Regents of the University of Minnesota v. LSI Corporation

[2] 3 UNITED STATES DISTRICT COURT 4 NORTHERN DISTRICT OF CALIFORNIA 5 SAN JOSE DIVISION [6] 7 REGENTS OF THE UNIVERSITY OF Case No. 5:18-cv-00821-EJD MINNESOTA, 8 CLAIM CONSTRUCTION ORDER Plaintiff, [9] v. [10] LSI CORPORATION, et al., [11] Defendants. [12] 13 Plaintiff Regents of the University of Minnesota (“UMN”) brings this suit against 14 Defendants LSI Corporation and Avago Technologies U.S. Inc. (collectively, “LSI”) for 15 infringement of U.S. Patent No. 5,859,601, entitled “Method and Apparatus for Implementing 16 Maximum Transition Run Codes” (the “ ’601 Patent”). The parties dispute the proper 17 construction of three terms. See Am. Joint Claim Construction Statement (“CC Statement”), ECF 18 No. 240. After careful consideration of the record and the parties’ arguments at hearing, the Court 19 construes the contested language of the patent-in-suit as set forth below. 20 I. BACKGROUND 21 A. Overview of the Technology 22 UMN is a public research university that occasionally patents and commercializes the 23 inventions of its researchers. First Am. Compl. (“FAC”) ¶¶ 4–6, 8, ECF No. 40. As relevant in 24 this matter, UMN is the owner and assignee of the ’601 Patent, for which Dr. Jaekyun Moon (a 25 former UMN professor) and Dr. Barrett J. Brickner (Dr. Moon’s former graduate student) are 26 named inventors. Id. ¶¶ 48, 53, 73 . 27 The ’601 Patent generally pertains to digital storage systems. CC Statement, Ex. A (’601 1 Patent) at col. 1:10–11, ECF No. 240-1. Such systems work by recording data as 1’s and 0’s, 2 though the formats with which they record that data can vary. The ’601 Patent discusses two such 3 recording formats. First, a storage device can use the Non-Return-to-Zero (“NRZ”) recording 4 format. ’601 Patent at col. 1:24–27; see also CC Statement, Ex. D (“McLaughlin Decl.”) ¶ 13. In 5 NRZ recording, a 1 is represented by positive magnetization within the storage device, and a 0 is 6 represented by negative magnetization within the storage device.1 ’601 Patent at col. 1:24–27. 7 Second, a storage device can use the Non-Return-to-Zero-Inversion (“NRZI”) recording format. 8 ’601 Patent at col. 1:29–32; see also McLaughlin Decl. ¶ 14. In this format, a 1 is represented by 9 a magnetic transition (i.e., from positive to negative), and a 0 is represented by a non-transition. 10 ’601 Patent at col. 1:29–32. 11 When a device attempts to read data on a digital storage system, noise that occurs during 12 the process can cause errors which, in turn, cause the device to misread the data. McLaughlin 13 Decl. ¶¶ 16, 25. The ’601 Patent addresses this issue by introducing Maximum Transition Run 14 (“MTR”) codes, which eliminate certain patterns of data that are particularly error prone. ’601 15 Patent at col. 2:43–47. In particular, MTR codes “impose[] a limit on the maximum number of 16 consecutive transitions that can occur,” id. at col. 2:59–60, through the use of ‘j’ and ‘k’ 17 constraints, often written as (j;k). Id. at col. 4:46–48. The j constraint places a limit on the 18 maximum number of consecutive transition allowable, and the k constraint places a limit on the 19 maximum number of non-transitions allowable. Id. at col. 10:55–59; see also McLaughlin Decl. 20 ¶ 29. 21 To achieve these constraints, the ’601 Patent teaches the encoding of “datawords” (pieces 22 of input data) into “codewords” (corresponding pieces of output data). ’601 Patent at col. 10:50– 23 51. The codewords are chosen such that, when the codewords are joined together in a string, the 24 (j;k) constraints are satisfied. This encoding can be performed by using either “block” codes or [25] 26 1 The parties dispute whether the ’601 Patent is limited to magnetic storage media, or if it encompasses other forms of digital storage. The Court uses magnetic storage as an exemplar in 27 this Overview because the ’601 Patent does the same, but the Court does not express any opinion on whether NRZ or any other recording formats are limited to magnetic storage. 1 “state-dependent” codes. Block codes are those where each dataword is mapped to a unique 2 codeword. Id. at col. 5:66–67. By contrast, state-dependent codes assign codewords to datawords 3 based on the previously used codeword, or “state,” of the system. See id. at col. 5:50–53, 61–66. 4 B. Procedural History 5 This case was originally filed in the District of Minnesota on August 25, 2016. ECF No. 1. 6 After it was transferred to this district, the Court stayed the matter pending inter partes review 7 (“IPR”) on May 11, 2018. ECF No. 211. The parties vigorously litigated the validity of certain 8 claims in the ’601 Patent in IPR proceedings, including two trips to the Federal Circuit. 9 Ultimately, the Patent Trial and Appeal Board (“PTAB”) held that claim 13 was unpatentable but 10 did not invalidate claims 14 and 17, which depend from claim 13. CC Statement, Ex. B (“PTAB 11 Decision”), ECF No. 240-2. LSI appealed that decision to the Federal Circuit, which affirmed. 12 LSI Corp. v. Regents of Univ. of Minn., 43 F.4th 1349 (2022). 13 On October 7, 2022, the Court lifted its IPR stay upon stipulation of the parties. ECF No. 14 218. Throughout May and June of 2023, the parties briefed claim construction. See Pl.’s Opening 15 Claim Construction Br. (“UMN Br.”), ECF No. 239; Defs.’ Responsive Claim Construction Br. 16 (“LSI Resp.”), ECF No. 247; Pl.’s Claim Construction Reply Br. (“UMN Reply”), ECF No. 249. 17 The Court held a tutorial and Markman hearing on July 19, 2023. ECF No. 256. 18 C. Claim Language 19 UMN asserts infringement of only claims 14 and 17 of the ’601 Patent. However, because 20 both claims depend from claim 13, the disputed terms for construction come only from claim 13, 21 which the Court reproduces below: 22 A method for encoding m-bit binary datawords into n-bit binary codewords in a recorded waveform, where m and n are preselected 23 positive integers such that n is greater than m, comprising the steps of: 24 receiving binary datawords; and producing sequences of n-bit codewords; 25 imposing a pair of constraints (j;k) on the encoded waveform; generating no more than j consecutive transitions of said sequence 26 in the recorded waveform such that j ≧ 2; and generating no more than k consecutive sample periods of said 27 sequences without a transition in the recorded waveform. 1 ’601 Patent at col. 10:47–59. 2 II. LEGAL STANDARD 3 Claim construction is a question of law to be decided by the court. Markman v. Westview 4 Instruments, Inc., 52 F.3d 967, 979 (Fed. Cir. 1995) (en banc), aff’d, 517 U.S. 370 (1996). “[T]he 5 interpretation to be given a term can only be determined and confirmed with a full understanding 6 of what the inventors actually invented and intended to envelop with the claim.” Phillips v. AWH 7 Corp., 415 F.3d 1303, 1316 (Fed. Cir. 2005) (citation omitted). Consequently, courts construe 8 claims in the manner that “most naturally aligns with the patent's description of the invention.” Id. 9 (citation omitted). 10 When construing disputed terms, courts begin with “the language of the asserted claim 11 itself.” Comark Commc’ns, Inc. v. Harris Corp., 156 F.3d 1182, 1186 (Fed. Cir. 1998) (citations 12 omitted). That is because “[i]t is a ‘bedrock principle’ of patent law that ‘the claims of a patent 13 define the invention to which the patentee is entitled the right to exclude.’” Phillips, 415 F.3d at 14 1312 (citation omitted). The words of a claim should be given their “ordinary and customary 15 meaning,” which is “the meaning that the term would have to a person of ordinary skill in the art 16 in question at the time of the invention.” Id. at 1312-13 (citations omitted). However, the person 17 of ordinary skill in the art does not work from a blank slate—rather, she “is deemed to read the 18 claim term not only in the context of the particular claim in which the disputed term appears, but 19 in the context of the entire patent, including the specification.” Id. at 1313 . Thus, courts “have 20 long emphasized the importance of the specification in claim construction.” David Netzer 21 Consulting Eng’r LLC v. Shell Oil Co., 824 F.3d 989, 993 (Fed. Cir. 2016) (citation omitted). 22 They have explained that the specification is “always highly relevant” and “[u]sually, it is 23 dispositive; it is the single best guide to the meaning of a disputed term.” Phillips, 415 F.3d at 24 1315 (quoting Vitronics Corp. v. Conceptronic, Inc., 90 F.3d 1576, 1582 (Fed. Cir. 1996)). 25 The prosecution history of a patent—which “consists of the complete record of the 26 proceedings before the PTO [Patent and Trademark Office]”—is also intrinsic evidence of a claim 27 term’s meaning. Id. at 1317. But since the prosecution history “represents an ongoing negotiation 1 between the PTO and the applicant, rather than the final product of that negotiation, it often lacks 2 the clarity of the specification and thus is less useful for claim construction purposes.” Id. 3 Nonetheless, there are times when prosecution history is controlling. Where a patentee disavows 4 certain meanings during prosecution, the doctrine of prosecution disclaimer “preclude[es] [her] 5 from recapturing through claim interpretation [those] meanings.” Omega Eng’g, Inc. v. Raytek 6 Corp., 334 F.3d 1314 , 1323 (Fed. Cir. 2003). To qualify as a disclaimer, such disavowal “must be 7 both clear and unmistakable.” 3M Innovative Props. Co. v. Tredegar Corp., 725 F.3d 1315 , 1325 8 (Fed. Cir. 2013). That is a high bar: the disavowal “must be ‘so clear as to show reasonable clarity 9 and deliberateness,’ and ‘so unmistakable as to be unambiguous evidence of disclaimer.’” 10 Genuine Enabling Tech. LLC v. Nintendo Co., 29 F.4th 1365, 1374 (Fed. Cir. 2022) (quoting 11 Omega, 334 F.3d at 1325). 12 Finally, the court is also authorized to consider extrinsic evidence, such as “expert and 13 inventor testimony, dictionaries, and learned treatises.” Markman, 52 F.3d at 980 (internal 14 citations omitted). Although the court may consider evidence extrinsic to the patent and 15 prosecution history, such evidence is considered “less significant than the intrinsic record” and 16 “less reliable than the patent and its prosecution history in determining how to read claim terms.” 17 Phillips, 415 F.3d at 1317–18 (citation omitted). Thus, while extrinsic evidence may be useful in 18 claim construction, ultimately “it is unlikely to result in a reliable interpretation of patent claim 19 scope unless considered in the context of the intrinsic evidence.” Id. at 1319 . 20 III. CONSTRUCTION OF DISPUTED TERMS 21 A. “producing sequences of n-bit codewords” 22 UMN’s Proposed LSI’s Proposed Court’s Construction 23 Construction Construction producing, as output, producing a series of producing, as output from the 24 sequences of bits, where each concatenated n-bit codewords, received binary datawords, nsequence is n-bits long such that each received m-bit bit codewords that are [25] binary dataword is encoded combined into a sequence 26 into an n-bit codeword in a manner that is not state27 dependent 1 The parties have two primary disagreements as to the construction of this term. First, they 2 disagree on whether state-dependent codes fall within the scope of the relevant claim language. 3 Second, they disagree on whether it is appropriate to include “concatenated” in the construction of 4 the term. The Court addresses each in turn. 5 1. State-Dependent Codes 6 UMN argues that there is nothing inherent in the claim language that excludes state7 dependent codes. UMN Br. at 9. And in fact, if the term were construed as LSI requests, that 8 construction would create internal inconsistency with dependent claim 19, which expressly claims 9 a state-dependent encoder. UMN Reply at 9. LSI largely ignores the text of the ’601 Patent, 10 instead focusing on what it views as a prosecution disclaimer. It contends that, in oral argument to 11 the Federal Circuit during the appeal of PTAB’s final IPR decision, UMN expressly disclaimed 12 state-dependent codes. LSI Resp. at 6–8. 13 a. Prosecution Disclaimer 14 The Court starts with LSI’s prosecution disclaimer argument. It is unusual that LSI bases 15 its argument on statements made in oral argument before the Federal Circuit. While the doctrine 16 of prosecution disclaimer has been extended to statements made by a patent owner during IPR 17 proceedings, Aylus Networks, Inc. v. Apple Inc., 856 F.3d 1353, 1361 (Fed. Cir. 2017), the parties 18 have not cited, and the Court has been unable to find, any authority extending the doctrine to oral 19 arguments before the Federal Circuit on an appeal from such IPR proceedings. The Court has its 20 doubts that prosecution disclaimer can arise from such oral arguments. But it need not—and does 21 not—seek to resolve those doubts now, because even assuming disclaimer could arise, LSI has not 22 shown that the requirements for disclaimer are met. 23 The purported disavowals of state-dependent codes were far from “clear and 24 unmistakable.” 3M Innovative Props., 725 F.3d at 1325 . Two of the supposed disavowals were 25 based on colloquies between UMN and the Federal Circuit panel in which they spoke in abstract 26 analogy. For example, in one exchange regarding anticipation by a prior art reference, UMN and 27 the panel discussed whether LSI was relying on both “A and B” or just “A” from the prior art 1 reference. Ex. 2 (“Oral Arg. Tr.”) at 20:16–22:5, ECF No. 247-1. And in another exchange, 2 UMN developed an extended metaphor comparing a four-legged stool without a back and a four3 legged stool with a back (a chair). Id. at 22:6–24:15. The disavowal there, according to LSI, 4 came when UMN stated, “We didn’t claim the back. . . . the ’601 patent doesn’t claim the back.” [5] Id. at 24:8–12. It is not at all apparent to the Court what UMN was referring to with “A” or “B” or 6 the “back,” and it is not at all obvious what UMN intended to convey with those analogies. 7 Nor is the third alleged disavowal—UMN’s statement that “I don’t think anyone argued 8 that, that the claims covered the state-dependent aspect of it”—any more explicit. Id. at 24:19–21. 9 To observe that nobody made an argument about state-dependent codes is not the same thing as 10 affirmatively disclaiming such codes from the scope of a term’s meaning. Mere non-argument is 11 hardly an unmistakable disclaimer. 12 Moreover, UMN seemed to suggest at points that state-dependent codes were encompassed 13 by the ’601 Patent. It argued: 14 I think what their argument is is that the state-dependent machine was different and beyond what was disclosed in the [’601] patent. Well, 15 as a matter of fact, that's wrong . . . . [16] Id. at 23:11–14. Thus, at best, UMN’s statements at oral argument were contradictory. That is not 17 enough to find prosecution disclaimer.2 18 b. Analysis of the Intrinsic Evidence 19 Having rejected LSI’s prosecution disclaimer argument, the Court turns to the intrinsic 20 evidence to construe the disputed term. Based on that evidence, the Court agrees with UMN. On 21 its face, nothing in the term, “producing sequences of n-bit codewords,” excludes state-dependent [22] 23 2 LSI briefly argues the PTAB expressly found that the claims of the ’601 Patent do not cover state-dependent codes, LSI Resp. at 6 (citing PTAB Decision at 38–39); however, a review of the 24 PTAB’s decision shows that is not the case. LSI also suggest that “[t]he Federal Circuit panel repeatedly confirmed UMN’s disclaimer during the hearing.” Id. But that argument carries little 25 weight. It should not need to be said that comments made by a panel during oral argument do not constitute a court’s holding. And the very nature of oral argument is that judges will ask probing 26 questions of both parties, sometimes playing devil’s advocate to encourage the parties to better develop a line of reasoning. Given the dynamic and exploratory nature of oral argument, stray 27 comments by judges during argument are not authority, either binding or persuasive, on which this Court can rely. 1 codes; indeed, the term says nothing at all about state-dependent codes. And the specification 2 contemplates the use of state-dependent codes as well. See, e.g., ’601 Patent at col. 4:61–62 3 (“[S]tate-dependent encoders . . . can be designed for the MTR constraint . . . .”) (emphasis 4 added); col. 5:48–67 (describing operation of a state-dependent encoder); col. 6:33–34 (describing 5 an embodiment where “conditional mappings create[] a state dependent encoder with two states”) 6 (emphasis added). Accordingly, the Court finds that state-dependent codes are not excluded from 7 the scope of the term. 8 2. Concatenation 9 LSI argues that the idea of concatenation must be part of the claim construction because 10 the (j;k) constraints must be satisfied at the boundaries of different codewords. LSI Resp. at 5. In 11 its view, concatenation conveys the idea that the codewords are connected from end-to-end and 12 therefore the constraints apply across connected codewords. Id. LSI’s concern is that, absent its 13 proposed concatenation language, a jury might erroneously believe that only each individual 14 codeword need satisfy the (j;k) constraints when the claims actually require the constraints to be 15 satisfied across the entire waveform generated by combining the codewords. UMN does not 16 address the specific boundary condition issue that LSI raises, only generically asserting that LSI 17 failed to cite sufficient evidence to support its construction. UMN Reply at 10. 18 There is ample evidence in the specification that a person of ordinary skill would 19 understand that the produced codewords must apply across the boundaries of different codewords 20 when those codewords are combined. See, e.g., ’601 Patent at col. 5:1–2 (an encoder must “meet 21 the MTR constraint at the codeword boundaries”) (emphasis added); id. at col. 5:48–50, col. 6:2– 22 5 (applying boundary conditions eliminates some potential codewords); id. at col. 7:59–61 23 (describing how an embodiment “satisfy[ies] the MTR j=2 constraint at codeword boundaries”) 24 (emphasis added). So, for example, if there are two codewords, “100” and “001,” it is not enough 25 for each individual codeword to meet the (j;k) constraints. Their combined forms, “100001” and 26 “001100,” must also satisfy the constraints. 27 The portions of the specification identified above inform the meaning of the word 1 “sequences” in the disputed term, making it clear that the idea of a sequence involves 2 concatenating, or combining, the codewords such that the (j;k) constraints must be evaluated 3 across the concatenated codewords rather than only across individual codewords. That said, the 4 Court is hesitant to include the word “concatenate” in its construction because that word is highly 5 technical and, though accurate, may not provide much guidance to a jury. Instead, the Court 6 draws inspiration from the language of claim 17. See Phillips, 415 F.3d at 1314 (“Other claims of 7 the patent in question, both asserted and unasserted, can also be valuable sources of enlightenment 8 as to the meaning of a claim term. . . . Because claim terms are normally used consistently 9 throughout the patent, the usage of a term in one claim can often illuminate the meaning of the 10 same term in other claims.”) (citations omitted). Claim 17 recites, “wherein the binary sequences 11 produced by combining codewords have no more than one of j consecutive transitions.” ’601 12 Patent at col. 11:1–3. That language more explicitly defines a “sequence” as a combination of 13 codewords, and the parties have agreed that the language of claim 17 can be construed as its plain 14 and ordinary meaning. CC Statement at 3.3 15 Therefore, the Court will use the language, “that are combined into a sequence,” rather 16 than “concatenated.” 17 * * * 18 After resolving the parties’ substantive disagreements over the disputed term, all that 19 remains is to settle on the final language of the construction. The Court begins with UMN’s 20 construction because it is more straightforward and will likely be easier for a jury to digest, but it 21 makes some modifications. First, because UMN’s construction describes the codewords as 22 “output,” the Court specifies the input to avoid confusion. Thus, the Court adds the words, “from 23 the received binary datawords,” after the word, “output.” Second, rather than describing the 24 codewords as “sequences of bits”—which may confuse a jury because a “sequence” in the 25 construction refers to an individual codeword while a “sequence” in the disputed term refers to a [27] 3 The parties mistakenly label the language of claim 17 as coming from claim 13. 1 collection of multiple codewords—the Court simply uses the word, “codeword.” The parties have 2 already agreed that “dataword” and “codeword” can be given their plain and ordinary meanings, 3 so the Court has no concern that use of those words in the construction will cause further 4 confusion. See CC Statement at 2. Finally, the Court adds, “that are combined into a sequence,” 5 to describe the collection of codewords, as it indicated in its discussion of concatenation above. 6 In sum, the Court construes this disputed term as: “producing, as output from the received 7 binary datawords, n-bit codewords that are combined into a sequence.” 8 B. “encoded waveform” 9 UMN’s Proposed LSI’s Proposed Court’s Construction Construction Construction [10] the recorded waveform the produced sequences of n- the recorded waveform 11 bit codewords 12 The parties’ dispute over this term centers on the question of how a court should interpret a 13 patentee’s choice to use different terms in a claim. LSI argues that that there “[t]here is an 14 inference, however, that two different terms used in a patent have different meanings.” LSI Resp. 15 at 9 (quoting Comaper Corp. v. Antec, Inc., 596 F.3d 1343, 1348 (Fed. Cir. 2010)). Since claim 16 13 uses both “recorded waveform” and “encoded waveform,” the two terms must have different 17 meanings. Id. at 9–10. UMN counters that the inference is rebuttable, and that in this case, there 18 is good reason to believe that the claim refers to a single waveform that is both encoded and 19 recorded such that “recorded waveform” and “encoded waveform” have the same meaning. 20 UMN has the better of the two arguments on this point. The Federal Circuit has held that, 21 while the use of two different terms in the same claim “gives rise to an inference that a different 22 meaning should be assigned to each” different term, “[t]hat inference [] is not conclusive; it is not 23 unknown for different words to be used to express similar concepts, even though it may be poor 24 drafting practice.” Bancorp Servs., L.L.C. v. Hartford Life Ins. Co., 359 F.3d 1367, 1373 (Fed. 25 Cir. 2004).4 That inference has been rebutted here. “Encoded waveform” is introduced with the [26] 27 4 LSI suggests that Bancorp is inapposite because it deals with indefiniteness rather than claim construction, and that in any case, the Supreme effectively overruled Bancorp in Nautilus, Inc. v. 1 definite article, “the,” which indicates that “encoded waveform” refers to an antecedent basis and 2 has the same meaning as that antecedent basis. X One, Inc. v. Uber Techs., Inc., 440 F. Supp. 3d 3 1019, 1034 (N.D. Cal. 2020) (citations omitted). The only waveform that can serve as an 4 antecedent basis within claim 13 is “a recorded waveform” in the preamble, supporting UMN’s 5 position that “encoded waveform” has the same meaning as “recorded waveform.” 6 Moreover, there is evidence that claim 13 refers only to a single waveform, and the use of 7 “recorded waveform” and “encoded waveform” merely emphasize different characteristics of that 8 single waveform. The preamble of claim 13 describes the invention as “encoding m-bit binary 9 datawords into . . . a recorded waveform,” showing that the “recorded waveform” is also encoded. 10 ’601 Patent at col. 10:47–48; see also McLaughlin Decl. ¶ 45. And the prosecution history, too, 11 supports that conclusion. In initially rejecting claim 13, the patent examiner interchangeably 12 referred to “encoded waveforms” and “recorded waveforms” without differentiating between the 13 two. UMN Br., Ex. 1 at 38, ECF No. 239-1. Likewise, in its response, UMN described its 14 invention by interchangeably referring to “encoded waveforms” and “recorded waveforms” (and 15 at one point, just “the waveform”), all without differentiation. Id. at 47 . Thus, the prosecution 16 history illustrates that drawing a distinction between the two terms was not important to either the 17 patent examiner or UMN.5 18 LSI’s authorities to the contrary are readily distinguishable. In Comaper, the two terms 19 that the court distinguished were “drive bay” and “drive bay slot.” 596 F.3d at 1348 . There, it 20 was appropriate to infer that the two had different meanings because a “bay” is not a “bay slot.” 21 Similarly, in Ethicon Endo-Surgery v. U.S. Surgical Corp., the terms were “pusher assembly” and [22] 23 Biosig Instruments, Inc., 572 U.S. 898 (2014). Neither suggestion is availing here. Bancorp is relevant here insofar as it explains how to interpret claim terms, and there is no reason to think that 24 canons of construction applied in the context of indefiniteness somehow cannot be applied in formal claim construction. And in Nautilus, the Supreme Court opined on the standard for 25 definiteness, not canons of claim construction; therefore Bancorp’s holdings on how to construe claim terms undoubtedly survive Nautilus. 26 5 The fact that UMN introduced “encoded waveform” into claim 13 only after the initial rejection does not suggest that it has a different meaning than “recorded waveform” because the prosecution 27 history does not indicate that UMN or the patent examiner ever drew a distinction between the two terms. 1 “pusher bar.” 93 F.3d 1572 , 1579 (Fed. Cir. 1996). Although the adjectives in Ethicon are the 2 same, the nouns, “assembly” and “bar,” are not. The difference in nouns strengthens the inference 3 that the two terms have different meanings since an assembly is plainly different than a bar. Those 4 circumstances are unlike the ones present here, where the terms “encoded waveform” and 5 “recorded waveform” are based on the same noun. They are both “waveforms,” and differ only in 6 that a different adjective is used to modify “waveform.” In such a situation, the Court can readily 7 conclude that there is a single waveform described alternately as “encoded” or “recorded.” 8 As such, the Court adopts UMN’s construction. 9 C. “transition” 10 UMN’s Proposed LSI’s Proposed Court’s Construction 11 Construction Construction a change in the recording plain and ordinary meaning; a plain and ordinary meaning; a 12 medium from one state to transition can be logically transition can be logically another, which can be represented in multiple ways represented in multiple ways [13] represented in different ways depending on the encoding depending on the recording 14 depending on the recording format used—a change from 0 format used—a change from 0 format to 1 or from 1 to 0 when using to 1 or from 1 to 0 when using 15 NRZ format, for example, or a NRZ format, for example, or a 1 when using NRZI format 1 when using NRZI format [16] 17 The parties’ dispute over this term centers on the question of whether the PTAB’s 18 construction of “transition” in the IPR proceedings has a collateral estoppel effect on the claim 19 construction here. As a threshold matter, the Court finds that the PTAB’s final IPR decision can 20 have a collateral estoppel effect here. In general, PTAB decisions may collaterally estop a federal 21 district court. Papst Licensing GMBH & Co. KG v. Samsung Elecs. Am., Inc., 924 F.3d 1243 , 22 1250–51 (Fed. Cir. 2019) (collecting cases). In the claim construction context, though, the Federal 23 Circuit has previously expressed doubt that estoppel could apply because “the [PTAB] applies the 24 broadest reasonable construction of the claims while the district courts apply a different standard 25 of claim construction as explored in Phillips.” SkyHawke Techs., LLC v. Deca Int'l Corp., 828 26 F.3d 1373, 1376 (Fed. Cir. 2016). Nonetheless, collateral estoppel can apply here because the 27 concerns highlighted in SkyHawke are not present. As the PTAB made clear, when construing 1 “transition,” it applied the same Phillips standard the Court applies now. PTAB Decision at 9. 2 Consequently, the PTAB’s construction collaterally estops the Court if the traditional elements of 3 collateral estoppel are met. 4 When dealing with collateral estoppel in a patent suit, courts apply regional circuit law 5 except as to aspects that have “special or unique application to patent cases.” Aspex Eyewear, Inc. 6 v. Zenni Optical Inc., 713 F.3d 1377, 1380 (Fed. Cir. 2013). Courts in this district typically apply 7 the three-part test set forth in Hydranautics v. FilmTex Corp., 204 F.3d 880 (9th Cir. 2000). 8 Finjan LLC v. SonicWall, Inc., 2021 WL 3111685 , at *2 & n.1 (N.D. Cal. July 22, 2021); see also 9 Droplets, Inc. v. Yahoo! Inc., 2019 WL 5781915 , at *2 & n.2 (N.D. Cal. Oct. 15, 2019). Under 10 that test, Defendants must show “(1) the issue necessarily decided at the previous proceeding is 11 identical to the one which is sought to be relitigated; (2) the first proceeding ended with a final 12 judgment on the merits; and (3) the party against whom collateral estoppel is asserted was a party 13 or in privity with a party at the first proceeding.” Hyrdanautics, 204 F.3d at 885 . 14 UMN does not contest most of these elements, nor could it. It is clear that term being 15 construed here (“transition”) is the same term that the PTAB construed, that the IPR proceedings 16 ended with a final judgment, and that the parties in the IPR proceedings and this action are the 17 same. See PTAB Decision at 10–21; LSI Corp., 43 F.4th 1349 . The only element of estoppel that 18 UMN contests is whether the construction of “transition” was “necessarily decided.” UMN Br. at 19 20; UMN Reply at 13. UMN’s position is that the PTAB’s decision logically required it to reject 20 UMN’s construction that “transition” referred to a magnetic transition, but the decision did not 21 logically require the PTAB to reject the alternate construction that UMN advances now. UMN Br. 22 at 20; UMN Reply at 13. The problem is that this argument has been squarely rejected. The 23 requirement that an issue be “necessarily decided” means only that the previous tribunal actually 24 decided the issue; where “a court hears ‘evidence and argument from both parties, and specifically 25 rule[s] on the issue,’ a party may not escape the rulings bin[d]ing effect on the ground that it was 26 not logically essential to the court’s ultimate determination.” Droplets, 2019 WL 5781915 , at *4 27 (quoting United States v. Johnson, 256 F.3d 895, 915 (9th Cir, 2001)). The PTAB expressly 1 addressed and rejected UMN’s alternate construction, so the issue was “necessarily decided.” 2 || PTAB Decision at 16-18. 3 As result, the Court is collaterally estopped from adopting UMN’s proposed 4 construction, and it adopts LSI’s construction, which is equivalent to the PTAB’s construction, 5 || with one modification. Namely, the Court uses “recording format” rather than “encoding format” 6 || to describe the NRZ and NRZI formats. The ’601 Patent consistently refers to NRZ and NRZI as 7 recording formats. See, e.g., °601 Patent at col. 1:24—25, 30, col. 7:52, 65, col. 11:5-6. And the 8 PTAB, too, refers to them as recording formats. See, e.g., PTAB Decision at 11, 14-15, 21. 9 || Therefore, making this minor modification is more faithful to the °601 Patent while still remaining 10 consistent with the PTAB’s construction. 11 IV. CONCLUSION a 12 For the reasons set forth above, the Court construes the disputed terms as follows: 4 Claim Terms Court’s Construction S producing sequences of n-bit codewords producing, as output from the received binary 2 15 datawords, n-bit codewords that are combined into a sequence [16] 17 transition plain and ordinary meaning; a transition can be logically represented in multiple ways 18 depending on the recording format used—a change from 0 to 1 or from 1 to 0 when using 19 NRZ format, for example, or a 1 when using NRZI format [20] 1 IT IS SO ORDERED. Dated: August 25, 2023 23 ( ) ( ( [24] EDWARD J. DAVILA 5 United States District Judge [27] 28 || Case No.: 5:18-cv-00821-EJD CLAIM CONSTRUCTION ORDER