Oral argument — INTERDIGITAL COMMUNICATIONS v. ITC

Good afternoon, and may it please the Court. This appeal focuses on two claim construction issues, but they are very unusual claim construction issues for three reasons. First of all, the Administrative Law Judge initially agreed with InterDigital that the ordinary plain meaning of both phrases should be used, and then he went ahead and added additional limitations to those claims, which I submit were unwarranted additional limitations. Second, those additional limitations not only excluded the preferred embodiment in InterDigital's patents, but excluded all embodiments. And third, the ALJ's constructions were, in my opinion, so clearly unwarranted that even the Commission, in its red brief, has run away from at least one of those constructions. Let me be more specific. Let me talk about the first limitation, the code signal limitation. The Administrative Law Judge initially agreed with InterDigital . . . JUSTICE SCALIA But before you get into your argument, Mr. Donner, just to get the lay of the land here, I wonder if you could tell me what your construction, your proposed construction of both the term code and the term spreading code is, so that I see what the baseline here is for work. MR. DONNER Our construction is that code is a sequence of chips, and our construction of spreading code is at least one code that can increase bandwidth but doesn't have to increase bandwidth. That's our definition. JUSTICE SCALIA And when you say a sequence of chips, you mean chips being defined as bits at a chip rate? MR. DONNER Exactly. Exactly. Now, the Administrative Law Judge initially construed the terms code and signal to be a sequence of chips, and consistent with what InterDigital had asked him to do. But then he added a limitation and he said, no, a code is not just a code, it's a spreading code, and a signal is a spreading code that's transmitted. Now, it is true that the preferred embodiment in the power ramp-up patents include spreading codes, and it is also true that the word spreading code is used a number of times in those patents. But it is also true that there's a clear, rational reason for not restricting the independent claims to spreading codes. JUSTICE SCALIA But does it actually in the specifications say that code is not limited to spreading code? Now, that's what I was looking for. MR. DONNER There are two things I would refer Your Honor to. One is that the 010 patent is incorporated by reference in the power ramp-up patents, and nobody disputes that, and the 010 patent has an express statement on page 504 of the appendix, and it says in column 17 on page 504, lines 46 and 7, the spreading code is a sequence of symbols called chips or chip values. Secondly, if you look at the spec itself, the 966 patent by way of example, if you look at column 3, lines 23 and 4 and 5, the short code is a sequence for detection by the base station, so on and so forth, which is a shorter period. JUSTICE SCALIA As long as it's the short code which doesn't spread, that that takes care of the definition, the broader definition of code, when they use the short code? MR. DONNER We say the short code can be defined as a sequence of chips. Now, it is true that the examples use spreading codes, but the reason for having a broad claim, we have broad claims which don't say spreading codes. We have 13 dependent claims in the collection of power ramp-up patents which further define code or signal as a spreading code. That's the only additional difference, and the normal doctrine of claim differentiation says that if the only difference is some more specific term, that creates a strong presumption that the two terms should be construed differently. The problem is that the codes, the short codes, are not used to spread. Nobody disputes that. Both NOPIA agrees with that, and the Commission agrees with that. During the power ramp-up procedure, the short codes are merely used to establish communication with the base station. A spreading function is not involved, and the mischief of construing code or signal as a spreading code is that what can happen is what happened in this case. The administrative law judge not only construed it as a spreading code, but he then went further and he said a spreading code is a spreading code used or intended to be used to increase bandwidth, which clearly does not happen in this case. So I submit that there is a good reason the doctrine of claim differentiation not only applies presumptively, but should be applied in this case to avoid that very mischief. But I would further note that even if the Court goes further and says, well, what if we limited it to spreading codes? Our position is that the bottom line, the bottom line of infringement, would not change even if it's limited to spreading codes. But spreading codes are not limited to spreading codes. So if the LJ defines spreading codes the way I answered Judge Bryson, and that is a spreading code is a code that can be used to increase bandwidth, but doesn't have to be used to increase bandwidth, then there would be no problem. But the administrative law judge went further and he said it must be used or that creates a problem, and we submit it is wrong for a number of reasons. First of all, none of the claims mention increasing bandwidth. Secondly, as I mentioned, the short codes and the codes used in the power ramp-up procedure do not increase bandwidth. On the contrary, they're disclosed as being used to establish communication, they perform timing functions, they perform discrimination functions between handsets, but they do not increase bandwidth. The end result of this is the administrative law judge's construction excludes all the embodiments in this patent, and there are many, many cases from this Court saying that ordinarily that's not a good way to proceed, to come up with a construction that excludes all of the disclosed embodiments. Now, the fact is that there is a ton of evidence, intrinsic and extrinsic, supporting the notion that the codes that are used in the specification and in the accused devices are, in fact, spreading codes. Let me talk about the accused devices. The administrative law judge, or the commission, in its brief on Red 37 said that a spreading code has certain characteristics, three characteristics. It operates at the chip rate, has a spreading factor, and is prepared from a spreading code generator. The briefs we filed make absolutely clear that that is applicable to the Pratch preamble. The CDMA standard, NOPIA technical specification, a number of patents, including a NOPIA patent, characterized the scrambling code in the Pratch preamble as a spreading code. For all those reasons... It is not used to increase bandwidth, but it can. It could, if you wanted to use it to increase bandwidth, because it's a sequence of chips, and as long as you want it to be modulated by a data stream transmitted at a lower rate than the spreading code, you It says the type of code that can be. The type that is intended to be used. Now, that is as close to what we say as you can get. It clearly is a move away from what the administrative law judge says. The scrambling code could fit into that category, and in fact, it's submitted as being a spreading code, and the signature part of the Pratch preamble is also described as a spreading code, even though in the power ramp-up procedure, they do not spread, just as our short codes do not spread, but they are the type series of ones and zeros in a stream that are chips that can be used to spread as long as you use it in conjunction with the data stream moving at a slower rate. So, the bottom line is, we submit that whether you... We submit sequence of chips is the correct answer. We further submit that even if it's a spreading code, we should prevail on the issue of whether code or signal covers the Pratch preamble and covers the Nokia devices. I'd like to move briefly into the increased power level part. The plain meaning was found by administrative law judge as a power level of a transmission is higher than that of a previous transmission. Now, that definition, that construction, says nothing about increase during transmission, and even more than interestingly, the administrative law judge says that the literal meaning of the claim does not restrict it to during the transmission. He said it was generic. Generic to stepwise, which is what the Nokia device does, or increase during transmission. But then, he nevertheless went ahead and he restricted it to increase during transmission, excluding stepwise. Now, had the administrative law judge stuck to his original construction, there clearly would be infringement. We point out on pages in blue 56 and 57, there's a chart of how the Nokia device operates, and it clearly would fit within the stepwise category. But the reason for his embellishing the claim, for further embellishing the claim, was based on a fundamental misunderstanding of the patent. The judge, and it's picked up in the briefs of the commission in Nokia, the judge says the whole reason for this invention is to prevent overshoot, power overshoot. And what prevents power overshoot is the increase during transmission. That is exactly right, and it says that in so many words in the summary of the invention in column 3. Now, what I understood the administrative law judge to be relying on in part, of course, is the fact that the references in the specification were to continuous increases in power, in the sense of not the discontinuous increases that are suggested by the stepwise. But I take it that your argument there is that whatever may have been the suggestion is to at least one type of increase in the specification. The claim language clearly encompasses both. Both. Our position is the claim language encompasses both. The judge's reason for restricting it was plainly an error, because the spec says it is the use of short codes that limits power increase. I personally believe that the word continuous doesn't necessarily restrict it to one or the other, to during the transmission or stepwise, although I acknowledge there is testimony to the contrary. So our case doesn't turn on that. I think I have run into my... Thank you. May it please the court, I think I'd like to start where this all started with the code limitation. Mr. Dunner said something I think that was important to get across, that he said that the code may be used but need not be used to increase... A spring code may be used but need not be used to increase the bandwidth of another signal. And that is exactly what the ALJ found when he said that it is a code that is used or intended to be used to increase the bandwidth of another signal. Not that it must be used, but... What does intended to be used in the context of a signal mean? Codes are inanimate objects. They generally don't have intents. Yes, Your Honor. I think the point is that it's... And that's why we try to elaborate a little bit in our brief that it is the type of code that may be used to increase the bandwidth of another signal, even if it is not used to do so. And the specification is clear on this. In the access procedure described in the specification, there's no spreading occurring at all during this part of the communication procedure. But that's their point, is it not? I mean, isn't this really the core of where they say that the ALJ may have strayed from what's in the specification? Right, Your Honor. The difference is a question of and what is the code doing at that time? And there is no dispute that there is no actual spreading occurring during this access procedure. Remember, this isn't the full communication procedure. This is simply when a subscriber unit, a handset, wants to contact a base station and say someone wants to make a call, and it does this access procedure. During this particular point of this entire call procedure, spreading codes are used, as Mr. Dunner said, for timing, for signaling, and also, very importantly, to identify the handset. The specification is clear, particularly at A409, excuse me, A408, in the entire first time, unique codes used to identify the subscriber unit. So the upshot is that by using these spreading codes during this point, the base station will already know which subscriber unit is involved in the communication. No later codes need to be, or no later type of code, if there were any later type of codes, which we don't think there are used in CDMA systems, need to be used. So this spreading code is used here and used later, but in this specific instance, the specification is clear that every single type of code that is described as being used is a spreading code. There are no other types of codes that are described as being used in the specification in the patents themselves or in the 010, which is incorporated. What do you make about the claim differentiation argument with specific reference to Claim 5, referring to the narrowing, presumably, of Claim 1 to a case in which the same code is a spreading code? Yes, Your Honor, and as the ALJ confronted this issue squarely, and he found that there is a presumption of claim differentiation, however, the specification overcomes that presumption by its clear definition, really, of what a code is. It says everywhere, code is spreading code. It never says anything else. Claim 5 is simply duplicative of Claim 1? That's what the ALJ found, Your Honor. Is that your position? Yes, Your Honor. It's redundant, which if you've ever drafted patents, it happens sometimes, but the specification is very straightforward. There is no question. And the ALJ's finding that spreading codes is an essential part of CDMA is really the linchpin here, in that CDMA, again, CDMA, or excuse me, spreading codes, are what identify the specific subscriber unit being used, or that is trying to make the communication, by spreading the data signals later on in the communication process, they enable multiple subscriber units to act at the same time over the same frequency band. But the short code doesn't spread, and that's not its role. Correct, Your Honor, and that's exactly the point, is that the function here of the short code is not to spread. Later on in the communication process, the same code, which is, again, the short code is described as being shorter, or a more compact portion of a larger spreading code. What is it that makes the short code a spreading code, in your view? What structurally makes it a spreading code? The fact that it is operating at a chip rate higher than the baseline rate, or what? Well, that is one reason, Your Honor, but... What else would be structurally necessary? If I were to look at a particular code, how would I identify this? Well, this is a spreading code. Okay, Your Honor, so again, yes, one reason that it's operating at a higher rate, so it will then, but more importantly, the 010 patent is actually much more explicit on this point as far as what a spreading code does. A spreading code is a specific type of code that not only is at a higher rate, but is also what they consider to be in... I can't find the exact language. Let's see, where is this? Oh, yes, on page A476 of the record, on column 2, lines 14 through 15, it says, these systems employ a family of orthogonal or quasi-orthogonal spreading codes, and that's important. But that's defining spreading code in terms of spreading codes. Besides the higher bandwidth, what is the structural feature of the spreading code that I should look for in order to make the determination whether it's a spreading code? Correct, and this is exactly the point. What features does a spreading code have to have to be a spreading code? Right, and so one is that it is at a higher bandwidth, that it is a quasi-orthogonal or quasi-orthogonal code, such that Nokia's brief actually explains this very clearly. I believe pages 11 through 15 of its brief, where it's explaining how data is spread and then recovered. And that's exactly, this type of code is what's used to form it. So unless the code has that orthogonality to it, it is not going to be able to be, for the data to be recovered later. So it can't just be any type of code that's at a higher bandwidth. It has to be a specific type of code. And that is what the ALJ recognized, in that these are not just any type of code. These are specific types of codes used exclusively, really, in CDMA systems that are spreading codes. Well, the short code is not orthogonal, is it? It's too short to be orthogonal, is it? Well, the short code, and this is at A409 of the 966 patent, at lines, again, 23 through 25, where it says, the short code is a sequence for detection by the base station, which has a much shorter period than a conventional spreading code. So again, these are spreading codes. It's merely a portion of a larger spreading code, but it's still a spreading code. If you have a spreading code that fulfills all the requirements and you take a piece of it, that doesn't lose any of its features. It's just a piece of it that's being used for a specific purpose here for this access procedure. I see my time has run, but... Yes. Thank you, Your Honor. May it please the Court, my name is Patrick Flynn, I'm the director of CDMA, and I'm going to talk to you about how to tell a spreading code, and how do you tell a spreading code from some other arbitrary sequence of ones and zeros. And we have to step back for a minute, and we have to understand how CDMA works. And this is what I think Ms. Valentine meant when she talked about a family of codes. For CDMA to work, for multiple users to share the same bandwidth, there has to be a pre-existing, pre-defined, known set of orthogonal codes that everybody knows what they are, everybody knows what they're using, and that's what is going to be assigned to an individual user. And so that the network and the base station can figure out one call from another one. So you have a pre-defined, finite set of codes that the system defines as the spreading codes for the system. And what the inventor said was, take those codes and either use an entire one as your access in the preamble, or use a piece of one in the preamble. Now to get to Judge Bryson, your question about claim differentiation, I'm actually going to differ from my colleague Ms. Valentine, because I do believe that Claim 5 is narrower in two very important respects than Claim 1. First of all, the language in question in that patent is not code, but signal. And yet the word they use that they want to say where the code is a spreading code, so we're talking about two different words, signal in the claim, and code in Claim 5. Another very important aspect of this is, let's assume that signal and code mean the same thing, and they don't, and they can't, for a reason I'll explain in a second, but let's assume that they do. The Claim 5 requires that it be an entire spreading code, whereas Claim 1 requires it to be either an entire spreading code, or part of a spreading code. So when they say the same code is a spreading code, they mean an entire spreading code, whereas Claim 1 is broader, it could be the entire spreading code, or a portion of the spreading code. But let me get... But let me make another important thing. In Claim 5 of the 847 patent, for example, or in Claim 1 of the 966 patent, again, recalling that what is accused of being, in essence, a shortcut is a signal, the language is where each of the successfully transmitted signals and the message are generated using a same code. Now, what is the same code that they accuse of infringing here? To understand that, we have to step back for a second and understand how this preamble is constructed. You have something called a signature, which is a 16-bit number, it's repeated 256 times, a scrambling code is applied to that, which is an arithmetic exclusive OR operation, produces a brand new number, that number is the actual preamble that's transmitted. So that's the signal that's transmitted. It is not the scrambling code, and it is not that signature. It is the result of the XOR combination of those two, which is, as we put it on our brief, like saying that when you transmit a 10, you're really transmitting a 10 and a 6, you're not. You're transmitting the result of an arithmetic operation. Now, their argument about how the same code is met, that limitation, is they say there is the same algorithm in the standard that generates that random scrambling code. So the code they're talking about being the same code is not the same sequence of ones and zeros. They're saying the same code there is the same algorithm. It's an entirely different apples and oranges comparison when they try to read the claim on what we do. Now, Judge Newman, you asked Mr. Dunner where in the specification it tells us what the meaning of code is, and Mr. Dunner pointed to you to the language in the 010 patent that talks about it as a sequence of chips. But as we pointed out in our brief, and it's important to remember, a sequence of chips is another way of saying spreading code because the 010 patent never calls anything a chip other than an actual spreading code in that family of predefined spreading codes I talked about, or a unit of time. Whenever it talks about the data that is being spread, that is the signal that is transmitted at the chip rate, but is the result of spreading the signal against the spreading code, they never call that a chip. They call that data at the chip rate. So the patent distinguishes chips from information at the chip rate. And when Mr. Dunner told you that that scrambling code, which is at the chip rate, could be used as a spreading code, he is absolutely wrong. And why is he wrong? He's wrong because if you took that scrambling code and tried to find it in that family of orthogonal codes that the system uses and knows and has to know, you won't find it there. So any old random set of ones and zeros at the chip rate can't be used as a spreading code because the system has to know the predefined universe of spreading codes. It's got to be one of them. But none of that is a short code, right? No, what you start with is, is it one of those predefined codes that the system uses for spreading? Because they have a relationship with each other. They have to interrelate, and it's got to be one of those universe of codes. And then as a short code, you simply take a sequence of it, a piece of it. And that's not any arbitrary sets of ones and zeros. And the inventors picked the system, they picked short codes as the place to get their codes because they know that they are known to the system and they don't have to generate a random set. They have hardware described in the 010 patent for producing this. Now I want to turn very quickly, if I can, to the increasing issue. I just want to make one comment on it. If I can ask you a question very briefly on this. The clean construction by the ALJ included this notion of intended to be used. Do you endorse that? I do, because what I think he was trying to capture was my point. Is that in a CDMA system, there are a set of codes that are used for, and are available to be used for spreading. And it's a finite, predefined, known set of numbers in a CDMA system. It has to be known. It has to be defined. And he's saying, those are the numbers you can use. Now, you pick one of those, you don't have to use it necessarily in one given instance for scrambling. You can make it do double duty as this access preamble, which is exactly what the specification called for. But it's got to be one of those known codes in the orthogonal family that the 010 patent talks about. You said used for scrambling? I meant spreading. I apologize. I absolutely meant spreading. You really don't. I really don't. Let's take that. I want to talk about this increasing issue. One thing that needs to be borne in mind, the judge was asked to construe the claims of four related patents, and some of those patents used the word increased. Others used the word increasing. And InterDigital asked Judge Rucker to construe those words to have the same meaning. And one of the reasons why, when you look at increased and increasing, the language of increased may by itself say that's fine, but when the ALJ was asked to construe increased and increasing to mean the same thing, it undercuts the argument of the plain meaning. And I see my time is up, and I thank the Court. Thank you. Any more questions? Any questions? Okay. Thank you. Let me address Mr. Flynn's comments first. He said that Claim 5 is narrower. He talked about the word signal versus the word code. Well, I don't understand his position because Claim 5 basically says the scribe unit of Claim 1, we're in the same code as the spreading code, and if you look at Claim 1 toward the end of the claim, it talks about the same code. The previous references are to the signal, and the ALJ made clear signal is the same as code. It's just transmitted codes. So I don't think that argument goes anyplace. Another point made is that he talked about the same code issue. I really didn't follow that. The same code issue was not before this Court. The Commission didn't reach it, and the ALJ didn't reach it on the question of infringement, and we pointed that out in our brief. He said the sequence of chips is another way of saying spreading code. Well, not the way the ALJ has defined it. The ALJ defined a spreading code as one which is intended to increase the width of a data signal. That's hardly the same as a sequence of chips. As to the scrambling code, we've pointed out document after document after document, including Nokia's own documents, that a scrambling code can be used to spread. In fact, if you take the scrambling code out of the Pratchett Preamble, you can use it. You can have the base station coordinated with the subscriber unit and use that scrambling code as a spreading code if you want to use it. It is not used in the Pratchett Preamble as a spreading code, although it's called a spreading code. And the interesting thing is that the ALJ said that on the question of claim construction, because these codes are used in CDMA systems, they are spreading codes, but when it came to infringement, even though the infringing devices are clearly used in CDMA systems, he said they didn't spread, clearly inconsistency between those two, and we pointed that out in our brief. As to Ms. Valentine, she said the ALJ's construction is used or intended to be used is the same as the type of code that can be used. I rest my case. There's no way in the world that those two phrases are the same, and I think the Commission is really running away from the ALJ's system. In any event, I think I've almost run out of time, and unless there are questions, I think I've covered the points I wanted to cover. Any questions? Okay, thank you, Mr. Dyer. Now, before the pardons go, all of the briefs have designations of confidential information. They aren't particularly consistent, but some of them seem to go right to the heart of things that we would want to refer to in explaining our decisions when we reach our decision. So I wanted to ask if those of you who have designated confidential information would take another look and see, we would respectfully ask to be preserved, but some of it seems to go a bit farther than perhaps is necessary. And so if you could get together, maybe prepare three sets of new briefs with new designations of what you view as confidential and not to be mentioned in any opinion you might come up with. It would be very helpful to us. If you want to keep it as it is, let us know. We will do that, Your Honor. Sure. I mean, I may be able to right now, Your Honor, say that anything having to do with what the standard requires Well, we don't know who marks what. Well, I'll coordinate with Mr. Flynn. We're on speaking terms. We are meeting. Good. And if you'll collaborate on that, we don't forget. Thank you. Thank you, Your Honor.