hashim reference teaches using an average, that the Prior Art WiMAX standard used an average, our experts' discussion of an LTE textbook that talks about error coding and interleaving, and just the notion that it doesn't make sense to do what their expert has suggested, which is to choose either the highest

[00:34:38] Speaker 01:
signal to noise ratio or the lowest signal to noise ratio to choose the modulation and coding parameters.

[00:34:46] Speaker 01:
Now what makes sense is to use the average.

[00:34:49] Speaker 01:
When you use the highest ratio, most of the sub-carriers won't work because you're trying to send the information through too fast.

[00:34:58] Speaker 01:
If you use the lowest ratio, then you kill the throughput because you're doing everything too slow.

[00:35:03] Speaker 01:
It makes sense to do the average

[00:35:07] Speaker 01:
so that you can have decent throughput while still having enough sub-carriers being able to send the data through.

[00:35:13] Speaker 01:
And that's how it would make sense to do and our expert went off for length about why it would make sense to do the average and why a person of ordinary skill in the art would have known to have used the average with explicit examples both from the side of the reference and from the WIMAC standard that this was done and it wasn't considered by the board.

[00:35:33] Speaker 01:
On top of all this, the 439 patent itself talks about four existing methods for setting joint modulation and coding parameters when you have multiple subbands.

[00:35:47] Speaker 01:
And of these four methods, two involve using an average.

[00:35:53] Speaker 01:
So again, we get back to the precept from KSR versus Teleflex.

[00:35:58] Speaker 01:
We have a finite number of known and predictable solutions to a problem, known and predictable design choices.

[00:36:06] Speaker 01:
It's obvious to try them, to use them.

[00:36:10] Speaker 01:
And so we've got plenty of evidence in the record that a person of ordinary skill in the art would have combined VGN with Lee using an average signal to noise ratio.

[00:36:19] Speaker 01:
And then finally, in the 30 seconds I've got left on the Hashim reference,

[00:36:24] Speaker 01:
We're only using Hashem to show that between the user equipment and the base station, which there's no dispute, those are the only two places that the joint parameters can be decided, that a person of ordinary school could have done it.

[00:36:40] Speaker 03:
But does that actually agree that those are the only two places?

[00:36:43] Speaker 03:
That's not self-evident to me.

[00:36:46] Speaker 03:
For example, neighboring handsets.

[00:36:51] Speaker 03:
I guess it's theoretically possibly... But was it actually agreed or just not disputed?

[00:36:56] Speaker 01:
There hasn't been any dispute or any suggestion that it would be done elsewhere.

[00:37:00] Speaker 03:
Okay.

[00:37:01] Speaker 01:
But in any event, and I see that my time is up, Hasham teaches, do it at the handset and gives a reason for doing so, which will save signaling overhead.

[00:37:10] Speaker 01:
Are there any further questions from the court?

[00:37:15] Speaker 02:
Thank you.

[00:37:16] Speaker 02:
The case is taken under submission.

[00:37:18] Speaker 02:
Thanks to both counsels.