UNITED STATES DISTRICT COURT FOR THE EASTERN DISTRICT OF TEXAS MARSHALL DIVISION

UNITED STATES DISTRICT COURT FOR THE EASTERN DISTRICT OF TEXAS MARSHALL DIVISION
PLAINTIFF OPTIWAVE SEMICONDUCTOR CORP.'S OPENING CLAIM CONSTRUCTION BRIEF
Filed: November 15, 2024 Docket Entry No. 44
Alan Ngo (CA Bar No. 287451; admitted pro hac vice) Marcus Delgado (TX Bar No. 24098317) Wendy Schultz (CA Bar No. 304219; admitted pro hac vice) Omar Siddiqui (CA Bar No. 311784; admitted pro hac vice) CALDERWOOD & HARKNESS LLP 350 South Grand Avenue, 48th Floor Los Angeles, CA 90071 Telephone: (213) 892-4100 Facsimile: (213) 892-4200
James "Jim" Callahan (TX Bar No. 03879250) WHITMORE & CALLAHAN LLP 110 North College Avenue, Suite 800 Tyler, TX 75702 Telephone: (903) 597-6622
Attorneys for Plaintiff Optiwave Semiconductor Corp.
Right-click to update Table of Contents
TABLE OF AUTHORITIES
Cases.  
Statutes and Rules
Treatises, Technical Dictionaries, and Standards
Journal Articles and Conference Papers
I. INTRODUCTION
Plaintiff Optiwave Semiconductor Corp. ("OWS") filed this patent infringement action against Defendant Lumicore Technologies, Inc. ("Lumicore") on April 22, 2024, asserting four United States utility patents covering innovations in optical transceiver and waveguide multiplexing technologies for silicon photonics applications. The asserted patents are: U.S. Patent No. 10,234,847 (the "'847 Patent"), U.S. Patent No. 10,891,193 (the "'193 Patent"), U.S. Patent No. 9,678,562 (the "'562 Patent"), and U.S. Patent No. 11,402,015 (the "'015 Patent").
These four patents represent foundational innovations in silicon-photonic integrated circuits designed for data-center interconnects — a market segment that has experienced explosive growth as hyperscale cloud operators demand ever-higher bandwidth and lower power consumption from their optical interconnect infrastructure. Specifically, the asserted patents address: (1) multi-channel waveguide multiplexing with low insertion loss, enabling high-density wavelength-division multiplexing on silicon photonic platforms (the '847 Patent); (2) high-bandwidth coherent optical transceivers monolithically or heterogeneously integrated on silicon substrates for next-generation data-center interconnect applications (the '193 Patent); (3) thermal management of photonic integrated circuits using embedded micro-channel heat-sink structures to solve the critical thermal dissipation challenges posed by densely integrated photonic circuits (the '562 Patent); and (4) polarization-diverse grating couplers for silicon photonic transceivers, addressing the challenge of polarization-dependent loss in fiber-to-chip coupling (the '015 Patent).
The Joint Claim Construction Statement filed October 28, 2024 (D.E. 40) identifies fourteen (14) disputed claim terms spanning the four asserted patents. The parties have agreed on the construction of two (2) of those terms, leaving twelve (12) terms for the Court's resolution at the Markman hearing currently scheduled for March 14, 2025. This brief addresses each of the twelve disputed terms and sets forth OWS's proposed constructions.
OWS's proposed constructions faithfully track the plain and ordinary meaning of the claim terms as understood by a person of ordinary skill in the art ("POSITA") at the time of each patent's priority date. OWS's constructions are firmly grounded in the intrinsic record — the claim language, specifications, and prosecution histories of the asserted patents — and, where appropriate, confirmed by extrinsic evidence, including technical dictionaries, Institute of Electrical and Electronics Engineers ("IEEE") standards, peer-reviewed literature, and expert testimony. In contrast, Lumicore's proposed constructions systematically import limitations from preferred embodiments into the independent claims, adopt unjustifiably narrow definitions that contradict the very specifications from which those claims arise, and selectively invoke extrinsic evidence in an attempt to override the clear intrinsic record.
A brief summary of the technology at issue provides essential context. Silicon photonics is the integration of optical components — including waveguides, modulators, photodetectors, and grating couplers — onto silicon substrates using fabrication processes compatible with complementary metal-oxide-semiconductor ("CMOS") manufacturing. This approach leverages the massive existing infrastructure for silicon chip fabrication to produce low-cost, high-performance photonic integrated circuits ("PICs") for high-speed data transmission. The primary commercial application of silicon photonics today is in pluggable transceiver modules for hyperscale data centers, where the technology enables data rates of 400 gigabits per second ("Gb/s"), 800 Gb/s, and 1.6 terabits per second ("Tb/s") per module.
OWS is the third-largest U.S. supplier of silicon-photonics transceiver modules. OWS holds 142 issued U.S. patents and has 58 pending U.S. patent applications. The four patents asserted in this action represent breakthrough innovations that have been commercially embodied in OWS's OWS-400ZR, OWS-800XR, and OWS-1600DR transceiver module families, which generated over $280 million in revenue in fiscal year 2023.
Lumicore's accused products are the LumiLink 400G-ZR+, LumiLink 800G-DR8, and LumiEdge 1600G-LR transceiver modules. As set forth in OWS's P.R. 3-1 Infringement Contentions served September 16, 2024, each of these accused products practices the asserted claims of the relevant patents under OWS's proposed claim constructions. The proper construction of the twelve disputed terms will significantly affect the infringement and validity analyses going forward, and OWS respectfully submits that its proposed constructions — rooted in the plain language of the claims and the full scope of the intrinsic record — should be adopted by the Court.
II. LEGAL STANDARD FOR CLAIM CONSTRUCTION
A. The Court's Role Under *Markman*
Claim construction is a matter of law exclusively for the Court. Markman v. Westview Instruments, Inc., 517 U.S. 370, 372 (1996). The Court's task is to determine the meaning and scope of disputed claim terms as they would be understood by a person of ordinary skill in the art at the time of the invention, in light of the intrinsic evidence and, where appropriate, extrinsic evidence. Id. at 388–89. The Federal Circuit has developed a well-established framework for claim construction, centered on the intrinsic evidence, which OWS respectfully summarizes below.
B. Intrinsic Evidence Framework
Under the framework established by the Federal Circuit's en banc decision in Phillips v. AWH Corp., 415 F.3d 1303 (Fed. Cir. 2005), claim terms are construed based on the intrinsic record, which consists of three components: (1) the claim language itself; (2) the specification, including the written description and drawings; and (3) the prosecution history. Id. at 1314–17. These three sources of intrinsic evidence are considered in an analysis that proceeds as follows.
First, the claims themselves provide substantial guidance. Claim terms are generally given their "plain and ordinary meaning" as understood by a person of ordinary skill in the art at the time of the invention. Phillips, 415 F.3d at 1312–13. The plain and ordinary meaning of a term is not determined in the abstract but rather "in the context of the patent, the prior art, and the particular claim in which the term appears." Id. at 1313; see also Nystrom v. Larkspur Holdings, 424 F.3d 1136, 1142–43 (Fed. Cir. 2005). Other claims in the patent — both asserted and unasserted — can illuminate the meaning of a disputed term, particularly through the doctrine of claim differentiation discussed below. Phillips, 415 F.3d at 1314–15.
Second, the specification is "the single best guide to the meaning of a disputed term." Phillips, 415 F.3d at 1315 (quoting Ridgeway & Co. v. Conceptronic, Inc., 90 F.3d 1576, 1582 (Fed. Cir. 1996)). The specification may reveal the inventor's special definition of a term, in which case that definition governs. Id. at 1316; Thorner v. Severin Industries, 669 F.3d 1362, 1365 (Fed. Cir. 2012). Alternatively, the specification may disavow particular claim scope, in which case the claims must be construed consistently with the disavowal. Phillips, 415 F.3d at 1316; Glenrock Enterprises v. Adriatic Partners LLC, 242 F.3d 1337, 1341 (Fed. Cir. 2001). However, any such disavowal must be "clear and unmistakable." Thorner, 669 F.3d at 1366.
Critically, courts "must not import limitations from the specification into the claims." Phillips, 415 F.3d at 1323. A particular embodiment described in the specification — even the preferred embodiment — does not define the scope of the claims unless the specification clearly and unmistakably limits the invention to that embodiment. Id. The Federal Circuit has repeatedly cautioned against this error. See, e.g., Sundial Group v. IPS Corp., 514 F.3d 1271, 1276 (Fed. Cir. 2008) ("It is well settled that the scope of a patent is not limited to its preferred embodiments."); Ashford Corporation v. Ashgrove Partners LLC, 158 F.3d 1243, 1248 (Fed. Cir. 1998); OnPoint Med., Inc. v. Calderon Group, 95 F.4th 1376, 1383 (Fed. Cir. 2024).
Third, the prosecution history "provides evidence of how the [Patent and Trademark Office ('PTO')] and the inventor understood the patent." Phillips, 415 F.3d at 1317. The prosecution history encompasses the complete record of proceedings before the PTO, including office actions, applicant responses, amendments, examiner interviews, and any reexamination or inter partes review proceedings. Id. Although the prosecution history "can often inform the meaning of the claim language by demonstrating how the inventor understood the invention and whether the inventor limited the invention in the course of prosecution, making the claim scope narrower than it would otherwise be," the prosecution history "often lacks the clarity of the specification and thus is less useful for claim construction purposes." Id.
C. The Role of Extrinsic Evidence
Extrinsic evidence may be used to assist the Court in understanding the technology, the meaning of terms to a POSITA, and the state of the prior art. Phillips, 415 F.3d at 1318–19. Extrinsic evidence includes expert testimony, inventor testimony, technical dictionaries, treatises, journal articles, and standards documents. Id. However, extrinsic evidence "is less significant than the intrinsic record in determining the legally operative meaning of claim language" and may not be used to "contradict claim meaning that is unambiguous in light of the intrinsic evidence." Id. at 1318 (quoting Vitronics, 90 F.3d at 1584). In particular, "a court should discount any expert testimony that is clearly at odds with the claim construction mandated by the claims themselves, the written description, and the prosecution history." Id. at 1318.
D. Claim Differentiation
Under the doctrine of claim differentiation, "the presence of a dependent claim that adds a particular limitation gives rise to a presumption that the limitation in question is not present in the independent claim." Ashford Enterprises v. Dunmore Corporation, 358 F.3d 898, 910 (Fed. Cir. 2004). This doctrine recognizes that each claim in a patent is presumed to have a different scope, and thus a construction of an independent claim that renders a dependent claim superfluous should be disfavored. See Phillips, 415 F.3d at 1314–15; Penmark & Co. v. Int'l Trade Comm'n, 690 F.3d 1318, 1324 (Fed. Cir. 2012). While the presumption of claim differentiation is not irrebuttable, it places a "heavy burden" on the party seeking to overcome it. Liebel-Flarsheim, 358 F.3d at 910.
E. Means-Plus-Function Limitations
Claim limitations that use the phrase "means for" performing a specified function are governed by 35 U.S.C. § 112(f) (pre-AIA § 112 ¶ 6). Under § 112(f), such limitations are construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof. Even in the absence of the word "means," a claim limitation may invoke § 112(f) if it "recites function without reciting sufficient structure for performing that function." Ostrander Holdings, 792 F.3d 1339, 1349 (Fed. Cir. 2015) (en banc). In Williamson, the Federal Circuit overruled the strong presumption against § 112(f) application that had previously attached when a limitation did not recite the word "means," holding instead that the standard inquiry is whether "the words of the claim are understood by persons of ordinary skill in the art to have a sufficiently definite meaning as the name for structure." Id. at 1349.
In this action, the parties have identified potential means-plus-function terms in Claim 9 of the '847 Patent and Claim 15 of the '193 Patent, which are addressed in Sections IV.A.5 and IV.B.6 below.
F. Prosecution History Estoppel
Where an applicant narrows a claim during prosecution to overcome a prior-art rejection, the applicant may be estopped from recapturing the surrendered claim scope through the doctrine of equivalents. Festo Corp. v. Fenmore & Co., 535 U.S. 722, 733–34 (2002). This principle is relevant to claim construction insofar as it informs the public notice function of the claims: the prosecution history is a publicly accessible record that competitors are entitled to rely upon in determining the scope of the patent. The relationship between prosecution history estoppel and claim construction in this case is addressed where relevant in the term-by-term analysis below, particularly with respect to the narrowing amendment made during prosecution of the '015 Patent (Claim 12) and the amendment during prosecution of the '193 Patent (Claims 3 and 7).
III. THE PERSON OF ORDINARY SKILL IN THE ART (POSITA)
Proper claim construction requires the Court to determine the meaning of disputed claim terms as they would be understood by a person of ordinary skill in the art at the time of the invention. Phillips, 415 F.3d at 1312–13. The parties' definitions of the POSITA therefore frame the entire claim construction analysis.
OWS proposes the following POSITA definition, supported by the Declaration of Dr. Caroline Ashford, Professor of Electrical Engineering & Photonics at the Northfield Institute of Technology in Evanston, Illinois (D.E. 47, "Ashford Decl."):
A person with at least a Master of Science (M.S.) degree or Doctor of Philosophy (Ph.D.) degree in electrical engineering, photonics, optical engineering, or a closely related field, and at least three (3) years of professional or academic experience in the design, fabrication, or characterization of silicon-photonic integrated circuits, including experience with waveguide structures, optical transceivers, and/or photonic IC packaging and thermal management.
See Ashford Decl. ¶¶ 18–25. This POSITA definition appropriately captures the level of skill possessed by the engineers and scientists working in the field of silicon photonics during the relevant time frames — a field that, while sophisticated, draws on well-established principles of semiconductor fabrication, integrated optics, and photonic device physics that are taught at the graduate level in electrical engineering and photonics programs. Dr. Ashford notes that the silicon photonics industry employs engineers and researchers with precisely this educational background and experience level, and that the patents-in-suit are written for and addressed to such persons. Ashford Decl. ¶¶ 22–25.
Dr. Ashford is eminently qualified to opine on the appropriate POSITA definition. She holds a Ph.D. in Electrical Engineering from the Massachusetts Institute of Technology (1998), with a dissertation on integrated photonic circuit design for wavelength-division multiplexing. She has served as a Professor of Electrical Engineering & Photonics at Northfield Institute of Technology since 2005, where she leads the Integrated Photonics Laboratory. Dr. Ashford has published over 180 peer-reviewed journal articles and conference papers on silicon photonics, waveguide design, and photonic IC characterization, and holds 12 U.S. patents in the field. She has served as an associate editor of the IEEE Journal of Lightwave Technology and the Journal of the Optical Society of America B, and has been a member of the technical program committees for the Optical Fiber Communication Conference (OFC) and the IEEE Photonics Conference. Ashford Decl. ¶¶ 1–17.
OWS's proposed POSITA definition is consistent with the educational backgrounds and experience of the named inventors of the asserted patents. Dr. Jian-Wei Huang, CTO of OWS and a named inventor on the '847, '193, and '015 Patents, holds a Ph.D. in Electrical Engineering from Stanford University (2003) and has over 20 years of experience in silicon photonics research and development. The co-inventors of the '562 Patent — Drs. Ananya Krishnan, Robert Teller, and Mei-Ling Fong, formerly of Solara Optics, Inc. — hold advanced degrees in electrical engineering and materials science with extensive experience in photonic IC thermal management. See Ashford Decl. ¶¶ 23–24; Declaration of Dr. Jian-Wei Huang (D.E. 48, "Huang Decl.") ¶¶ 4–9.
OWS's POSITA definition is also consistent with the prior art of record. The prior art references cited during prosecution of the asserted patents — including Takahashi et al. (IEEE Photonics Technology Letters, 2014), Wei (OFC 2013), U.S. Patent No. 8,412,007 (Bergmann), and PCT/EP2015/071234 (Simonsen) — were authored by individuals with M.S. or Ph.D. degrees in electrical engineering or photonics and years of experience in silicon-photonic IC design. Ashford Decl. ¶ 25.
The relevant time frame for assessing the POSITA's understanding varies by patent. For the '562 Patent, which was filed on November 22, 2014, the relevant date is circa 2014. For the '847 Patent, which was filed on September 14, 2016, the relevant date is circa 2016. For the '193 Patent, which was filed on April 7, 2017, and claims priority to Provisional Application No. 62/499,117 filed on the same date, the relevant date is circa April 2017. For the '015 Patent, which was filed on August 30, 2019, and claims priority to the same provisional application, the relevant date is likewise circa April 2017 for terms whose construction is informed by the disclosure of the provisional application, and circa August 2019 for terms that rely on matter added in the non-provisional application.
Lumicore's expert, Dr. Rajesh Sundaram, Professor of Electrical & Computer Engineering at Pacific Western University, proposes a POSITA definition that is unduly specialized and narrow. As OWS understands Lumicore's position from the parties' meet-and-confer discussions and the Joint Claim Construction Statement, Dr. Sundaram's proposed POSITA definition would require a Ph.D. in photonics or optical engineering (excluding electrical engineering more broadly), at least five (5) years of professional experience specifically in the design and fabrication of silicon-photonic waveguide multiplexers or coherent transceiver architectures, and familiarity with advanced CMOS fabrication nodes at 45 nm or below. This proposed definition is artificially narrow in several respects. First, it excludes the substantial population of skilled engineers with M.S. degrees who actively design and characterize silicon-photonic circuits in industry — a population that the patents-in-suit are plainly addressed to. Second, by requiring experience specifically in waveguide multiplexers or coherent transceivers, it ignores the reality that the silicon photonics field draws on a common body of knowledge shared across sub-disciplines, including waveguide design, photodetection, thermal management, and packaging. Third, the requirement of familiarity with "advanced CMOS fabrication nodes at 45 nm or below" introduces an irrelevant limitation: while silicon photonics leverages CMOS-compatible processes, the specific transistor node is not central to the optical design principles at issue in this case. See Ashford Decl. ¶¶ 26–27.
The practical effect of Lumicore's unduly narrow POSITA definition is to inflate the level of skill in the art to such a degree that commonly understood terms — such as "waveguide multiplexer," "integrated on a silicon substrate," and "polarization-diverse grating coupler" — would be given artificially technical and restrictive meanings that do not correspond to how these terms are used in the field. OWS respectfully submits that this Court should adopt OWS's proposed POSITA definition as the appropriate baseline for construing the disputed claim terms.
IV. CONSTRUCTION OF THE DISPUTED CLAIM TERMS
A. Terms from the '847 Patent (U.S. Patent No. 10,234,847): "Multi-Channel Silicon Photonic Waveguide Multiplexer with Low Insertion Loss"
Background. The '847 Patent, entitled "Multi-Channel Silicon Photonic Waveguide Multiplexer with Low Insertion Loss," was filed on September 14, 2016, and issued on March 19, 2019. The original assignee is Optiwave Semiconductor Corp. The '847 Patent is in the same patent family as the '562 Patent, being a continuation-in-part of the application that matured into the '562 Patent. The '847 Patent discloses innovations in combining multiple optical signals of different wavelengths onto a single silicon photonic waveguide with minimized insertion loss, enabling high-density wavelength-division multiplexing ("WDM") on silicon photonics platforms. The patent addresses the particular challenges of achieving low-loss multiplexing in the context of silicon waveguides, which have higher propagation losses and tighter fabrication tolerances than silica-based waveguides. The asserted claims of the '847 Patent are Claims 1, 3, 5, 9, and 14.
A.1. "waveguide multiplexer".  OWS's Proposed Construction: "an optical device that combines multiple optical signals of different wavelengths onto a single waveguide"
Lumicore's Proposed Construction: "an arrayed waveguide grating that combines optical signals using a slab coupler and phase-shifted waveguide array"
##### (a) Claim Language
Independent Claim 1 of the '847 Patent recites, in relevant part:
"A silicon photonic integrated circuit comprising: a waveguide multiplexer configured to combine a plurality of optical channels, each optical channel carrying an optical signal at a distinct wavelength, onto a common output waveguide, wherein the waveguide multiplexer is configured to achieve an insertion loss of less than 1.5 dB per channel across a wavelength range spanning the C-band . . . ."
The claim language itself is revealing. The claim recites "a waveguide multiplexer" — not "an arrayed waveguide grating," not "a slab-coupler-based multiplexer," and not any other specific multiplexer architecture. The term is used in its generic, functional sense: it identifies a device that performs the function of combining (multiplexing) multiple optical channels onto a common waveguide. The claim language does not specify or require any particular internal architecture or physical mechanism for achieving the multiplexing function. The only structural limitation on the multiplexer is that it must be "configured to achieve an insertion loss of less than 1.5 dB per channel across a wavelength range spanning the C-band" — a performance characteristic, not a structural limitation on the type of multiplexer.
A POSITA reading Claim 1 would understand "waveguide multiplexer" as a generic term encompassing any optical device that combines multiple signals of different wavelengths onto a single waveguide, consistent with the broad, plain-meaning use of this term in the art. See Ashford Decl. ¶ 28.
##### (b) Specification
The specification of the '847 Patent confirms that "waveguide multiplexer" is a broad, genus-level term. The specification describes multiple embodiments of waveguide multiplexers, including, but not limited to:
•  Arrayed waveguide gratings ("AWGs"): Col. 4, ll. 15–45 describes an embodiment in which the waveguide multiplexer comprises an AWG with a first slab coupler, a phased array of waveguides with incrementally varying path lengths, and a second slab coupler. This is described as "one embodiment" of the waveguide multiplexer. Col. 4, l. 15 ("In one embodiment, the waveguide multiplexer 110 comprises an arrayed waveguide grating . . . .").
•  Echelle gratings: Col. 5, ll. 40–60 describes an alternative embodiment in which "the waveguide multiplexer comprises an echelle diffraction grating formed in the silicon photonic substrate, the grating facets providing wavelength-dependent diffraction to combine multiple channels." This embodiment is described as "particularly suited for compact multiplexer designs."
•  Mach-Zehnder interferometer ("MZI")-based combiners: Col. 6, ll. 5–30 describes "yet another embodiment" in which the waveguide multiplexer comprises a cascade of Mach-Zehnder interferometric stages, each stage configured to combine two adjacent wavelength channels, with the stages cascaded in a binary-tree arrangement.
•  Cascaded ring-resonator multiplexers: Col. 7, ll. 10–35 describes "a further embodiment" in which the waveguide multiplexer comprises a plurality of microring resonators, each tuned to a specific channel wavelength, arranged along a common bus waveguide.
Most significantly, the specification contains an express, definitional statement of the breadth of the term. At column 3, lines 55–62, the specification states:
"The waveguide multiplexer may take any of several forms known to those skilled in the art, including but not limited to arrayed waveguide gratings, echelle diffraction gratings, cascaded ring resonators, and Mach-Zehnder-based combiners. The particular multiplexer architecture may be selected based on the required channel count, channel spacing, insertion loss budget, and available die area."
This language is unambiguous. The inventors defined "waveguide multiplexer" as a term encompassing any multiplexer architecture known to a POSITA. The specification's use of the phrase "including but not limited to" followed by a non-exhaustive list of four different architectures confirms that the term is not limited to any single architecture. See Phillips, 415 F.3d at 1316 (the specification "may reveal a special definition given to a claim term by the patentee that differs from the meaning it would otherwise possess"); see also Westvale Enterprises v. AgiLight, Inc., 750 F.3d 1304, 1310 (Fed. Cir. 2014) (construing term broadly where specification described multiple embodiments).
##### (c) Prosecution History
The prosecution history of the '847 Patent further supports OWS's broad construction. During prosecution, the examiner issued an initial rejection under 35 U.S.C. § 102(a)(1) based on Nakamura (U.S. Patent No. 9,122,003), which disclosed an AWG-based multiplexer on a silicon photonic substrate. In response, the applicant amended Claim 1 to add the insertion-loss limitation ("configured to achieve an insertion loss of less than 1.5 dB per channel across a wavelength range spanning the C-band") and argued that the claimed invention was distinguishable from Nakamura because Nakamura's AWG did not achieve the recited low insertion loss. See Prosecution History, Response to Office Action dated March 12, 2018, at pp. 7–9 (D.E. 45, Appx. at A-0124 to A-0126).
Critically, the applicant did not narrow the term "waveguide multiplexer" to any specific architecture. The applicant's remarks make clear that the point of novelty was the performance of the multiplexer (low insertion loss), not the type of multiplexer:
"The claimed waveguide multiplexer is distinguished from Nakamura not by the multiplexer architecture per se, but by the achievement of insertion loss of less than 1.5 dB per channel across the C-band, which Nakamura neither teaches nor suggests. The claimed low insertion loss is achieved through the novel techniques described in the specification, including optimized waveguide transition tapers (col. 8, ll. 25–50) and apodized grating designs (col. 9, ll. 5–30), which are applicable to any multiplexer architecture."
Response to Office Action dated March 12, 2018, at p. 8 (emphasis added). This prosecution history statement expressly confirms that the applicant understood "waveguide multiplexer" as encompassing any multiplexer architecture, and that the applicant did not intend to narrow the term.
##### (d) Extrinsic Evidence
Standard technical dictionaries and references used by persons of ordinary skill in the art define "waveguide multiplexer" broadly, consistent with OWS's proposed construction. The Fiber Optic Reference Guide (3d ed., Penrose Holdings, 2002), a widely used industry reference, defines "multiplexer" as "[a] device that combines two or more signals into a single output" and defines "waveguide multiplexer" as "[a] multiplexer implemented in waveguide form, combining multiple optical signals onto a common waveguide." See Ashford Decl. ¶ 30. Predding & Knights, Silicon Photonics: An Introduction (Crowne Albany Industries, 2004), similarly describes waveguide multiplexers as a genus of devices that includes multiple architectures such as AWGs, echelle gratings, and ring-resonator-based combiners. Ashford Decl. ¶ 31. Doerr, "Silicon Photonic Integration in Telecommunications," Frontiers in Physics, vol. 3, art. 37 (2015), surveys the landscape of silicon photonic multiplexer architectures and uses the generic term "waveguide multiplexer" to refer to the entire class. Ashford Decl. ¶ 32.
Dr. Ashford confirms that, in her expert opinion, a POSITA at the time of the invention would have understood "waveguide multiplexer" to mean any optical device that combines multiple optical signals of different wavelengths onto a single waveguide, without limitation to a specific architecture. Ashford Decl. ¶¶ 28–35.
##### (e) Claim Differentiation
The doctrine of claim differentiation provides compelling support for OWS's broad construction and independently refutes Lumicore's narrow construction. Dependent Claim 3 of the '847 Patent specifies: "The silicon photonic integrated circuit of claim 1, wherein the waveguide multiplexer comprises an arrayed waveguide grating." Dependent Claim 5 specifies: "The silicon photonic integrated circuit of claim 1, wherein the waveguide multiplexer comprises a cascaded ring-resonator structure."
The presence of these dependent claims, each adding a specific multiplexer-architecture limitation, gives rise to a strong presumption that independent Claim 1 does not require any particular multiplexer architecture. Liebel-Flarsheim, 358 F.3d at 910. Lumicore's proposed construction of "waveguide multiplexer" as "an arrayed waveguide grating that combines optical signals using a slab coupler and phase-shifted waveguide array" would render dependent Claim 3 entirely superfluous, because independent Claim 1 would already require an AWG. This result is precisely the type of claim-construction error that the doctrine of claim differentiation is designed to prevent. See id.; Phillips, 415 F.3d at 1314–15; Aspex Eyewear, Inc. v. Strathmore Industries, 672 F.3d 1335, 1341 (Fed. Cir. 2012) ("[W]e have frequently relied on the doctrine of claim differentiation when construing claims.").
##### (f) Rebuttal of Lumicore's Construction
Lumicore's proposed construction — "an arrayed waveguide grating that combines optical signals using a slab coupler and phase-shifted waveguide array" — is improper for multiple independent reasons.
First, Lumicore's construction improperly imports a single preferred embodiment into the independent claim. The specification describes four different multiplexer architectures — AWGs, echelle gratings, MZI-based combiners, and ring-resonator multiplexers — and expressly states that the "waveguide multiplexer may take any of several forms." Col. 3, ll. 55–62. Lumicore's construction limits the claim to only one of those forms, in direct contravention of the Federal Circuit's admonition that courts "must not import limitations from the specification into the claims." Phillips, 415 F.3d at 1323.
Second, Lumicore's construction renders dependent Claim 3 superfluous, violating the doctrine of claim differentiation as discussed above.
Third, Lumicore's construction adds structural requirements — "slab coupler" and "phase-shifted waveguide array" — that appear nowhere in the claim language. Even within the AWG embodiment, these are specific sub-components of the AWG architecture described in the specification, not elements recited in the claims. Importing sub-component-level detail from the specification into the claims goes far beyond even an impermissible embodiment import.
Fourth, as discussed above, the prosecution history affirmatively demonstrates that the applicant did not narrow "waveguide multiplexer" to AWGs. The applicant's remarks during prosecution expressly stated that the invention's novelty lay in low insertion loss "applicable to any multiplexer architecture," not in the selection of a particular architecture.
Fifth, the extrinsic evidence uniformly supports OWS's broader construction, and Lumicore has identified no technical dictionary, textbook, or standard that limits "waveguide multiplexer" to AWG architectures.
##### (g) Conclusion
For all of the foregoing reasons, the Court should adopt OWS's proposed construction of "waveguide multiplexer": "an optical device that combines multiple optical signals of different wavelengths onto a single waveguide."
A.2. "insertion loss of less than [X] dB per channel".  OWS's Proposed Construction: "the reduction in optical power, expressed in decibels, experienced by each optical channel as it passes through the waveguide multiplexer, measured as the ratio of output power to input power for each channel"
Lumicore's Proposed Construction: "the reduction in optical power measured at the output of the waveguide multiplexer under standardized test conditions at 25°C with a single-mode fiber input, excluding coupling losses from fiber-to-chip interfaces"
##### (a) Claim Language
Independent Claim 1 of the '847 Patent recites that the waveguide multiplexer is "configured to achieve an insertion loss of less than 1.5 dB per channel across a wavelength range spanning the C-band." The claim recites a performance threshold — less than 1.5 dB per channel — but does not specify particular measurement conditions, test temperatures, or the exclusion of coupling losses.
The term "insertion loss" is a well-understood term of art in the optical and photonics community. A POSITA would understand "insertion loss" in this context as the total optical power loss incurred by each channel as it passes through the multiplexer from its input port to the common output waveguide. Ashford Decl. ¶ 36.
##### (b) Specification
The specification of the '847 Patent defines insertion loss consistently with its ordinary meaning in the art. At column 2, lines 40–48, the specification states:
"As used herein, 'insertion loss' refers to the loss of optical power resulting from the insertion of the waveguide multiplexer in the optical path, and is expressed in decibels (dB) as the ratio of the optical power at the output of the multiplexer to the optical power at the input of the multiplexer for a given channel wavelength."
The specification does not limit the measurement to any particular temperature, does not require single-mode fiber input (as opposed to on-chip waveguide input), and does not require the exclusion of fiber-to-chip coupling losses. To the contrary, the specification describes insertion-loss measurements in multiple configurations, including both fiber-coupled and on-chip test configurations. Col. 10, ll. 5–30 describes a measurement setup using on-chip test structures with integrated photodetectors; col. 10, ll. 35–55 describes a fiber-coupled measurement configuration using external tunable lasers and optical spectrum analyzers.
The specification also describes insertion-loss performance at various temperatures, including room temperature (25°C), elevated temperatures (up to 85°C), and cryogenic temperatures (down to -40°C), demonstrating that the claimed insertion-loss performance is intended to be achieved across a range of operating conditions. Col. 11, ll. 10–45. There is no indication that the specification limits the insertion-loss measurement to any single temperature.
##### (c) Prosecution History
During prosecution, the insertion-loss limitation was added by amendment to distinguish the prior art (Nakamura). The applicant's remarks focused on the claimed multiplexer's ability to achieve less than 1.5 dB per channel insertion loss — a performance metric — without specifying or narrowing the conditions under which the measurement is to be performed. The examiner accepted this amendment and allowed the claims without requiring any specific measurement protocol. See Prosecution History, Response to Office Action dated March 12, 2018, at pp. 7–9; Notice of Allowance dated June 22, 2018.
##### (d) Extrinsic Evidence
The term "insertion loss" has a standard, universally accepted meaning in the optical engineering community. IEEE Standard 802.3cu-2020, for example, defines insertion loss as "the decrease in power between the input and the output of a device, measured in decibels." Ashford Decl. ¶ 37. This definition does not require any specific measurement conditions or the exclusion of particular loss components. Similarly, the Fiber Optic Reference Guide defines insertion loss as "the loss of signal power resulting from the insertion of a device in a transmission line, expressed in dB." Ashford Decl. ¶ 38.
Dr. Ashford opines that Lumicore's attempt to import specific measurement conditions (25°C, single-mode fiber input, exclusion of coupling losses) into the construction of "insertion loss" has no basis in the plain and ordinary meaning of the term, the specification, or the prosecution history. Ashford Decl. ¶¶ 36–41.
##### (e) Rebuttal of Lumicore's Construction
Lumicore's proposed construction improperly adds three extraneous limitations: (1) measurement at 25°C only, (2) single-mode fiber input, and (3) exclusion of coupling losses from fiber-to-chip interfaces. None of these limitations appears in the claim language. None is required by the specification, which describes measurements at multiple temperatures and in both fiber-coupled and on-chip configurations. None was required by the prosecution history. Lumicore's construction would transform a straightforward performance metric — insertion loss — into a narrow measurement protocol that finds no support in the intrinsic record.
The practical effect of Lumicore's construction is transparent: by restricting the measurement conditions, Lumicore seeks to argue that its accused products do not meet the insertion-loss threshold when measured under the artificially narrow conditions it proposes, even though they would meet the threshold under the standard measurement protocols used in the industry. This is precisely the type of litigation-driven construction that the Federal Circuit has rejected. See Severin Holdings v. Quillon Corporation, 289 F.3d 801, 812 (Fed. Cir. 2002) (rejecting construction designed to avoid infringement where not supported by intrinsic record).
##### (f) Conclusion
The Court should adopt OWS's proposed construction: "the reduction in optical power, expressed in decibels, experienced by each optical channel as it passes through the waveguide multiplexer, measured as the ratio of output power to input power for each channel."
A.3. "plurality of optical channels".  OWS's Proposed Construction: "two or more optical channels"
Lumicore's Proposed Construction: "four or more optical channels operating at distinct wavelengths within a single wavelength band"
##### (a) Claim Language
Independent Claim 1 recites "a plurality of optical channels, each optical channel carrying an optical signal at a distinct wavelength." The term "plurality" is one of the most well-established terms in patent drafting, universally understood to mean "two or more." The claim language does not specify a minimum number of channels beyond two, and does not require that all channels operate within a "single wavelength band."
##### (b) Specification
The specification describes embodiments with various channel counts, including 4-channel (col. 4, ll. 50–55), 8-channel (col. 5, ll. 10–15), 16-channel (col. 6, ll. 35–40), and 32-channel (col. 7, ll. 45–50) configurations. The specification does not define "plurality" as requiring any minimum number greater than two. The specification also describes embodiments in which channels span multiple wavelength bands (e.g., C-band and L-band), directly contradicting Lumicore's proposed "single wavelength band" limitation. Col. 8, ll. 55–65 ("In one embodiment, the waveguide multiplexer is configured to operate across both the C-band (approximately 1530–1565 nm) and the L-band (approximately 1565–1625 nm), combining channels from both bands onto the common output waveguide.").
##### (c) Prosecution History
The prosecution history contains no indication that the applicant defined or narrowed "plurality" to mean four or more, or limited the channels to a single wavelength band. No amendments were made with respect to this limitation.
##### (d) Extrinsic Evidence
Dr. Ashford confirms that "plurality" is universally understood in the art to mean "two or more." Ashford Decl. ¶ 43. No technical dictionary or standard defines "plurality of optical channels" as requiring four or more channels or operation within a single wavelength band. Id.
##### (e) Rebuttal of Lumicore's Construction
Lumicore's proposed construction improperly raises the minimum channel count from two to four and adds a "single wavelength band" limitation that is contradicted by the specification. The clear purpose of Lumicore's construction is to exclude configurations in which its accused products operate with fewer than four channels or span multiple wavelength bands. There is no basis for this construction in the intrinsic or extrinsic evidence.
##### (f) Conclusion
The Court should adopt OWS's proposed construction: "two or more optical channels."
A.4. Claim Differentiation Analysis — '847 Patent Independent Claim 1 vs. Dependent Claims 3 and 5.  As discussed in detail in Section IV.A.1 above, the claim-differentiation analysis strongly supports OWS's broad construction of "waveguide multiplexer" in independent Claim 1. Dependent Claims 3 and 5 each narrow the generic "waveguide multiplexer" of Claim 1 to specific architectures:
•  Claim 3: "The silicon photonic integrated circuit of claim 1, wherein the waveguide multiplexer comprises an arrayed waveguide grating."
•  Claim 5: "The silicon photonic integrated circuit of claim 1, wherein the waveguide multiplexer comprises a cascaded ring-resonator structure."
Lumicore argues that the independent claim should be construed narrowly because, it contends, the specification primarily discusses the AWG embodiment and the AWG is the "core" multiplexer architecture disclosed. But this argument conflates the specification's detailed disclosure of multiple embodiments with an intent to limit the claims to only one. The Federal Circuit has squarely rejected the argument that the degree of detail devoted to a particular embodiment limits the independent claim to that embodiment. See Pemberton Group v. Hollowell Industries, 483 F.3d 800, 807 (Fed. Cir. 2007) ("That the specification may describe a particular embodiment in greater detail does not mean that the claim is limited to that embodiment."); Phillips, 415 F.3d at 1323.
Moreover, the specification devotes substantial disclosure to each of the four multiplexer architectures identified — AWG, echelle grating, MZI-based, and ring-resonator — providing sufficient written description and enablement support for the full scope of Claim 1. The fact that the AWG embodiment may be described in somewhat greater detail is unremarkable, as it represents the preferred commercial implementation, but this does not limit the independent claim.
Lumicore has not overcome the "heavy burden" required to overcome the presumption of claim differentiation. Liebel-Flarsheim, 358 F.3d at 910. There is no intrinsic evidence — no express definition, no disavowal, no prosecution history narrowing — that would rebut the presumption that Claim 1 is broader than Claims 3 and 5.
A.5. Means-Plus-Function Analysis — Claim 9 of the '847 Patent.  Claim 9 of the '847 Patent recites, in pertinent part: "a channel equalization unit for adjusting the optical power of each of the plurality of optical channels to achieve a uniform output power level across all channels." The parties dispute whether this limitation invokes 35 U.S.C. § 112(f).
##### (a) Whether § 112(f) Is Invoked
Although Claim 9 does not recite the word "means," the limitation recites a "unit for" performing a specified function — "adjusting the optical power of each of the plurality of optical channels to achieve a uniform output power level across all channels." The term "unit" is a nonce word that, standing alone, does not connote sufficiently definite structure to a POSITA. See Williamson, 792 F.3d at 1350 (holding that the term "module for" invoked § 112(f) because "module" is a nonce word that does not connote structure).
OWS acknowledges that this limitation likely invokes § 112(f). The question then becomes whether the specification discloses sufficient corresponding structure.
##### (b) Corresponding Structure in the Specification
The specification discloses the corresponding structure for the "channel equalization unit" at column 9, lines 15–45, which describes:
"The channel equalization unit 160 comprises a plurality of variable optical attenuators (VOAs) 162, each VOA coupled to a respective input waveguide carrying one of the optical channels. Each VOA 162 comprises a Mach-Zehnder interferometric structure with a thermo-optic phase shifter 164 on one arm. A feedback control circuit 166 monitors the optical power of each channel at the output of the waveguide multiplexer 110 using integrated photodetectors 168, and adjusts the attenuation of each VOA 162 to equalize the optical power across all channels."
The specification further describes an alternative structure at column 9, lines 50–65: "In an alternative embodiment, the channel equalization unit comprises a semiconductor optical amplifier (SOA) array, each SOA coupled to a respective channel, with a control circuit adjusting the gain of each SOA to equalize channel powers."
Thus, the corresponding structures disclosed in the specification are: (1) a plurality of variable optical attenuators (VOAs) with Mach-Zehnder interferometric structures, thermo-optic phase shifters, feedback control circuits, and integrated photodetectors; and (2) a semiconductor optical amplifier (SOA) array with a control circuit. Under § 112(f), the claim limitation covers these disclosed structures and their equivalents.
OWS's construction of this means-plus-function limitation is: Function: adjusting the optical power of each of the plurality of optical channels to achieve a uniform output power level across all channels. Corresponding Structure: a plurality of variable optical attenuators (VOAs) comprising Mach-Zehnder interferometric structures with thermo-optic phase shifters and a feedback control circuit with integrated photodetectors, or a semiconductor optical amplifier (SOA) array with a control circuit, and equivalents thereof. '847 Patent, col. 9, ll. 15–65.
B. Terms from the '193 Patent (U.S. Patent No. 10,891,193): "High-Bandwidth Coherent Optical Transceiver Integrated on Silicon Substrate"
Background. The '193 Patent, entitled "High-Bandwidth Coherent Optical Transceiver Integrated on Silicon Substrate," was filed on April 7, 2017, and issued on January 12, 2021. It claims priority to U.S. Provisional Application No. 62/499,117, filed April 7, 2017, which is shared with the '015 Patent. The original assignee is Optiwave Semiconductor Corp. The '193 Patent discloses a coherent optical transceiver architecture that can be monolithically or heterogeneously integrated on a silicon substrate, enabling high-bandwidth operation at data rates of 400 Gb/s and above for data-center interconnect applications. The key innovation is the integration of a coherent detection front end — including a 90-degree optical hybrid, balanced photodetectors, and transimpedance amplifiers — together with a coherent modulation transmitter on a common silicon photonic platform. The asserted claims of the '193 Patent are Claims 1, 3, 7, 10, 15, and 18.
B.1. "integrated on a silicon substrate".  OWS's Proposed Construction: "formed on or within a silicon-based wafer using semiconductor fabrication processes"
Lumicore's Proposed Construction: "monolithically fabricated on a single silicon die without hybrid bonding or heterogeneous integration"
##### (a) Claim Language
Independent Claim 1 of the '193 Patent recites: "A coherent optical transceiver integrated on a silicon substrate, the transceiver comprising: a coherent modulator . . . ; a coherent detection module . . . ; and a local oscillator laser coupled to the coherent detection module . . . ." The claim uses the phrase "integrated on a silicon substrate" without specifying the method of integration. The term "integrated on" is broad and encompassing: it describes a spatial and functional relationship between the transceiver components and the silicon substrate, but does not dictate the fabrication process by which that relationship is achieved.
A POSITA would understand "integrated on a silicon substrate" to mean that the transceiver components are formed on or within a silicon-based wafer using any of the semiconductor fabrication processes known in the art for integrating photonic and electronic components on silicon, including monolithic, heterogeneous, and hybrid integration. Ashford Decl. ¶ 42.
##### (b) Specification
The specification of the '193 Patent is unambiguous on this point. It describes three distinct integration approaches:
•  Monolithic integration: Col. 5, ll. 20–50 describes a preferred embodiment in which "all photonic and electronic components of the coherent optical transceiver are monolithically integrated on a single silicon-on-insulator (SOI) wafer using standard CMOS-compatible fabrication processes." This embodiment uses silicon waveguides, germanium photodetectors epitaxially grown on silicon, and silicon-based modulators (carrier-depletion or carrier-injection type).
•  Heterogeneous integration: Col. 7, ll. 1–25 describes an alternative embodiment using "heterogeneous integration, in which III-V semiconductor materials (such as InP or InGaAsP) are bonded to the silicon substrate via wafer bonding techniques (e.g., direct bonding, adhesive bonding, or oxide-mediated bonding)." In this embodiment, the laser source and, optionally, the photodetectors are fabricated from III-V materials bonded to the silicon photonic platform.
•  Hybrid integration: Col. 8, ll. 10–30 describes a further alternative in which "separately fabricated components, such as a III-V laser die and a transimpedance amplifier (TIA) die, are attached to the silicon photonic substrate using micro-bump or flip-chip bonding." This hybrid approach allows the use of best-in-class components from different material platforms on a common silicon substrate.
The specification contains an express statement of breadth at column 4, lines 58–63:
"The transceiver may be integrated on the silicon substrate using any suitable fabrication technique, including monolithic integration, heterogeneous integration via wafer bonding, or hybrid integration using micro-bump or flip-chip bonding. The choice of integration technique may depend on the performance requirements of the particular component and the available fabrication infrastructure."
This language is clear: the patentee defined "integrated on a silicon substrate" to encompass all three integration approaches. Any construction that limits the term to monolithic integration alone would exclude two of the three expressly disclosed and enabled embodiments.
##### (c) Prosecution History
During prosecution of the '193 Patent, the examiner did not require the applicant to limit "integrated on a silicon substrate" to monolithic integration. The prosecution history does not contain any narrowing amendment or argument with respect to this term. The applicant distinguished the prior art based on the coherent detection architecture — specifically, the integration of the 90-degree optical hybrid, balanced photodetectors, and local oscillator laser on the silicon platform — not on the particular method of integration. See Prosecution History, Response to Office Action dated February 14, 2020, at pp. 5–11 (D.E. 45, Appx. at A-0287 to A-0293).
There is no clear and unmistakable disavowal of heterogeneous or hybrid integration in the prosecution history. The applicant's arguments during prosecution consistently referred to the "silicon photonic platform" and the "silicon substrate" without any statement limiting these terms to monolithic fabrication.
##### (d) Extrinsic Evidence
The extrinsic evidence uniformly supports OWS's construction. The term "integrated on a silicon substrate" is widely understood in the silicon photonics community to encompass monolithic, heterogeneous, and hybrid integration approaches. See Ashford Decl. ¶¶ 42–55. Dr. Ashford cites numerous survey articles in the IEEE Journal of Lightwave Technology and conference proceedings from the Optical Fiber Communication Conference (OFC) that use the phrase "silicon photonic integration" to encompass all three approaches. Ashford Decl. ¶¶ 45–50.
Indeed, as Dr. Ashford explains, by the priority date of the '193 Patent (April 2017), heterogeneous integration was the predominant commercial approach for integrating laser sources on silicon photonic platforms, because silicon's indirect bandgap makes monolithic laser integration exceptionally challenging. Ashford Decl. ¶ 51. A construction that excluded heterogeneous integration would exclude the very approach that was most commercially significant at the time of the invention. Id.
##### (e) Rebuttal of Lumicore's Construction
Lumicore's proposed construction — "monolithically fabricated on a single silicon die without hybrid bonding or heterogeneous integration" — is improper for several independent reasons.
First, it improperly imports the preferred monolithic-integration embodiment into the claims, excluding two of the three integration methods expressly disclosed and enabled in the specification. This is the paradigmatic claim-construction error cautioned against in Phillips. 415 F.3d at 1323.
Second, Lumicore's construction would effectively read out of the claims the heterogeneous-integration and hybrid-integration embodiments that the patentee described and enabled. The Federal Circuit has held that a construction that excludes a disclosed embodiment is "rarely, if ever, correct." Oatey, 514 F.3d at 1276; Acumed, 483 F.3d at 806 ("[A] claim interpretation that excludes a preferred embodiment from the scope of the claim is rarely, if ever, correct." (internal citation omitted)).
Third, there is no clear and unmistakable disavowal of heterogeneous or hybrid integration in the specification or prosecution history. The specification's express statement that the transceiver "may be integrated on the silicon substrate using any suitable fabrication technique, including monolithic integration, heterogeneous integration via wafer bonding, or hybrid integration" (col. 4, ll. 58–63) is the opposite of a disavowal — it is an affirmative embrace of all three approaches.
Fourth, the practical motivation behind Lumicore's construction is transparent. Lumicore's LumiLink 400G-ZR+ and LumiLink 800G-DR8 transceiver modules employ heterogeneous integration, bonding III-V laser dies onto silicon photonic substrates using wafer bonding techniques. By limiting "integrated on a silicon substrate" to monolithic fabrication, Lumicore seeks to exclude its own products from the claim scope. This litigation-driven construction finds no support in the intrinsic record.
##### (f) No Disavowal
OWS emphasizes that the threshold for finding disavowal of claim scope is high: the disavowal must be "clear and unmistakable." Thorner, 669 F.3d at 1366. There is no statement in the specification or prosecution history of the '193 Patent that clearly and unmistakably limits "integrated on a silicon substrate" to monolithic integration. The specification's description of the monolithic embodiment as "preferred" (col. 5, l. 22) does not constitute disavowal; the Federal Circuit has repeatedly held that the mere description of a preferred embodiment does not limit the claims. Oatey, 514 F.3d at 1276.
##### (g) Conclusion
The Court should adopt OWS's proposed construction: "formed on or within a silicon-based wafer using semiconductor fabrication processes."
B.2. "Coherent detection module".  OWS's Proposed Construction: "a module that detects optical signals using coherent detection, which involves mixing the received signal with a local oscillator to recover amplitude, phase, and polarization information"
Lumicore's Proposed Construction: "a module comprising a 90-degree optical hybrid, balanced photodetectors, and transimpedance amplifiers, that recovers both in-phase and quadrature components of the received signal"
##### (a) Claim Language
Independent Claim 1 of the '193 Patent recites "a coherent detection module configured to receive an optical input signal and mix the optical input signal with a local oscillator signal to generate electrical output signals representative of the amplitude and phase of the optical input signal." The claim language defines the coherent detection module by its function — receiving an optical signal, mixing it with a local oscillator, and generating electrical outputs representative of amplitude and phase — rather than by specifying its internal sub-components.
A POSITA would understand "coherent detection module" as a functional unit that performs coherent detection, without being limited to any single implementation of the internal optical and electronic sub-components. Ashford Decl. ¶ 56.
##### (b) Specification
The specification describes multiple embodiments of the coherent detection module. The preferred embodiment (col. 6, ll. 5–35) includes a 90-degree optical hybrid, balanced photodetectors, and transimpedance amplifiers (TIAs), which together recover in-phase (I) and quadrature (Q) components of the received signal. An alternative embodiment (col. 8, ll. 35–55) describes a simplified coherent detection module using a 3×3 coupler-based hybrid instead of a 90-degree optical hybrid, which the specification notes is suitable for lower-complexity modulation formats such as QPSK. A further alternative (col. 9, ll. 5–20) describes an embodiment using single-ended photodetectors rather than balanced photodetectors, with subsequent digital signal processing to compensate for the loss of common-mode rejection.
The specification does not define "coherent detection module" as requiring any specific sub-component or architecture. Rather, the specification defines the module by its function — coherent detection — and discloses multiple structural implementations capable of performing that function.
##### (c) Prosecution History and Relevance to Doctrine of Equivalents
During prosecution, the applicant amended Claims 3 and 7 to narrow certain aspects of the coherent detection module in response to a § 102 rejection over Takahashi et al., "Dense WDM Silicon Photonic MUX" (IEEE Photonics Technology Letters, March 2014). Specifically, dependent Claim 3 was amended to require "wherein the cohere