COMMIL USA, LLC, Plaintiff–Appellee v. CISCO SYSTEMS, INC., Defendant–Appellant.
This case returns to us on remand from the Supreme Court. See Commil USA, LLC v. Cisco Sys., Inc., 135 S.Ct. 1920 (2015). While our previous opinion had remanded this case to the district court for a new trial, Commil USA, LLC v. Cisco Sys., Inc., 720 F.3d 1361, 1372 (Fed.Cir.2013), when we received the case back from the Supreme Court we granted Cisco's request to retain the case and address Cisco's remaining non-infringement arguments which we had declined to address in our previous opinion. ECF No. 101. We now conclude that substantial evidence does not support the jury's finding that Cisco's devices, when used, perform the “running” step of the asserted claims. The district court's judgment is therefore reversed.1
Because much of the relevant background is set forth in our previous opinion, we summarize only briefly here the facts and posture of this case.
Commil owns U.S. Patent No. 6,430,395 (“ '395 patent”), which relates to a method of providing faster and more reliable handoffs of mobile devices from one base station to another as a mobile device moves throughout a network area. In 2007, Commil brought this action against Cisco, which makes and sells wireless networking equipment. In a first jury trial, Commil alleged that Cisco directly infringed the '395 patent by making and using networking equipment, and also that Cisco induced its customers to infringe by selling them the infringing equipment. The jury concluded that Commil's patent was valid, that Cisco was liable for direct but not induced infringement, and awarded $3.7 million in damages. Commil then filed a motion for a new trial on induced infringement and damages, which the district court granted. The second jury concluded that Cisco was liable for induced infringement and awarded $63.7 million in damages.
Cisco thereafter appealed to us, raising a number of issues. A split panel affirmed in part, vacated in part, and remanded for a new trial. Commil, 720 F.3d at 1361. Because we were remanding for a new trial, we did not reach certain of Cisco's arguments on non-infringement and damages. Id. at 1372. Following that decision, Commil sought certiorari on the limited question of whether a defendant's belief that a patent is invalid is a defense to induced infringement. The Supreme Court granted certiorari, reversed the majority's decision on that issue, and vacated and remanded back to us.
Upon return to this court, Cisco requested that we address its non-infringement arguments that a majority of this panel previously had declined to decide. Specifically, Cisco contends that Commil cannot prevail on its infringement charges because neither Cisco nor its customers directly infringe by performing both method steps. We granted Cisco's request, and now address those arguments. We review the jury's determinations of infringement for substantial evidence. ACCO Brands, Inc. v. ABA Locks Mfrs. Co., 501 F.3d 1307, 1311 (Fed.Cir.2007).
Claim 1, the sole independent claim of the '395 patent, contains two steps: a “dividing” step and a “running” step. Specifically, Claim 1 reads:
In a wireless communication system comprising at least two Base Stations, at least one Switch in communication with the Base Stations, a method of communicating between mobile units and the Base Stations comprising:
dividing a short-range communication protocol into a low-level protocol for performing tasks that require accurate time synchronization and a high-level protocol which does not require accurate time synchronization; and
for each connection of a mobile unit with a Base Station, running an instance of the low-level protocol at the Base Station connected with the mobile unit and running an instance of the high-level protocol at the Switch.
'395 patent col. 39 ll. 16–29.
We begin with the running step.2 The district court construed the running step as requiring “for each connection of a mobile unit with a Base Station, running at the Base Station a copy of the low-level protocol supporting only that connection and running at the Switch a corresponding separate copy of the high level protocol supporting only that connection.” J.A. 2 (emphasis added). Cisco contends that this step is never performed when its system is used, because its system employs a single copy of the protocol to support all the connected devices.
We agree with Cisco. First, Cisco's engineer testified that Cisco's system runs only one copy of the protocol to support multiple connected mobile devices. Specifically, he testified that Cisco's system “needs and uses only one copy of the protocol to support all 2,000 of those mobile devices․ The standard was written in a fashion that only one copy of the protocol is necessary to implement the standard.” J.A. 6268.
Although Commil sought to establish otherwise through expert testimony, that testimony falls far short of supporting the jury verdict here. Specifically, Commil's expert testified that, at most, Cisco's devices track separate state information for each connected device. He opined that: “[T]he instructions, the protocol ․ it's a state machine. So this communication state that it is invoking in that communication represents a copy of the protocol that's unique to that one device that it's communicating with.” J.A. 6176; see also id. (“[A]ll of that information, with regard to that state that it's using for the communication, is its own copy of the protocol that's unique to that one communication path․”). But tracking separate state information for each device does not provide substantial evidence to satisfy a limitation that requires running a separate protocol copy for each device. Moreover, when pressed, Commil's expert conceded that Cisco's system supports multiple connected devices at the same time, but only runs one copy of the protocol at any one time. J.A. 6204, 6018. In light of this testimony, a reasonable jury could not have found that Cisco's devices run a separate copy of the protocol for each connected device.
Nor is the jury's verdict supported by Commil's additional contentions. Commil argues that overturning the verdict requires reading a “simultaneity requirement” into the claims that is not there. Not so. In finding for Cisco, we do not require simultaneity; rather, we simply adhere to the construction of the district court, which requires a separate copy of the protocol to be run for each connected device. Commil also contends that discarding its expert's testimony about state information requires improperly presuming that “protocol” and “information” are mutually exclusive. But in making this argument, it is Commil who departs from the governing constructions in this case. Specifically, the district court construed “short-range communication protocol” to mean “a set of procedures required to initiate and maintain short-range communication between two or more devices.” J.A. 1. In all of Commil's expert testimony, nowhere does Commil's expert provide evidence or reasonable support for his opinion that tracking separate state information for each device is the same as running, for each connected device, a separate “set of procedures required to initiate and maintain short-range communication between two or more devices.”
Because we find none of Commil's other arguments persuasive, we conclude that substantial evidence does not support the jury's necessary finding that Cisco's devices, when used, perform the “running” step of the claims. Because this conclusion precludes liability under either of Commil's direct or inducement theories, we reverse the judgment of the district court.
1. We do not disturb, and therefore reinstate from our original decision, everything other than (i) the portion affected by the Supreme Court's decision, Section II(B), and (ii) our direction that the case be remanded for a new trial.
2. Because we find the running step is not performed by either Cisco or its customers, we need not reach the party's additional arguments on the dividing step.
PROST, Chief Judge.