Invt Spe LLC v. Itc

• Court: Court of Appeals for the Federal Circuit
• Decided: August 31, 2022
• Docket: 20-1903
• Status: Published

Opinion

Case: 20-1903 Document: 98 Page: 1 Filed: 08/31/2022

United States Court of Appeals for the Federal Circuit ______________________

INVT SPE LLC, Appellant

v.

INTERNATIONAL TRADE COMMISSION, Appellee

HTC AMERICA, INC., HTC CORPORATION, Intervenors ______________________

2020-1903 ______________________

Appeal from the United States International Trade Commission in Investigation No. 337-TA-1138. ______________________

Decided: August 31, 2022 ______________________

JEFFREY A. LAMKEN, MoloLamken LLP, Washington, DC, argued for appellant. Also represented by LUCAS M. WALKER; SARA MARGOLIS, New York, NY; JOHN K. HARTING, BRENDA L. JOLY, CYRUS ALCORN MORTON, CHRISTOPHER SEIDL, Robins Kaplan LLP, Minneapolis, MN.

RICHARD P. HADORN, Office of the General Counsel, United States International Trade Commission, Washing- ton, DC, argued for appellee. Also represented by DOMINIC Case: 20-1903 Document: 98 Page: 2 Filed: 08/31/2022

L. BIANCHI, WAYNE W. HERRINGTON.

CHARLES M. MCMAHON, McDermott, Will & Emery LLP, Chicago, IL, argued for all intervenors. Also repre- sented by MARTIN BADER, STEPHEN S. KORNICZKY, ERICKA SCHULZ, Sheppard, Mullin, Richter & Hampton LLP, San Diego, CA; EDWARD V. ANDERSON, Palo Alto, CA. ______________________

Before NEWMAN, TARANTO, and CHEN, Circuit Judges. Chen, Circuit Judge. BACKGROUND Complainant INVT SPE LLC (INVT) appeals from a determination by the International Trade Commission (Commission or ITC) in Investigation No. 337-TA-1138, Certain LTE- and 3G-Compliant Cellular Communications Devices, that respondents Apple Inc., HTC Corporation, HTC America, Inc., ZTE Corporation, and ZTE (USA) Inc. did not violate 19 U.S.C. § 1337 (section 337) by the impor- tation and sale of personal electronic devices, such as smartphones, smart watches, and tablets. INVT’s com- plaint alleged that these devices infringed five INVT pa- tents, only two of which are at issue in this appeal—U.S. Patent Nos. 6,760,590 (’590 patent) and 7,848,439 (’439 pa- tent). In a final initial determination (FID), the adminis- trative law judge (ALJ) determined that the accused devices did not infringe claims 3 and 4 of the ’590 patent and claims 1 and 2 of the ’439 patent. In the Matter of Cer- tain LTE- and 3G-Compliant Cellular Communications Devices, Initial Determination on Violation of Section 337, No. 337-TA-1138, 2020 WL 1504741, at *2 (Feb. 18, 2020) (FID). The ALJ also determined that INVT had failed to meet the technical prong of the domestic industry require- ment as to those claims. Id. INVT petitioned the Commis- sion for review of those findings, J.A. 1787–1815, 1831–56, which the Commission decided not to review, In the Matter Case: 20-1903 Document: 98 Page: 3 Filed: 08/31/2022

INVT SPE LLC v. ITC 3

of Certain LTE- and 3G-Compliant Cellular Communica- tions Devices, Notice of a Commission Determination to Re- view in Part a Final Initial Determination Finding No Violation of Section 337 and, on Review, to Affirm the Final Initial Determination’s Findings of No Violation; Termina- tion of the Investigation, No. 337-TA-1138, 2020 WL 4582313, at *2 (June 1, 2020) (Commission Decision). The Commission affirmed the ultimate finding of no violation of section 337. See id. at *3. INVT appeals from this final determination. All five respondents intervened, but Apple, Inc., ZTE (USA) Inc., and ZTE Corporation have since withdrawn as parties, leaving HTC Corporation and HTC America as intervenors. See ECF Nos. 67, 93 (orders grant- ing motions to withdraw). We affirm the Commission’s determination that there was no section 337 violation with respect to the ’439 patent because INVT failed to show infringement and the exist- ence of domestic industry. We agree with INVT’s argument on appeal that the asserted ’439 claims are drawn to “capa- bility.” However, we disagree with INVT on infringement. For infringement purposes, a computer-implemented claim drawn to a functional capability requires some showing that the accused computer-implemented device is pro- grammed or otherwise configured, without modification, to perform the claimed function when in operation. We affirm the noninfringement finding in this case because INVT failed to introduce any evidence to establish that the ac- cused devices, when put into operation, will ever perform the particular functions recited in the asserted claims. We find the Commission’s determination with respect to the ’590 patent moot based on the patent’s expiration, and thus vacate and remand as to that patent. A. ’590 Patent Before this decision issued, the ’590 patent expired on March 5, 2022. See Letter from the Office of the General Counsel Attorney for ITC, ECF No. 78; Appellant’s Suppl. Case: 20-1903 Document: 98 Page: 4 Filed: 08/31/2022

Br. 1, ECF No. 84. For the reasons discussed, infra, the appeal as it relates to the ’590 patent is moot. We vacate the Commission’s decision as to that patent and remand with instructions to dismiss as moot the relevant portion of the complaint. B. ’439 Patent The ’439 patent relates to wireless communication sys- tems, specifically an improvement to adaptive modulation and coding (AMC), which is a technique used to transmit signals in an orthogonal frequency division multiplexing (OFDM) system. ’439 patent col. 1 ll. 7–14. In an OFDM system, the frequency bandwidth is di- vided into subcarriers. A subcarrier is a narrow subdivi- sion of a communication system’s available frequency spectrum (bandwidth). Id. col. 1 ll. 25–26. Groups of sub- carriers in neighboring positions within the frequency do- main are referred to as subbands. Id. col. 2 ll. 18–22. AMC involves adjusting parameters, such as a modulation scheme or a coding rate, in response to changing conditions that impact the channel quality. Id. col. 1 ll. 34–52, 65–67. The prior art included AMC based on subcarrier and sub- bands divisions of the communication system bandwidth. See id. col. 1 l. 53 – col. 2 l. 49. The ’439 patent is directed to AMC based on subband groups. See id. col. 5 l. 9 – col. 6 l. 44, col. 7 l. 32 – col. 10 l. 26. This means that that the modulation scheme and coding rate are determined per subband group as the min- imum unit of adaptivity, rather than per subcarriers or subbands. Id. col. 7 l. 32 – col. 12 l. 24; see id. col. 2 ll. 4–8; id. col. 2 ll. 12–25, col. 7 l. 65 – col. 8 l. 2, col. 8 ll. 41–48, col. 10 ll. 21–26. Subband groups are made up of multiple subbands, although not necessarily subbands in neighbor- ing positions. See id. col. 7 ll. 43–46; col. 10 l. 26 – col. 11 l. 3. A subband group might consist of a plurality of neigh- boring subbands, id. col. 10 ll. 33–49, Fig. 8, or a plurality of subbands at predetermined intervals, id. col. 10 ll. 50– Case: 20-1903 Document: 98 Page: 5 Filed: 08/31/2022

INVT SPE LLC v. ITC 5

61, Fig. 9, or even all of the subbands as a single subband group, id. col. 10 l. 62 – col. 11 l. 3, Fig. 10. Claim 1 of the ’439 patent recites: 1. A communication apparatus comprising: [a] a channel estimating section that carries out a channel estimation per subband; [b] a parameter deciding section that decides mod- ulation parameters and coding parameters per sub- band group comprised of a plurality of the subbands, based on a result of the channel estima- tion per subband; [c] a parameter information transmission section that transmits, to a communicating party, param- eter information indicating the modulation param- eters and the coding parameters decided at the parameter deciding section; [d] a receiving section that receives a signal con- taining data modulated and encoded on a per sub- band group basis at the communicating party using the modulation parameters and the coding param- eters of the parameter information transmitted at the parameter information transmission section; [e] a data obtaining section that demodulates and decodes the received signal received at the receiv- ing section on a per subband group basis using the modulation parameters and the coding parameters decided at the parameter deciding section, and ob- tains the data contained in the received signal; and [f] a pattern storage section that stores in advance patterns for selecting subbands constituting the subband groups wherein the parameter deciding section decides the modulation parameters and the coding parameters per subband group comprised of the subbands selected based on the patterns stored Case: 20-1903 Document: 98 Page: 6 Filed: 08/31/2022

in the pattern storage section. (bold bracketed letters added). A “communication apparatus,” as recited in claim 1, can be a user device, and a “communicating party” can be a base station. See id. col. 2 ll. 54–60 (describing a base station as the transmission side and a mobile terminal as the receiving side). The ’439 patent describes a user device (receiving side) determining the adaptive parameters and sending the parameters to the base station (transmission side); the base station encoding data using those parame- ters and sending the encoded data back to the user device (receiving side); and the user device (receiving side) decod- ing the data using those parameters. See id. col. 9 l. 13 – col. 12 l. 24. C. LTE Standard The accused devices are communication devices, in- cluding smartphones, tablets, smart watches, and comput- ers, that use the 3G and LTE 1 standards. FID, at *2, *9. INVT’s infringement theory for the ’439 patent is based in part on alleging that the asserted claims of the ’439 patent are standard essential, i.e., subject matter essential to practicing the LTE standard. Id. at *58. In the LTE standard, the smallest portion of the com- munications spectrum is referred to as a subcarrier, like in the ’439 patent. However, a group of subcarriers (sub- bands, in the ’439 patent) is referred to as a “resource block.” See Appellant’s Br. 35; Intervenors’ Br. 12. A group of resource blocks (subband groups, in the ’439 patent) is referred to as “LTE subbands.” See Appellant’s Br. 35; In- tervenors’ Br. 12. At a high level, under the LTE standard, the user

1 LTE is an acronym for “Long-Term Evolution.” FID, at *22 n.14. Case: 20-1903 Document: 98 Page: 7 Filed: 08/31/2022

INVT SPE LLC v. ITC 7

device chooses one of 16 possible combination of modula- tion and coding parameters, per LTE subband, to send to the base station. See FID, at *67. The base station, in turn, selects a combination of modulation and coding parame- ters, but from a much greater number of possible combina- tions. See id. at *60–61 (noting almost 3,000 possible values for calculating the coding parameters). As dis- cussed later, there is no evidence in the record that at least one modulation and coding parameter combination that the base station can select matches a combination that the user device can select. See infra Part II.C.1 & C.2. After selecting its parameters, the base station modulates and encodes the data using those selected parameters and then sends the modulated and encoded data and its selected pa- rameters to the user device. FID, at *59–61. The user de- vice demodulates and decodes the data using the parameters selected by the base station. Id. The user device and the base station send and receive parameter information using values referred to as CQI, DCI, and TBS. See id. at *67–68 (citing J.A. 10865–66; JA 11858; J.A. 13169–70; J.A. 1175); Appellant’s Br. 36–37; Intervenors’ Br. 60. The initial selection of parameters by the user device is transmitted to the base station as a CQI 2 index. The CQI index corresponds to one of the 16 possible combination of modulation and coding parameters. These 16 possible combinations are the only combinations the user device can choose. FID, at *60. When the base station returns parameters to the user device, it does so in a DCI 3

2 CQI is an acronym for “channel quality indicator.” FID, at *58 n.56. The respondents’ expert explained that the CQI report to the base station indicates the “maximum rate at which the base station can send real data to the UE [user equipment].” Id. at *59. 3 DCI is an acronym for a “downlink control indica- tor.” FID, at *67. Case: 20-1903 Document: 98 Page: 8 Filed: 08/31/2022

message. Id. at *67; see Intervenor’s Br. 60; Appellant’s Br. 37 (citing J.A. 10867). The DCI message includes an as- signment of resource blocks, which tells the user device which portions of the bandwidth to use to receive data from the base station. FID, at *67. The DCI message also in- cludes an MCS 4 index—one MCS value for the entire re- source block assignment. Id. at *59, *68. There are 32 possible values for the MCS index, each one associated with a modulation scheme (Qm)5 and a TBS 6 index. Id. at *59. The TBS index corresponds to a TBS table, which in- cludes almost 3,000 entries. Id. at *60–61. 7 Based on the resource block assignment and the TBS value, the user de- vice calculates the coding parameter. Id. at *60. The FID depicts and discusses in detail the LTE standard’s CQI, MCS, and TBS tables. See id. at *59–61. In summary, based on the DCI message, MCS index, TBS index, and the resource block assignment, the user de- vice receives information about the modulation and coding

4 MCS is an acronym for “modulation and coding scheme.” FID, at *59 n.57, *68. 5 European Telecommunications Standards Insti- tute (ETSI), ETSI TS 136 211 V8.4.0 (2008–11)—LTE; Evolved Universal Terrestrial Radio Access (E-UTRA); Physical channels and modulation (3GPP TS 36.211 ver- sion 8.4.0 Release 8) 8 (2008), https://www.etsi.org/de- liver/etsi_ts/136200_136299/136211/08.04.00_60/ts_13621 1v080400p.pdf (“Qm Modulation order: 2 for QPSK, 4 for 16QAM and 6 for 64QAM transmissions”). 6 TBS is an acronym for “transport block size.” FID, at *59 n.58. 7 “[T]he reason for so many entries is ‘to give the base station a great deal of flexibility in terms of the downlink assignment, including the assignment of the code rate to be used for all the resources blocks.’” FID, at *60 (quoting J.A. 11882:15–20). Case: 20-1903 Document: 98 Page: 9 Filed: 08/31/2022

INVT SPE LLC v. ITC 9

parameters that were selected by the base station to mod- ulate and encode the data that the user device receives from the base station. Id. at *68. The user device demod- ulates and decodes the received data using the modulation and coding parameters selected by the base station. Id. Regardless of the number of LTE subbands included in the resource block assignment, one set of modulation and cod- ing parameters is used to modulate and encode and demod- ulate and decode the data. Id. (citing J.A. 10871–74; J.A. 11861, 11864–65, 11867). As will be discussed, the evi- dence does not show that the user device ever receives data modulated and encoded with the same parameters initially selected by the user device. See infra Part II.C.1 & C.2. D. ALJ’s Final Initial Determination (FID) Relevant to the disposition of this appeal, the ALJ’s FID found that the accused products did not infringe the asserted claims of the ’439 patent. FID, at *58–71. INVT asserted two infringement theories: (1) the ’439 claims are essential to the practice of the LTE standard, see id. at *58– 63, and (2) the accused products practice the asserted claims, see id. at *63–71. The ALJ found that independent claim 1 of the ’439 patent is not essential to the LTE stand- ard. Id. at *63. Therefore, INVT could not show infringe- ment by relying on the fact that the accused products were LTE-compliant. Id. In addition, the ALJ found that INVT had failed to prove infringement under a normal infringe- ment analysis, which analyzes the accused products in view of the asserted claims. Id. at *63, *71. In the proceeding below, the dispute over whether the asserted ’439 claims are essential to practicing the LTE standard was limited to limitations [d] and [e]. See J.A. 1713–14 (“Respondents do not dispute the essentiality or infringement of elements 1(a), 1(b), 1(c), 1(f), and 1(g). Tr. (Acampora) 1907:21–1908:10 (‘Q. Right. You limited your opinions to 1.d and 1.e in Claim 1, right? A. As far as my opinions on noninfringement [and essentiality] are Case: 20-1903 Document: 98 Page: 10 Filed: 08/31/2022

concerned, that’s correct.’).” (emphasis omitted)); FID, at *58–63 (analyzing whether limitations [d] and [e] are es- sential to the LTE standard). Prior to the evidentiary hearing, the ALJ addressed limitation [e], which recites a communication apparatus’s “data obtaining section that demodulates and decodes the received signal . . . using the . . . parameters decided at the parameter deciding section” of the communication appa- ratus. J.A. 252–54 (Order No. 52). Although the ALJ de- clined to provide an explicit construction, the ALJ explained that the claim limitation could be met even when the parameters used to demodulate and decode are decided by the communicating party (i.e., a base station), so long as the communicating party chooses “those very parameters” decided by the communication apparatus (i.e., a user de- vice) for communication between the two entities. J.A. 254 (emphasis added). The ALJ reasoned that the plain lan- guage dictated the result, noting that “claim 1 does not re- quire the [user device’s] ‘parameter deciding section’ to serve as the final or ultimate decision maker with respect to exchanges that occur in the communication system . . . As INVT asserted, ‘Claim 1 is silent on the operation of the communicating party.’” J.A. 254 n.3. With this interpretation, the ALJ held that INVT failed to show claim 1 was essential to the LTE standard. FID, at *58–63. Specifically, the ALJ found that INVT “failed to present evidence that the modulation and coding parame- ters corresponding to the CQI index the [user device] ini- tially reports to the base station are the ‘very parameters’ that the base station ultimately determines are appropri- ate” to meet limitations [d] and [e]. Id. at *58 (internal footnote omitted). Under the LTE standard, the CQI trans- mitted from an LTE user device includes only 16 possible modulation and coding parameter combinations, whereas there is a “much larger” number of possible modulation and coding parameter combinations from which the LTE base station can select and send to the user device as part of the Case: 20-1903 Document: 98 Page: 11 Filed: 08/31/2022

INVT SPE LLC v. ITC 11

DCI message. Id. at *60–61. The ALJ found that “even if the base station happens to consider the CQI reported by the [user device], it does not appear to be likely, must [sic] less required, for the base station to choose the ‘very pa- rameters’ initially decided by the [user device].” Id. at *61. The ALJ also rejected INVT’s argument that, despite the significant mismatch between the possible parameters the user device is able to select and the base station is able to select, the claim was standard essential because an LTE- compliant user device has the capability to receive data modulated and encoded using the parameters decided by the user device and to demodulate and decode that data. Id. at *61. The ALJ stated that “patent essentiality cannot, as a matter of law, be established merely by showing that the asserted standard is capable of meeting the claim, as mere capability of a claimed feature is ipso facto not tanta- mount to the requirement that the claimed feature must be mandatory.” Id. at *61. “Patent essentiality,” the ALJ ex- plained, requires the standard to “necessarily” meet the el- ements of the claim. Id. The ALJ also found that the language of claim 1 was not drawn to capability, as further support for the conclusion that “mere capability in this in- stance does not equate to infringement.” Id. at *61–62. The ALJ found claim 1 was not drawn to capability based on the fact that its language (“using,” “decided”) did not fol- low the “for performing”-type language (“for preventing,” “for obtaining”) of the claims in our Finjan decision. Id. (citing Finjan, Inc. v. Secure Computing Corp., 626 F.3d 1197, 1204–05 (Fed. Cir. 2010)). Because INVT could not rely on the accused products’ compliance with the LTE standard to establish infringe- ment, the ALJ stated that INVT was required to show that actual operation of the accused products meets every limi- tation of the asserted claims. Id. at *63. Under this anal- ysis, the respondents’ arguments were focused on limitations [d] and [e] not being met by the accused prod- ucts. Id. at *67; id. at *71–72 (noting parties’ agreement Case: 20-1903 Document: 98 Page: 12 Filed: 08/31/2022

and respondents’ lack of argument for limitations [a] through [c]); see also id. at *73–74 (noting INVT’s lack of evidence and argument separate from the standard-essen- tial ones for limitation [f]). The ALJ agreed with respond- ents that the accused products did not meet limitations [d] or [e] because they did not receive data modulated and en- coded “on a per subband group basis” nor demodulate and decode data “on a per subband group basis.” Id. at *67–69. Specifically, the per-subband-group limitation was not met because a single MCS (indicating a single modulation and coding scheme) was used and sent to the user device for an entire resource block assignment (spanning one or multiple LTE subbands, i.e., subband groups). Id. at *67–69. The ALJ also found that limitation [e] was not met based on INVT’s failure to show that the information in the DCI re- ceived and used by the user device from the base station is informed by the CQI that the user device previously trans- mitted to the base station. Id. at *69–70. The evidentiary failure was because of a failure to analyze source code gov- erning the operation of the base station (as opposed to the source code of the user device). Id. at *70. Relying on a similar analysis, the ALJ found that INVT failed to meet the technical prong of the domestic industry requirement. First, even if the representative product, a Samsung Galaxy S9, complies with the LTE standard, claim 1 is not standard essential. Id. at *73. Second, INVT failed to show that the S9’s actual operation meets the “on a per subband group basis” requirement of limitations [d] and [e]. Id. at *74. INVT appeals. We have jurisdiction under 28 U.S.C. § 1295(a)(6). DISCUSSION I The ’590 patent expired on March 5, 2022. The ITC has a limited statutory mandate and can only grant prospective Case: 20-1903 Document: 98 Page: 13 Filed: 08/31/2022

INVT SPE LLC v. ITC 13

relief. Tessera, Inc. v. Int’l Trade Comm’n, 646 F.3d 1357, 1371 (Fed. Cir. 2011); Tex. Instruments, Inc. v. Int’l Trade Comm’n, 851 F.2d 342, 344 (Fed. Cir. 1988) (citing 19 U.S.C. § 1337(d)–(f)). “The ITC can issue only an exclusion order barring future importation or a cease and desist order barring future conduct. If the violation of section 337 in- volves patent infringement, neither of the above remedies is applicable once the patent expires.” Tex. Instruments, 851 F.2d at 344. The expiration of the ’590 patent, there- fore, has rendered this appeal moot with respect to that pa- tent. See id. INVT argues that its appeal regarding the ’590 patent is not moot even after the patent’s expiration because of pending district-court litigation that was stayed in favor of the ITC investigation. Appellant’s Suppl. Br. 4–5, ECF No. 84. However, we have previously held that because ITC decisions on patent infringement or invalidity do not have preclusive effect on district court litigation, a decision by this court does not have enough “collateral consequences” to avert mootness, even though a pending district court case involves the same issues. Hyosung TNS v. Int’l Trade Comm’n, 926 F.3d 1353, 1358–59 (Fed. Cir. 2019) (citing Bio-Technology General Corp. v. Genentech, Inc., 80 F.3d 1553, 1563–64 (Fed. Cir. 1996); then citing Tex. Instru- ments Inc. v. Cypress Semiconductor Corp., 90 F.3d 1558, 1568–69 (Fed. Cir. 1996); and then citing Tandon Corp. v. U.S. Int’l Trade Comm’n, 831 F.2d 1017, 1019 (Fed. Cir. 1987)). INVT relies on Microsoft and Powertech. Appellant’s Suppl. Br. 5, ECF No. 84 (quoting Microsoft Corp. v. Int’l Trade Comm’n, No. 2012-1445, 2014 WL 10209132 (Fed. Cir. Jan 3, 2014) (per curiam); and then citing Powertech Tech. Inc. v. Tessera, Inc., 660 F.3d 1301 (Fed. Cir. 2011)). Neither helps INVT avoid mootness. Microsoft is a non- precedential opinion, where the patent expired after the court’s decision on the merits, and no party raised the issue of the patent’s imminent expiration or mootness until after Case: 20-1903 Document: 98 Page: 14 Filed: 08/31/2022

the petition for rehearing en banc was denied and two days before the mandate was set to issue. 2014 WL 10209132, at *2; see Hyosung, 926 F.3d at 1359 n.3; Microsoft Corp., No. 2012-1445, ECF. No. 104. Powertech did not involve mootness. It addressed the fact that a Federal Circuit de- cision on an appeal from the ITC can have precedential ef- fect on district courts as to certain other issues but also reaffirmed that ITC determinations of patent infringement and validity do not have preclusive effect on district courts, even when affirmed by the Federal Circuit. 660 F.3d at 1307–08 (explaining that district courts were bound by the legal precedent set forth in the prior decision Tessera, that a licensed sale does not become unauthorized and infring- ing merely because the licensee falls behind on royalty pay- ments). Here, like in Hyosung, we see no “potential for collateral consequences resulting from the possible stare decisis effect of our decision, if precedential” that prevents the appeal from becoming moot. 926 F.3d at 1359. Because the ’590 patent portion of this appeal is mooted due to the intervening happenstance of the patent’s expiration, we vacate the ITC’s decision as to that patent and remand with instructions to dismiss as moot the rele- vant portion of the complaint. See Tessera, 646 F.3d at 1371; U.S. Bancorp Mortg. Co. v. Bonner Mall P’ship, 513 U.S. 18, 25 & n.3 (1994); United States v. Munsingwear, Inc., 340 U.S. 36, 39–40 (1950); see also Oral Arg. 48:48– 49:18. II On appeal, INVT argues that the ALJ erred in finding that the accused products do not infringe claim 1 of the ’439 patent based on a combination of misinterpreting the claim, to require performance by (not merely capability of) an accused product, and misapplying law on standard es- sential patents. Appellant’s Br. 59–67. INVT also chal- lenges the ALJ’s “actual operation” noninfringement findings, specifically that the “per subband group basis” Case: 20-1903 Document: 98 Page: 15 Filed: 08/31/2022

INVT SPE LLC v. ITC 15

requirement of limitations [d] and [e] and the “pattern stor- age section” of limitation [f] are not met by the accused products. Appellant’s Br. 67–72. INVT’s arguments do not prevail. We agree with INVT that the asserted ’439 claims are drawn to “capability,” and not to actual operation as the ALJ found. However, INVT has failed to show that the accused LTE-compliant devices have the capability required by the claims. Therefore, whether under a theory of the claims being standard essen- tial or the claims being met by the accused devices, INVT has not proven infringement. A According to INVT, claim 1 of the ’439 patent requires only that the accused LTE-compliant devices are capable of receiving, from a base station, data modulated and encoded with the same parameters decided by the user device, and capable of demodulating and decoding that data using those parameters. See Appellant’s Br. 62–65. Our cases have held that sometimes a device only needs to be “capable of operating” according to a claimed limitation, for a finding of infringement. See Finjan, 626 F.3d at 1204. Other times, a device does not infringe unless it actually operates as claimed. See ParkerVision, Inc. v. Qualcomm Inc., 903 F.3d 1354, 1361 (Fed. Cir. 2018) (dis- cussing Ball Aerosol & Specialty Container, Inc. v. Ltd. Brands, Inc., 555 F.3d 984 (Fed. Cir. 2009)). Whether in- fringement requires actual performance of the recited func- tions by the accused device depends on the claim language. See Finjan, 626 F.3d 1204 (citing Fantasy Sports Props. v. Sportsline.com, Inc., 287 F.3d 1108, 1118 (Fed. Cir. 2002)); see also ParkerVision, Inc., 903 F.3d at 1361 (articulating a distinction between configuration-type and capability- type claims). Possibly the most straightforward example of this is the common distinction between method claims and Case: 20-1903 Document: 98 Page: 16 Filed: 08/31/2022

apparatus claims. See, e.g., Finjan, 626 F.3d at 1203–05. In Finjan, this court held that Finjan’s “non-method claims describe capabilities without requiring that any software components be ‘active’ or ‘enabled.’” Id. at 1204–05. The court, therefore, upheld a finding of infringement for prod- ucts in which the accused proactive-scanning software module was locked when sold by the defendants. Id. at 1205; see also id. at 1203–04 (distinguishing Southwest Software, Inc. v. Harlequin Inc., 226 F.3d 1280, 1291 (Fed. Cir. 2000), in which the accused software product required a manual step to activate the patented feature, but the claim at issue was a method claim). In contrast, Finjan’s method claims were not infringed by the accused products because those claims required actual performance of each claimed step; there was no evidence that proactive scan- ning was performed in the United States. Id. at 1206 (“To infringe a method claim, a person must have practiced all steps of the claimed method.” (quoting Lucent Techs. V. Gateway, Inc., 580 F.3d 1301, 1317 (Fed. Cir. 2009))). Cf. ParkerVision, Inc., 903 F.3d at 1363 (similarly differentiat- ing apparatus claims and method claims in the patent in- validity context). However, differences exist between apparatus claims as well, depending on the claim language. We have con- strued some apparatus claims to require an infringing de- vice to actually perform and operate according to the functional terms recited in the claim. The intervenors cite two such cases, Cross Medical and Ball Aerosol. See Inter- venors’ Br. 64, 67–68. We have construed other apparatus claims to require only capability, such as in Finjan and Sil- icon Graphics. Ball Aerosol and Cross Medical both involve mechani- cal apparatus claims. According to these two cases, inter- venors argue, the ’439 claims should be construed to require actual operation of the functions recited in limita- tions [d] and [e] in order for there to be infringement. In- tervenors’ Br. 64–69. In Cross Medical, we rejected the Case: 20-1903 Document: 98 Page: 17 Filed: 08/31/2022

INVT SPE LLC v. ITC 17

argument that the limitation “anchor seat means which has a lower bone interface operatively joined to said bone segment” could be met merely by an interface that was ca- pable of contacting bone. Cross Med. Prods., Inc. v. Med- tronic Sofamor Danek, Inc., 424 F.3d 1293, 1305–06 (Fed. Cir. 2005). Instead, the claim language “operatively joined” required that the interface and the bone segment be connected and in contact such that the device effectively performed posterior stabilization. Id. at 1306. Direct in- fringement did not occur until the device was connected to the bone, which a surgeon performed, not the allegedly in- fringing device maker. Id. at 1310–14. In Ball Aerosol, the apparatus claim recited a specific physical relationship between elements of a candle holder, in which protrusions in the bottom of a candle holder are resting upon the cover of the candle tin, the cover having been removed from the rest of the tin. Ball Aerosol, 555 F.3d at 994–95. The claim recited not only a particular physical relation between two components of the device, but included a limitation resembling a particular use of those components. Id. at 987–88 (“the cover, when re- moved, being placed upon the surface with the holder being set upon the cover for the cover to support the holder above the surface”). We ultimately construed the Ball Aerosol claim to be a configuration-type claim, which required showing that the candle holder was actually placed on its cover. Id. at 995 (finding no infringement although the ac- cused Travel Candle had a removable cover that the candle holder was capable of being placed on, because there was no evidence that the Travel Candle was in fact ever placed in such a configuration). The candle holder makers and sellers did not infringe the claim because no infringement occurred until the candle holder was positioned on top of the cover. In both of those cases, actual operation of arranging el- ements in a particular way was required because of the claim language. Neither case involved claim language Case: 20-1903 Document: 98 Page: 18 Filed: 08/31/2022

resembling the language in INVT’s claims. Intervenors emphasize that the Cross Medical claims included the pre- sent-tense verb “has,” similar to the present-tense verbs “demodulates” and “decodes” in the ’439 claims. But this court’s construction in Cross Medical was based on “opera- tively joined,” not “has.” Intervenors do not provide any analysis of Ball Aerosol. While the Ball Aerosol claim is atypical for its recitation, within a mechanical device claim, of how a component is to be used or operated—some- what akin to computer device claims, which commonly in- clude functional language—it ultimately has little relevance for construing INVT’s claims. Because of the nature of the technology, computer and software claims typically use functional language to define the invention. Functional language is used to define and delimit otherwise generic or interchangeable general pur- pose computer hardware, which can be programmed to per- form an unlimited array of functions. In other words, the recited operative steps a computer- or software-based de- vice undertakes is what defines what a computer-imple- mented invention is. We have frequently construed such functional language as not requiring actual performance of those operative steps for infringement purposes. Moreo- ver, we have not required claims to adhere to a specific grammatical form to find that the claim is drawn to capa- bility, contrary to the Commission’s and the intervenors’ contentions. In Finjan, accused products sold with locked software modules still infringed the device claims because the claims described capabilities and did not require software components be active or enabled. Finjan, 626 F.3d at 1204–05. The claims used “for performing”-type claim lan- guage—“a logical engine for preventing execution,” “a com- munications engine for obtaining a Downloadable,” and “a linking engine . . . for forming a sandbox package.” Id. at 1205. Based on the language, we held that the system claims recite “software components with specific purposes.” Case: 20-1903 Document: 98 Page: 19 Filed: 08/31/2022

INVT SPE LLC v. ITC 19

Id. Because defendants admitted that program code for the relevant function (proactive scanning) was literally present on all accused devices, the claim was infringed “in the same way that an automobile engine for propulsion exists in a car even when the car is turned off.” Id. at 1205. The pres- ence of that programming in the accused products was all that was necessary for establishing infringement. But “for performing”-type language like in Finjan is not the only way for a computer-implemented claim to be di- rected to capability, as intervenors contend. See, e.g., In- tervenors’ Br. 66–68. In Silicon Graphics, the computer apparatus claim recited “a rasterization circuit coupled to the processor that rasterizes the primitive according to a rasterization process which operates on a floating point for- mat.” Silicon Graphics, Inc. v. ATI Techs., Inc., 607 F.3d 784, 788 (Fed. Cir. 2010) (emphasis added and omitted). The accused product was a graphics processor that could not perform rasterization unless combined with an operat- ing system. Id. at 794. We held that infringement did not turn on whether the accused product was actually perform- ing the rasterization but, simply, whether it included a ras- terization circuit for doing so. Id. at 795. We held that the apparatus claim directed to a computer, claimed in func- tional terms, is nonetheless infringed so long as the product is “designed ‘in such a way as to enable a user of that [prod- uct] to utilize the function . . . without having to modify [the product]”—i.e., capable of the functions. Id. (original alter- ations) (quoting Fantasy Sports, 287 F.3d at 1118). We noted that “[w]here, as here, a product includes the struc- tural means for performing a claimed function, it can still infringe ‘separate and apart’ from the operating system that is needed to use the product.” Id. Both appellee and intervenors ignore the fact that Silicon Graphics treats “that rasterizes” as capability-type language. See Appel- lee’s Br. 53 (noting other parts of the claim which use “for performing”-type language); Intervenors’ Br. 67–68 (same, conflating the rasterization circuit and the processor, to Case: 20-1903 Document: 98 Page: 20 Filed: 08/31/2022

which the rasterization circuit is coupled). The FID relied primarily on Finjan and the absence of “for performing”-type claim language, providing little else for why the ’439 claims require actual operation. The in- tervenors argue that simple-present-tense words are enough to require actual operation, see Intervenors’ Br. 65, 67, but Silicon Graphics shows that is not true. 8 In fact, based on just claim language, we see very little significance in the difference between a limitation that might recite “a data obtaining section for demodulating and decoding” (Finjan-style) and one that recites “a data obtaining section that demodulates and decodes” (the actual ’439 claim lan- guage), for determining on which side of the capability/ac- tual-operation line the claims fall. 9

8 See also, e.g., MasterMine Software, Inc. v. Mi- crosoft Corp., 874 F.3d 1307, 1315–16 (Fed. Cir. 2017) (“Though claim 8 includes active verbs—presents, receives, and generates—these verbs represent permissible func- tional language used to describe capabilities of the ‘report- ing module.’”); UltimatePointer, L.L.C. v. Nintendo Co., 816 F.3d 816, 826–27 (Fed. Cir. 2016) (holding that “an image sensor, said image sensor generating data” reflected the ca- pability of the claimed device). 9 The other cases cited by the intervenors are not rel- evant or distinguishable. Acco Brands involved induced in- fringement, where a showing of underlying direct infringement was required based on an end user’s actual use of an infringing mode versus a noninfringing mode. Evidence of the infringing capability of the device was therefore insufficient. ACCO Brands, Inc. v. ABA Locks Mfrs. Co., 501 F.3d 1307, 1313–14 (Fed. Cir. 2007). In Te- lemac Cellular, this court held that a patentee could not rely on an alleged capability of the accused device to per- form a claimed function where the device needed to be mod- ified to infringe. Telemac Cellular Corp. v. Topp Telecom, Case: 20-1903 Document: 98 Page: 21 Filed: 08/31/2022

INVT SPE LLC v. ITC 21

Like with the claims in Finjan and Silicon Graphics, we find that the asserted ’439 claims are directed to capa- bility—as in a device that includes “software components with specific purposes,” programmed to have the ability to perform the operative steps: namely, a “receiving section” that can receive a signal containing data modulated and encoded using the parameters decided at user device, and a “data obtaining section” that can demodulate and decode the signal using the parameters decided at the user device. Finjan, 626 F.3d at 1204–05. In other words, the claims recite a device with the capability of performing the recited functions when in operation without any modification or further programming. B While we agree with INVT that its claims are “capabil- ity” claims, that does not mean we agree with its conception of what necessarily are the capabilities of an infringing de- vice. We thus find it necessary to further construe the as- serted ’439 patent claims to clarify that determining a user

Inc., 247 F.3d 1316, 1330 (Fed. Cir. 2001). Similarly, in High Tech Medical Instrumentation, the court held that in- fringement could not be based on an unintended alteration of the device. High Tech Med. Instrumentation, Inc. v. New Image Indus., Inc., 49 F.3d 1551, 1555 (Fed. Cir. 1995). In ViaTech, a nonprecedential opinion, the claims required a dynamic license database and it was undisputed that the accused product lacked a database. The fact that the prod- uct had the ability to eventually create or generate a data- base (with the installation of Windows on the device) could not support a finding of infringement. ViaTech Techs. Inc. v. Microsoft Corp., 733 F. App’x 542, 551–52 (Fed. Cir. 2018). These cases do not inform whether the ’439 claims are correctly construed to be directly infringed by the sale of a device that has the pre-existing capability, without modification, to perform the claimed functions. Case: 20-1903 Document: 98 Page: 22 Filed: 08/31/2022

device’s capability involves analyzing a base station’s oper- ation. INVT argues that the Commission erred by consid- ering the operation of base stations in determining whether the asserted ’439 claims were infringed. See Ap- pellant’s Br. 61 (“[T]he asserted claims are directed to de- vices, not base station . . . . Whether a base station is required to take certain actions is irrelevant to whether the accused user devices meet the claim limitations.”). The in- tervenors argue, on the other hand, that whether the ac- cused products are capable of meeting the claim limitations depends on how the base stations operate. See Intervenors’ Br. 71 (“[W]hether Respondents’ product do (or even can) demodulate/decode data using the parameter combinations from the CQI table depends on whether base stations them- selves even can modulate/encode data using those combina- tions.”). We agree with the intervenors that, although the asserted ’439 claims do not include the base station itself, the base station’s operation is a part of the infringement analysis. To determine whether an accused device is a device with the “capability” of performing the recited functions, it must be able to perform those functions when it is activated and put into operation. See infra Section C. Here, that means that the accused device receives and then decodes and demodulates a data signal with a particular claimed protocol—using the same parameters it had previously chosen. In this case, the user device’s capability is depend- ent on the base station’s capability. The base station is part of “the environment” in which the user device must function. Advanced Software Design Corp. v. Fiserv, Inc., 641 F.3d 1374 (Fed. Cir. 2011). The claims have specific requirements for the data signal that the user device’s receiving section and data obtaining sec- tion handle and process when the device is activated and put into operation. That received data signal must be mod- ulated and encoded with specific parameters—and not by the claimed user device but by a separate base station. See Case: 20-1903 Document: 98 Page: 23 Filed: 08/31/2022

INVT SPE LLC v. ITC 23

’439 patent, claim 1, limitation [d] (“a signal containing data modulated and encoded on a per subband group basis at the [base station] using the modulation parameters and the coding parameters of the parameter information trans- mitted at the parameter information transmission sec- tion”). To understand whether a user device can ever receive a data signal with the particularized characteris- tics set forth in the claim, it is necessary to know whether the base station (i.e., the communicating party) is capable of transmitting that particular type of data signal to the user device. Therefore, although the recited base station is not “a limitation on the claimed invention itself,” Nazomi Commcn’s, Inc. v. Nokia Corp., 739 F.3d 1339, 1345 (Fed. Cir. 2014), in the sense that an infringer would not need to, for instance, use, make, or sell the base station, the base station’s operation affects whether the claims are met, see, e.g., Advanced Software Design, 641 F.3d at 1373–74. In Advanced Software, the claimed invention was for validating a check, to prevent check fraud, and involved ei- ther decrypting or encrypting information on the check. The preamble of the claim set out that the check included “selection information [that] is encrypted” to generate a control code and a “control code [which] is printed on the [check].” We held that these steps in the preamble “de- fine[d] the financial instrument that the claimed system validates” as opposed to setting forth steps that would have to be performed by the accused infringer. Advanced Soft- ware, 641 F.3d at 1373–74 & 1374 n.1. Nevertheless, the accused infringer would infringe “only by validating checks that [had] been encrypted and printed in accordance with steps described in the preamble.” Id. at 1374. Like in Advanced Software, the claimed device of the ’439 patent operates in an environment that involves ac- tions of another device (the communicating party, i.e., the base station). The claimed device’s capability of perform- ing the recited functions depends on being supplied a cer- tain modulated and encoded signal, which, in turn, Case: 20-1903 Document: 98 Page: 24 Filed: 08/31/2022

requires the supplier (the communicating party) to actually supply that signal. Because the communicating party (base station) generates the necessary environment, its op- erations must be known to determine whether the accused device infringes, i.e., is capable of performing the claimed functions. C We find, on this record, that INVT has not shown in- fringement, even though the claims are drawn to capabil- ity. “[W]here claim language recites ‘capability, as opposed to actual operation,’ an apparatus that is ‘reasonably capa- ble’ of performing the claimed functions ‘without signifi- cant alterations’ can infringe those claims.” ParkerVision, Inc., 903 F.3d at 1362. In contexts involving software func- tionality, we have never suggested that reasonable capabil- ity can be established without any evidence or undisputed knowledge of an instance that the accused product per- forms the claimed function when placed in operation. For example, in Ericsson, a capability claim limitation was met because of a finding of the accused device’s capability to perform the function “some of the time” based on “proof that it was in fact so used by some device users.” Ericsson, Inc. v. D-Link Sys., Inc., 773 F.3d 1201, 1216 (Fed. Cir. 2014). Similar support underlies infringement in Versata, Finjan, ParkerVision, and Fantasy Sports Properties, as additional examples. See Versata Software, Inc. v. SAP Am., Inc., 717 F.3d 1255, 1263 (Fed. Cir. 2013) (noting that customers did and were expected to perform the function- ality); Finjan, 626 F.3d at 1203 (“Defendants argue that in- fringement occurred only when customers purchased keys and unlocked proactive scanning modules . . . .” (emphasis added)); ParkerVision, 903 F.3d at 1360 (noting, in the in- validity context that “it is undisputed that [the reference] Nozawa’s device necessarily will produce a periodic signal that contains integer multiples of the fundamental Case: 20-1903 Document: 98 Page: 25 Filed: 08/31/2022

INVT SPE LLC v. ITC 25

frequency under some, albeit not all, conditions” (emphases added)); Fantasy Sports Props., Inc. v. Sportsline.com, Inc., 287 F.3d 1108, 1119 (treating the claim as drawn to soft- ware with the “ability” to award a particular type of bonus points and finding a need for further factfinding regarding evidence of such bonus point awarding being performed without modification). 10 Relatedly, we have rejected find- ing infringement based on an accused product being “merely capable of being modified in a manner that in- fringes.” Fantasy Sports, 287 F.3d at 1117–18 (emphases added) (citing High Tech Med. Instrumentation, Inc. v. New Image Indus., Inc., 49 F.3d 1551, 1555–56 (Fed. Cir. 1995); and then citing Telemac Cellular Corp. v. Topp Telecom, Inc., 247 F.3d 1316, 1330 (Fed. Cir. 2001); and discussing Intel Corp. v. ITC, 946 F.2d 821 (Fed. Cir. 1991)). While we have noted the principle that “apparatus claims cover what a device is, not what a device does,” Hewlett-Packard Co. v. Bausch & Lomb Inc., 909 F.2d 1464, 1468 (Fed. Cir. 1990), as a useful reminder that the focus of apparatus claims is the structure and not the op- eration or use, with computer- or software-based

10 The appellant cites Silicon Graphics for the propo- sition that direct infringement does not require the perfor- mance of all elements in the apparatus claims. Appellant’s Br. 63. In Silicon Graphics, we noted that, in addition to actual use of a product, infringement of an apparatus claim occurs when the invention is made or sold in the United States. That is what we were referring to when we said “even absent its use (or performance),” an apparatus claim directed to a computer with functional terms is nonetheless infringed. Silicon Graphics, Inc. v. ATI Techs., Inc., 607 F.3d 784, 794 (Fed. Cir. 2010). Summary judgment was improper because there was a possible factual dispute as to whether the accused products, once run, could rasterize and store the data as functionally claimed. See id. at 795. Case: 20-1903 Document: 98 Page: 26 Filed: 08/31/2022

inventions, apparatus claims routinely depend on func- tional claiming to describe the apparatus. It is the func- tional terms that distinguish a general purpose computer, “which can be programmed to perform very different tasks in very different ways” from a special purpose computer that is programmed to perform the particular function. Aristocrat Techs. Australia Pty Ltd. v. Int’l Game Tech., 521 F.3d 1328, 1333 (Fed. Cir. 2008). Therefore, what the de- vice does (and how it does it) is highly relevant to under- standing what the device is, in the computer/software context. Because we require claim limitations to have some teeth and meaning, proof of reasonable capability of per- forming claimed functions requires, at least as a general matter, proof that an accused product—when put into op- eration—in fact executes all of the claimed functions at least some of the time or at least once in the claim-required environment. For the ’439 claims, the receiving section in limitation [d] and the data obtaining section in limitation [e] require receiving and handling a data signal from the base station that is modulated and encoded using parameters that were decided by the user device. INVT has failed to show that under the LTE standard, a user device ever receives and handles such a data signal, i.e., infringement based on the claim being essential to the standard, as discussed next in Subsection 1. Nor has INVT shown that the accused prod- ucts receive such a data signal, i.e., infringement based on comparing the claims to the actual accused products, as discussed subsequently in Subsection 2. Those failures mean that infringement has not been shown. 1 Infringement can be proven based on an accused prod- uct’s use of an industry standard if the asserted claim is standard essential. See Fujitsu Ltd. v. Netgear Inc., 620 F.3d 1321, 1326–29 (Fed. Cir. 2010); Godo Kaisha IP Bridge 1 v. TCL Commcn’s Tech. Holdings Ltd., 967 F.3d Case: 20-1903 Document: 98 Page: 27 Filed: 08/31/2022

INVT SPE LLC v. ITC 27

1380, 1383–84 (Fed. Cir. 2020). Claims are standard essential if “the reach of the claims includes any device that practices the standard.” Fujitsu, 620 F.3d at 1327. In other words, “all implemen- tations of a standard infringe the claim” and the “patent covers every possible implementation of a standard.” Id. 1327–28. In Fujitsu, this court explained that while “claims should be compared to the accused product to de- termine infringement,” “if an accused product operates in accordance with a standard, then comparing the claims to that standard is the same as comparing the claims to the accused product.” Id. at 1327. Therefore, once a claim is found to be standard essential, that is sufficient to find in- fringement for any standard-compliant device. Id. at 1328. Sometimes, “an industry standard does not provide the level of specificity required to establish that practicing that standard would always result in infringement. Or . . . the relevant section of the standard is optional, and standards compliance alone would not establish” infringement. Id. at 1327–28. In such instances, the patent owner cannot es- tablish infringement simply by arguing that the product practices the standard but “must compare the claims to the accused products or, if appropriate, prove that the accused products implement any relevant optional sections of the standard.” Id. at 1328. According to INVT, the ’439 claims are standard essen- tial because, under the LTE standard, all LTE-compliant devices must be capable of receiving, demodulating, and decoding data using any of the available modulation and coding parameters in LTE, including parameters originally decided by the LTE-compliant user device. See Appellant’s Br. 62–63, 66–67. But because INVT failed to provide any evidence showing that a base station in fact ever sends the user device a data signal that is modulated and encoded using parameters that the user device decided, INVT has failed to prove the required capability. Case: 20-1903 Document: 98 Page: 28 Filed: 08/31/2022

INVT only provides two record citations for the conten- tion that an LTE-compliant user device is necessarily ca- pable of demodulating and decoding data using parameters it originally sends to the base station. The first citation is to the testimony of its expert, Dr. Vojcic. 11 See Appellant’s Br. 63 (citing J.A. 10827 (827:8–13)); see also id. at 60, 66. This testimony merely states that an LTE user device is capable of receiving, and then demodulating and decoding using, the parameters it sent to the base station because those parameters are parameters found in the LTE stand- ard. Without any evidence as to whether the base station operating under the LTE standard ever selects the same parameters chosen by the user device, there is no evidence that a standard-compliant user device ever receives data

11 Q. Okay. So do [user devices] and LTE have to be able to handle the case where the scheduler does assign re- sources based on measurements reported by the [user de- vice]? A. [User device] would have to, absolutely, be able to receive -- to receive, using modulation and coding parame- ters that were decided by the user device, but in some cases it might also -- it should be able to receive other modulation and coding parameters. Q. Do [user devices] under LTE basically have to be able to receive and obtain data under any of the available modulation and coding parameters of LTE? A. They should -- yes, they must be capable of receiv- ing any. Q. And are all the accused products, based on your analysis, capable of receiving demodulating and decoding using the parameters decided by the [user device]? A. Yes, Counsel. J.A. 10826–27 (826:23–827:18). Case: 20-1903 Document: 98 Page: 29 Filed: 08/31/2022

INVT SPE LLC v. ITC 29

modulated and encoded with the claimed parameters. 12 The second citation is to the testimony of respondents’ expert, Dr. Acampora. See Appellant’s Br. 63 (citing J.A. 11918–19 (1918:1–1919:14)); see also FID, at *60, *66–67. Contrary to INVT’s characterization, this evidence does not show Dr. Acampora agreeing that the LTE standard con- templates that all LTE-compliant devices must be able to demodulate and decode using any of the available modula- tion and coding parameters of LTE, including the parame- ters originally decided by the device (i.e., the CQI parameters). Preceding the portion of his testimony cited by INVT, Dr. Acampora testified that an LTE user device does not have to be able to demodulate and decode using the param- eters shown in the CQI table, the table of the 16 combina- tions of parameters the user device can choose and transmit to the base station. J.A. 11915 (1915:2–6). It is true Dr. Acampora agreed that an LTE device must be able

12 Testimony by Dr. Vojcic additionally cited by INVT’s counsel during the hearing is unhelpful. Oral Arg. 13:49–14:46 (citing J.A. 10881; 10878; 10820); 15:25–16:10. Dr. Vojcic vaguely testified that for the base station, it “would be typical or it would try whenever it could to ac- complish that [i.e., sending the very CQI values back to the mobile] because that would optimize system capacity . . . .” J.A. 10881–82 (881:13–882:3); see also J.A. 10878 (878:5– 9) (“the base station will tend to use this reported CQI be- cause that would maximize its capacity and throughput and so on”); J.A. 10820 (820:11–22) (identifying the CQI ta- ble). And, ultimately, Dr. Vojcic rested on the idea that “who makes the ultimate decision is not really relevant for the infringement of the claim, because the [user device] must be capable of receiving what it decided when base sta- tion decides the same, and that’s -- that’s my explanation.” J.A. 10881–82 (881:13–882:3). Case: 20-1903 Document: 98 Page: 30 Filed: 08/31/2022

to demodulate QPSK, 16QAM, and 64QAM—in other words, the user device must be able to handle the modula- tion parameters listed in the CQI table. J.A. 11915 (1915:7–19). 13 But, Dr. Acampora disagreed that the user device must be able to handle the coding parameters shown in the CQI table. J.A. 11915–17 (1915:20–1917:23). Dr. Acampora supported his conclusion with a detailed ex- planation of how the coding parameter chosen by the base station is calculated based on several things, including the MCS parameter, the TBS table, and the resource block al- location (both the MCS parameter and the resource block allocation decided by the base station and then sent to the user device), J.A. 11916 (1916:2–14), and, importantly, that he is unaware of such calculation ever resulting in a coding parameter that matches a coding parameter in the CQI table, J.A. 11919 (1919:12–13); see also J.A. 11877–85. In response, INVT’s counsel posed a hypothetical, which Dr. Acampora ultimately agreed with. This hypo- thetical included as an assumption that there is a MCS index that the base station can choose and send to the user device that could result in a coding parameter that is listed in the CQI table. 14 Not fighting the hypothetical, Dr.

13 QPSK, 16AM, and 64AM are listed in the MCS ta- ble, the table for the parameters the base station sends to the user device. See supra footnote 5 (referring to ETSI documentation on parameter Qm in the LTE standard); J.A. 13157–58 (MCS table). 14 Q. I heard you the first time. My question was still a little different. I’m asking you if an LTE-compliant device receives data modulated QPSK with a coding rate of 78, will it be able to demodulate and decode that data? A. That’s different than the question you asked earlier. So if the [user device] concludes that the base sta- tion -- and I’ll put a little finer point on your question -- if Case: 20-1903 Document: 98 Page: 31 Filed: 08/31/2022

INVT SPE LLC v. ITC 31

Acampora agreed that if there is an MCS value, a corre- sponding TBS value, and resource block allocation that could ultimately result in a coding parameter that matches one listed in the CQI table (based on the appropriate calcu- lations), the user device would be required to be able to de- code using a CQI parameter. See J.A. 11917–19 (1917:24– 1919:14). In other words: If the LTE standard provides for the user device to receive the same parameters the user de- vice had decided, the user device that complies with the

the [user device] does the calculation I described, concludes that the code rate selected by the base station is 78 divided by 102.4, if that’s one of the possible outcomes of the com- putation, then the [user device] must be able to demodulate using the indicated modulation index and that code rate. Q. Right. You’re still talking about the base station a lot, but I assume the answer is true for any of these coding rates. If this modulation in this table and the associated coding rate is what the base station says to use, all [user devices], all LTE-compliant [user devices] and all the ac- cused products have to be able to handle that and demodu- late and decode using those parameters, right? A. If the base station had chosen one of those, based upon how much data it has to send, transfer block size, re- source block assignment so forth and so on, if the base sta- tion determines the TBS and the modulation rate, then if that combination -- which also includes how many resource blocks are being assigned -- happens to correspond with, say, to row 3, QPSK and 1093 divided by 102.4, if that hap- pens, then the [user device] must be capable of demodulat- ing using the corresponding modulation index and the code rate, yes. If that should ever happen. I don’t know if it ever does happen. Q. I’ll take that as a yes, Dr. Acampora. . . . J.A. 11917–19 (1917:24–1919:14) (emphases added). Case: 20-1903 Document: 98 Page: 32 Filed: 08/31/2022

LTE standard must be capable of receiving the parameters it decided. Like the testimony of Dr. Vojcic, Dr. Acampora’s testimony does not show that the base station, operating under the LTE standard, ever selects the same parameters chosen by the user device, and that a standard-compliant user device ever receives data modulated and encoded with the claimed parameters. The ALJ’s FID findings confirm there is no factual sup- port for INVT’s position that LTE user devices necessarily have the ability to receive the claimed data signal modu- lated and encoded with the parameters originally decided by the user device. See FID, at *58–61 (citing J.A. 11881– 84; J.A. 13157–63) (analyzing infringement based on inter- preting the claims to require actual operation). Although the ALJ did not decide the question of capability of an LTE user device—because the ALJ construed the claims to re- quire actual operation—the ALJ’s findings are nonetheless consistent with Dr. Acampora’s point. INVT has not shown that the LTE standard includes the possibility for a combi- nation of MCS, TBS, and resource block allocation values from a base station that results in a user device using a CQI coding parameter. See id. For that reason, INVT has failed to show that an LTE standard-compliant user device is capable of meeting the claimed functional language when that device is put into operation under the standard, and that the claims are essential to the LTE standard. 15

15 We need not address a potential additional issue that may arise in the context of standards and patented computer-implemented inventions—must accused devices operating under a standard perform the claimed functions all of the time or just some of the time to support an in- fringement finding, where the claim is drawn to capability. See FID, at *61 (ALJ stating that “mere capability of a claimed feature is ipso facto not tantamount to the require- ment that the claimed feature must be mandatory”). In Case: 20-1903 Document: 98 Page: 33 Filed: 08/31/2022

INVT SPE LLC v. ITC 33

2 Because the ALJ found the claims were not essential to the LTE standard (which we agree with), the ALJ stated that INVT must prove that the “actual operation” of the accused products meets every claim limitation. We con- strue the claims to be directed to capability and, thus, to be more precise, INVT was required to prove infringement in the ordinary manner, which involves “compar[ing] the claims to the accused products.” Fujitsu, 620 F.3d at 1328. The same rationale detailed above regarding the require- ment for reasonable capability again applies. An accused device cannot meet the required capabilities of the ’439 claims without evidence or agreement that the accused de- vice performs, without modification, the recited functions at least once when it is in operation. We see no reversible error in the ALJ’s finding that INVT failed to prove that the accused products actually practice limitation [e], Ap- pellee’s Br. 14, 16, for reasons discussed by the ALJ’s FID, FID, at *69–70. The ALJ credited the intervenors’ expert’s testimony that without the benefit of reviewing source code governing the operation of the base station’s communication with an accused device, it was not possible to determine whether the DCI transmitted to the user device was “informed in any way by a CQI value that was sent earlier” from the accused user device. Id. at *70. It was, therefore, not pos- sible to know whether the base station decides to use the “very parameters” included in the CQI without reviewing base station source code. Id. The ALJ relied on this signif- icant evidentiary gap about whether and how the base sta- tion was influenced by the CQI code to find that INVT failed to prove by a preponderance of the evidence that the

this case, INVT failed to prove that the accused products ever perform the requirements of claim limitations [d] and [e]. Case: 20-1903 Document: 98 Page: 34 Filed: 08/31/2022

accused products practice limitation [e]. This evidentiary gap mirrors and reinforces our earlier observation that there is nothing to support INVT’s contention that the ac- cused products are reasonably capable of receiving a data signal from the base station that is modulated and encoded using the parameters that are originally decided by the user device, and demodulating and decoding using the same parameters. Accordingly, we affirm the ALJ’s finding of noninfringement as supported by substantial evidence. D Because limitation [e] is not met due to INVT’s afore- mentioned failure of evidence, we need not address INVT’s other infringement arguments regarding “per subband group basis” in limitations [d] and [e] and the “pattern stor- age section” in limitation [f]. In addition, because the parties agree that the domes- tic industry findings fall with the noninfringement find- ings, Appellant’s Reply Br. 35, we affirm the Commission on finding no domestic industry. Lastly, we note that Chenery does not preclude our af- firmance of the Commission’s decision although we do so on a different claim construction (drawn to capability) than that underlying the Commission’s decision. See SEC v. Chenery Corp., 332 U.S. 194, 204 (1947). We make no fac- tual findings or decisions on any element of discretion that is the factfinding agency’s to make. Relying on Dr. Acampora’s testimony quoted earlier, see supra Section C.1, the ALJ found that “it does not appear to be likely, must [sic] less required, for the base station to choose the ‘very parameters’ initially decided by the UE.” FID, at *61. The ALJ erred in construing claim limitations as requiring more than capability for infringement, a legal question. The ALJ’s factual findings for noninfringement, however, are equally applicable under the correct claim construction because, as a matter of law, reasonable capability cannot be proven in light of the total absence of evidence put on by Case: 20-1903 Document: 98 Page: 35 Filed: 08/31/2022

INVT SPE LLC v. ITC 35

INVT. Therefore, our affirmance is not “on a basis contain- ing any element of discretion—including discretion to find facts and interpret statutory ambiguities—that is not the basis the agency used,” that would improperly “remove the discretionary judgment from the agency to the court.” Koyo Seiko Co. v. United States, 95 F.3d 1094, 1101 (Fed. Cir. 1996) (quoting ICC v. Brotherhood of Locomotive Engi- neers, 482 U.S. 270, 283 (1987)); see also Mayfield v. Ni- cholson, 444 F.3d 1328, 1336 (Fed. Cir. 2006) (noting the possibility of a court “conclud[ing] that there was no viola- tion of the Chenery doctrine on the ground that ‘it is clear . . . the agency would have reached the same ultimate re- sult under the court’s legal theory.” (quoting Grabis v. OPM, 424 F.3d 1265, 1270 (Fed. Cir. 2005); and citing Koyo Seiko Co., 95 F.3d at 1100–01))). We have not “scour[ed] the record to find some alternative basis to reach the same result.” Borovsky v. Holder, 612 F.3d 917, 921 (7th Cir. 2010). CONCLUSION For the foregoing reasons, we affirm the Commission’s determination of no violation of section 337 and a lack of domestic industry as to the ’439 patent because INVT has not shown that the relevant LTE devices infringe. With respect to the expired ’590 patent, we vacate the Commis- sion’s determination and remand with instructions to dis- miss the relevant portion of the complaint. AFFIRMED IN PART, VACATED IN PART WITH REMAND COSTS Costs against Appellant.