Oral argument — Utstarcom v. Starent Networks

To the benefit of all counsel, please be assured that we have read your briefs and we're familiar with the facts. So, your time is limited, so please get right to the legal issues you would like to press upon us today. For those who may not be familiar with our light system, when the yellow light comes on, that means you are beginning to use up your rebuttal time. The red light means you're out of time, both for your original argument and if you want a rebuttal, for a rebuttal. We'll hear argument first then in number 06-1295, UTstarcom v. Starent Networks. Ms. Esty for the amendment. Good afternoon, Your Honors, Mr. Leffinger, Ms. Bryson, and Judge Lynn, members of the Bar and those in attendance. My name is Joanna Esty. I am associated with the firm of Venable and I represent UTstarcom in this matter. UTstarcom has appealed the finding of non-infringement by Starent Networks Corporation of its patent, U.S. Patent 6628671. In particular, UTstarcom contends that the claim interpretation of the lower court is incorrect and that the lower court judge incorrectly and inconsistently applied the claim interpretation to the election infringing device. If you look at the abstract, the abstract in relevant parts says that the invention deals with a network access or providing remote access to an IP network. In order to do that, there is a PPP connection for a remote client and that PPP connection and the PPP state associated with it is switched rather than reconnecting and renegotiating the state. And the abstract says this context switching of the active PPP session allows the mobile user to seamlessly move about the wireless network without having to renegotiate the link control protocols and network control protocols every time they move out of range of one radio tower and into range of another. Let me ask you this. Isn't the crux of the debate here a question about whether the patent covers a system that has two PPP states compared with the accused structure that has one? At bottom, isn't that what this comes down to? Well, Your Honor, the crux of the appeal is whether or not a PPP session having the same PPP state as a previous PPP session is considered the same session. If I said a moment ago state, I meant session. I'm sorry, Your Honor. Let me start again. Isn't this, at bottom, a question of whether there are, in the accused structure, there's one session, whereas the patented structure describes a system that has two different sessions, the first of which will go dormant, the second of which is established, and then the PPP state is transferred from the first session to the second session? Isn't that, at bottom, what the difference is, that in the patent system, there is the claim and the specification talks about a structure or a system that has two different sessions, whereas the accused operates under the 3G protocols with one session? Your Honor, I believe that that's the issue as it was raised on the appeal, but I believe that you have to look at the underlying layer on PPP session and exactly what a PPP session is. A PPP session is a negotiated state. That means I've negotiated my link at one end, I've negotiated my link at the other end, and I have that link information that allows me to pass data along it through an IP address, and it's the passing of that IP address without having to renegotiate the link at one end and the link at the other end that really is the essence of the invention. Well, but are you saying, did I hear you say in there or imply that a PPP state is also in effect the same thing as a PPP session? That is to say, if you have the same PPP state, you necessarily have the same PPP session? Yes, Your Honor. And I – if you read the specification, you can see that the patent attorney that drafted the application struggled with how to differentiate what was time-differentiated but actually had the same PPP state information. And he doesn't do it really in a way that's completely consistent, but it is understandable to one ordinary skill in the art. If you look at – let me just – if I can interrupt you because I want to pursue this question of the equivalence of equivalences of that where the parallel between state and session. Now, in the abstract, you read from the abstract and you keep going a little bit, you come across the language that says that the invention involves the initiation of a new active – well, I'm sorry – that there is a switching the PPP state to a new session, which now I understand your argument is, well, session there doesn't always refer to PPP session. But what's your explanation for that language consistent with the view that PPP state means the same thing as PPP session? What the patentee is trying to get across, Your Honor, if you look at the abstract again, the context switching of the – in the PPP session is what we're looking at. So if you're switching the context, you're taking the state information and you're passing it on. But in that abstract, what is the meaning of new session? It sounds like session number two as opposed to same session. And that's why I'm saying that the author of the application struggled with how to define a PPP session when it has the same state information. He didn't make very clear that new doesn't mean new. Well, Your Honor, he did say in the specification he talks about the fact that his session – he reconnects the PPP session. And he does that – if you look at column two at lines 26 through 32, he talks about the way that the invention works. And he's talking about reconnecting the session. So he's talking about essentially passing the state information, the link control protocols with the network control protocols and the IP address. Why didn't he say it switched over to the same old acting PPP session? That's what you're telling me. And that's what he said. Yes, Your Honor. If he was struggling that hard, why didn't he quickly introduce the word new again? Why didn't he just say over to the same old active PPP session? Your Honor, he wanted to distinguish the fact that the session continues at a second port. And that's why he's saying you reconnect the session at a second port. He's gone from an active PPP session that went dormant, i.e. went inactive, right? Yes, Your Honor. Now, is that happening in the accused device? Does the accused device go to sleep for a while and then wake up again? Yes, it does, Your Honor. The dormant is the lack of active data transfer back and forth. And the active data transfer is information going back and forth. So the fact, if I go out of range of the network access server and the cell tower, if I go out of range of the tower, I'm actually dormant. Even though I'm still trying to ping the mobile device, the mobile device is not responding because I'm in a tunnel. That's the definition of dormancy, is that the mobile device is not actively communicating back and forth. If you look at the definition as the judge applied it to session and PPP session, what she did was she said, every session is a PPP session, you claim one, therefore I get to other sessions. You know, I struggled with this because I said, what was the patent attorney doing when he was drafting this? Where was he trying to go with this? Why didn't he put in there, you start a session by a communication from a cell phone to a tower and down to your base station. The reason for that is that he wanted the network access server and the methodology of the network access server to be found in all of the limitations of the claim. So that you didn't have a limitation where you had to reach out for a mobile phone. We don't need to decide any of this if we agree with the respondent that you waived the right to challenge the court limitation issue, right? Your Honor, I don't believe those... What's your best argument is that? Why did you actually raise the court limitation challenge in your blue brief? Your Honor, we raised the court limitation in the docketing statement under section 6 where the court properly found on, infringement based on the claims limitations. We raised it in the statement of issues. Right, but that wouldn't allow you to bring into the brief any issue. Your Honor, we also... If you're essentially saying, well, we just challenged a similar judgment. Your Honor... And the respondent is blindsided because they don't quite know what you're going to bring up in your brief. Your Honor, we did argue in the opening brief, and we put into the claims, that the first court is the connection between the network equipment and the PDSN. Where is that? That's in the claim chart, page 33, rows 2 and column 2. Then we also cited to our declaration of Paul Francis. And then we cited to the historic call flow at A2093. And then in the opening brief, we explained after the claim chart, that there's a time period for the PPP section associated with the first court. And another time period for the same... Where's that? Where's that? On page 37, Your Honor, of the brief. Then we also looked at association with the location of the second court. But, Your Honor, I think that the reason the lower court judge got it wrong is... And you have to go to... You have to move from the direct infringement part of the opinion into the part of the opinion that deals with doctrine of equivalence. In the doctrine of equivalence section... And I racked my brain on this. The doctrine of equivalence section, she actually... She defined court as a logical or physical connection. Your Honor, we have an argument on the doctrine of equivalence at pages 43 through 45 of our brief. But the doctrine of equivalence argument that the judge made, she used two different definitions for court. And both of those attached court to the network... The new call message setup that was coming in to the network access server. That's what she equates to. She has two different definitions of the function of court. So it's no wonder she couldn't make a decision. She didn't understand what a logical connection is. She actually says at pages 8, 19, and 8, 20... She first talks about the function of the court being the acceptance of the new call message. And then she talks about the function of the court at page 20. She's got a separate definition for it. Page 20, she says it's differentiated between the data packets that flow in and flow out of the ST16 with a particular user in session. That's the function she uses on page 20. And then the function she uses on page 19, she defines that court function as being receiving after said PPP session has become dormant. Well, that function's inconsistent because she's read the claim incorrectly by determining that session and PPP session are the same. Thank you, Your Honor. Very well. I'm going to take the rest of your time. We'll hear from the appellee, Mr. Yarks. Thank you, Your Honor. Let me address the waiver issue first and why it's important in this case. In the gray brief, the appellants have put forth a theory of infringement of ports, which is different from any one they had ever raised. In the gray brief for the first time, they suggest that a port includes something called an Ethernet port. However, if you look at what they urged before the district court in summary judgment, the word Ethernet port was never there. Before the district court in summary judgment, opposing summary judgment of non-infringement, they said that the port was some sort of memory and some sort of software. That's what the court addressed. That's what the court found wasn't there. That was even a change, what they urged in the preliminary injunction motion, which was just this thing called an RP session. So the reason it's important in this case, especially, that the waiver argument be used to preclude them from addressing that issue is they have never pointed out that it's legitimate. If you may interrupt just for a second. As I listen to you describe the three different definitions, it's not clear to me that they're not consistent, i.e., that you can describe the port in each of those ways as being some feature, either functional or structural, of the port as they conceive it. The port, I take it, what's going on. I mean, I have full appreciation for this. But my understanding is you're taking port from the old physical device, which was one to a customer, into a system which is software-driven that allows a single, what used to be a single port, to be used, thanks to some fancy switching, by multiple incoming signals. Isn't that basically what you got? Packets are going to be run down the same circuit. Isn't that essentially it? Well, not quite, Your Honor. I just don't think so. The type of ports that are referred to here are specific ports that the patent talks are not associating with, among other things, with PPP session. And in addition, it's something, for example, about the way the new call message is received. When they urged at the beginning, in a preliminary injunction, that that port could be in this thing called an RP session, the district court promptly found that that made no sense. Why? Because the new call setup message that the claim requires to be received over the port is the very message required to set up the RP session. So you didn't have a port until you had the session, yet the claim required you to receive the signal to set up the session over that port. So it made no sense logically. They didn't attack that. We come now to the claim construction hearing. And after the claim construction, they were allowed to put in what they called their final infringement contentions. And their final infringement contentions, they abandoned that argument. They said, no, really, what we're talking about. In addition, one of the reasons they abandoned that argument is the court said, this RP session you're talking about is really outside of the network access circle. Yet the claim is required to be inside the network access circle. So they changed their theory. Their new theory was, it was some undescribed memory. They didn't say what the memory was, just some undescribed memory. And it was some undescribed software. They did not say what the software was. The court rejected that as well, too. They said, well, that can't be the port. Because there's nothing that indicates how this software that's inside can receive anything like the new call center message. Now, your honor is correct in that perhaps what's being referred to here by the court is some logical idea in which you can multiply several things off. But that's not what they claim. But that is what I understood their position to be with respect to court. Now. Say it a second later. Now being the gray brief, or now being the? Yeah, now. No. Now being the gray brief is, it's some RP session, which, again, they never explained why the court was wrong in rejecting that, plus a physical port called the Ethernet port. And there's no proof. They put no proof in here that they ever explained to the district court that that was their basis of infringement. It's not in their final infringement contingent. It wasn't what they put in front of the court in response to the motion for summary judgment. They're essentially coming before this court here in the gray brief with a new theory of infringement, which is why the waiver is particularly important. Why they didn't preserve it. Why they never explained to this court why the district court had erred, either in their claim construction report or in finding that there was no issue of fact. And that's why it's particularly important. Because there is evidence of record, but of course we weren't allowed to present it, that our device does indeed not associate PPP sessions with courts. I'd like to go back to the question about PPP state and PPP session and the distinctions that you see between a PPP state, which as I understand it is a series of negotiated protocols, and a PPP session, which I understand is a connectivity, if you will, between a device and the network access server. Can you clarify or comment on that for me? Sure. Your Honor is exactly correct. What the PPP state is, is a set of variables which define essentially how you're going to communicate. So for example, and this is not what it is, but just because it's easier for me to understand. We're going to talk in French. We're going to use complete sentences. And you're not going to start until I say Roger over and out. That's the state variables explain how we can talk to each other. The session is a specific instance in which you and I are going to talk using those rules. So the state is necessary to define the session, but the same state can define many sessions. It can define our communication here. And if we want to have the same rules, the exact same state can define my communication with Judge Bryson. So the state is not the same in the session. The state is merely the collection of variables, or as indicated here in Column 2 of protocols, Column 2 lines, I guess, 8 through 13 of protocols, that define how we're going to communicate. So it would be entirely consistent to have one state apply to three different sessions so that you don't have to waste your time. If I move over here so I can use the microphone, then I don't have to waste my time renegotiating. We're going to talk in French. We're going to talk in complete sentences. And I'm going to say Roger over and out. I guess this is all relevant in the context of the various standards that have applied over the years. As Ivan said, the 2.5G standard sets up a system where, in effect, you have a distinct connection. And during periods where there isn't data traffic, you have underutilized bandwidth. Therefore, some provision had to be made to sort of put that system in a dormant state and use that bandwidth for other calls at the same time. And that the 3G standard is unlike that because there's, in effect, data going simultaneously all the time. And that accounts for the difference between what's described in this patent and what is implemented by the accused structure. Am I correct? That accounts for much of the differences. The fact of the matter is there is no dormancy of PPP sessions because there's no need to enter that state to reclaim any resources. This frame really used to exist. It needed to be reclaimed because if it wasn't being used, your bandwidth went down. And there's a new Ethernet which, if you look at it, is like a train with a bunch of boxcars. If you don't use a boxcar, the next person will. It's a multiplexed function. Multiplexed function. What answer? Well, just the last question is, well, given those distinctions, I guess the debate here all along has been whether the claims, even though written in the context of a system designed for 2.5G, nonetheless are broad enough to cover a system that can implement the 3G standard. And the claim talks about PPP session and then seemingly another session and then the handoff of the PPP state from the first session to the second session. And the question is, is that appropriately descriptive of the 3G system that's accused? No, it's not. Because the undisputed facts are that there's only one session in 3GPP. There's no dispute about that. That's the position they originally took in the preliminary injunction motion. It's what the documents say, and it's what uniformly exists, except if you magically now divide the one PPP session into two, as they try. And second of all— That one session doesn't go dormant? It never goes dormant. The PPP session doesn't go dormant. There's no need for it. Other stuff goes dormant, which they refer to, but it's not the PPP session. Your mobile device may go dormant, but the session's still active. When you say the session is active—and this really gets to the heart of it for me— what is it that is different between a session that is active and a two-session system in which the state remains the same? That is to say you have—well, you understand. Well, I think you're confusing two things. The difference between a session going active and one going dormant is— and this is the agreed-upon construction of the court— is a dormant PPP session is one in which not only is there no active data transfer, but it has to be for a predetermined period of time. So you have to know ahead of time that after five milliseconds, there's no traffic. That's dormant. And in addition, that allows the reallocation of resources. So there's other requirements other than you're not talking. For example, if we're on a telephone call and we stop talking for a while, it's still active even though we're not talking. But that's what I'm really—how do we know we call that active if we don't have to do this elaborate renegotiation because of the appellant's invention? Because you're not able now to reallocate resources. Even though we haven't talked for a while, the phone company can't come in and say, I want to use that bandwidth for something else. They're still active according to the system. It's still an active PPP session. That's the beauty of 3G. What would happen with respect to the 2.5G that makes for two sessions as opposed to one? Is that when you have this single circuit, which was previously dedicated to session number one, you say, well, the session is then transferred to a different circuit? Is that what marks the difference? Yeah, but let me explain this. That doesn't have to be dormant. The issue of 2.5G is after a certain period of time, you can go dormant, so you can use that connection for something else. Now, what happens is the reason you have two different sections is because when a new call comes in and comes into a different port, that now is a different session because it's in a different port. So that's what makes the difference in sessions is the fact that it comes through a different port. That doesn't necessarily mean a physical different connection. It's just a different port which can be an electrical address or an address in a computer-driven system. I think back in the 2.5G area, and I will confess to you that now we're getting beyond my understanding of technology, but I believe that back in the 2.5G area, it was semi-random which physical port it came in on because, again, it had to come not only to a different radio tower but to a different base station, and it possibly could have come in through the same physical port, but if it came into a different physical port, it was a different session. And in the 2.5G, it was the physical realm. So there's two different things. There was the two different sessions which was required, but in addition, there was a dormancy which you needed to do to release these link layers. I'm sorry, prime relays, so somebody else could use them. Now, in the multiplexed system, the 3.0, I take it that as soon as you and I stop talking, the system recognizes that we're not using the boxcars and it throws somebody else's conversation into the boxcars. Exactly so. And maybe they interspersed our conversations. Maybe they interspersed 10 or 20 conversations. All that can be done by, I gather, the software that does the directing back and forth based on the addresses as to which boxcar has whose conversation chunks. I believe that's it. Now, would that be deemed to be a kind of dormancy on our part, a dormancy on the part of the PPP session, a kind of dormancy in the communication that would be recognized as an opportunity for the multiplexer to put in somebody else's data? It's not recognized that in the area, in the field. Nobody calls it dormancy. What do they call it? They don't call it anything. They just call it maximum use of available bandwidth. So the software does not pay attention to the fact of whether we're talking or not other than to simply see that there's open space to be used. My understanding of it, again, I will hedge this with the fact that I'm no longer an engineer. It's not my field. But you've got a lot of people that want to come in if there's a complicated algorithm that decides who's first, who's second based on priorities, based on a lot of stuff. And if you're just not in line, they don't even know that you're dormant. They just know that you're not in line. Okay. Is there a difference between a PPP session and a session? And if so, is that difference the fact that a PPP session is one in which a PPP state has been established? I think in the context of these claims, there's no difference between a PPP session and a session. Outside this pattern, or in the context of this pattern, there's no difference. Outside this pattern, I would acknowledge that the word session has a broader meaning. But in this pattern, every time it's a session, it's talked about a PPP session. Well, if session has a broader meaning, then does PPP session have a narrower meaning? And is that narrower meaning determined by the existence of a PPP state? The PPP session means that there is a negotiated PPP state. That's correct. But the only session in the claims where you talk about transferring the state to a session, the only session that can receive a state is a PPP session. But isn't the point that the appellant is making that the claim talks about establishing a PPP session, and then when later on the claim talks about another session to which the PPP state is transferred, it doesn't refer to that as a PPP session. And one can read into that that, well, yes, the distinction was being made in the claim that you have one PPP session established, and then later on you establish another session, in the broadest sense, and it doesn't become a PPP session until the PPP state is transferred. The reason I would disagree with that is because when you read the several other instances of the specification, they're in our brief, but they're in column two, for example, lines 28 through 34, column two, lines 53 through 57. They talk not about it doesn't become a session until you get the PPP state. They call it a PPP session as you're transferring it in. It does seem throughout the specification, repeatedly, the second session is referred to as a PPP session. Yes. And referred to not as being defined by the transfer of the PPP state, but simply a second session where the first session remains and remains as a dormant session. Exactly so. It is more logical to assume that the patent draftsman just used session as a shorthand for PPP session than to go through the machinations that you have to go through for a comments argument. Thank you. I will distort. I think you have about two and a half minutes. We'll give just that amount of time. Great. Thank you, Your Honor. I would just like to address the issue of whether PPP session and session are the same thing. I think that the communication sessions and when I read the patent, when it says you make a call, a call is initiated by the mobile device sending a communication, that's considered essentially the beginning of a session. But the drafter of this application did not want to incorporate into the claim the mobile device. So in the second limitation, it says receiving after the PPP session has become dormant. That's when the session starts. And it starts with a new call setup message. So session is broader than PPP session. And then you have the transfer into this broader session that started with the receipt of the new call message, the broader session going where the PPP session is transferred with the same state information so that the state information can continue, the IP address can continue, and the communication can continue without renegotiation of the PPP state. But even if that's true, doesn't that mean that there are two sessions? There's the original session and then there's the subsequently established session. According to the way the claim is drafted, you have a transfer of the PPP session because you have a transfer of the state information. But, Your Honor, the transfer of the PPP state is the transfer of the PPP session. If you interpret that session to not be PPP session, it's the broader session that starts with the receipt of the new call message setup and continues on. The trouble that I have with that argument, quite candidly, is that the specification repeatedly refers to that second session as a PPP session. He does refer to it as new, but he also refers to it as being transferred. And the claim itself— It's not new vis-à-vis the original session. I mean, the use of the term new certainly suggests different. Yes, Your Honor. And second. But it's a new because it comes into a new port, not just a new PPP session. And then, Your Honor, just one last comment. Even if you interpreted it to be two PPP sessions with the same state, under the doctrine of equivalence, because you are using the same PPP state information, you can say that it functions the same way, in the same means, to get the same results. The court would not allow that argument to be made because she limited the scope of the claims by what essentially wound up being a prosecution disclaimer rather than a prosecution history estoppel. And prosecution disclaimer is the matter of law, and it needs specific language in the term of the prosecution history disavowing that claim scope. And she didn't go there. Thank you. Case is submitted.