Ramot At Tel Aviv University Ltd v. Cisco Systems Inc

I. Executive Summary and Procedural Information

• Parties & Counsel:
  • Plaintiff: Ramot at Tel Aviv University, Ltd. (Israel)
  • Defendant: Cisco Systems, Inc. (Delaware)
  • Plaintiff’s Counsel: Capshaw DeRieux, LLP; Bunsow De Mory LLP
• Case Identification: 1:23-cv-00012, E.D. Tex., 08/15/2022
• Venue Allegations: Plaintiff alleges venue is proper in the Eastern District of Texas because Defendant Cisco maintains regular and established places of business in the district, including facilities in Richardson and Allen, Texas, and has committed acts of infringement in the district.
• Core Dispute: Plaintiff alleges that Defendant’s high-speed optical transceiver modules, digital signal processors (DSPs), and related components infringe a patent for linearizing the output of optical modulators through a digital mapping technique that pre-compensates for known signal distortions.
• Technical Context: The technology addresses the inherent non-linear response of optical modulators, a critical problem in high-bandwidth fiber-optic communication systems where signal fidelity is paramount for achieving faster data rates.
• Key Procedural History: The complaint details an extensive litigation history between the parties involving related predecessor patents. This history includes a 2019 lawsuit, multiple unsuccessful Inter Partes Review (IPR) petitions filed by Cisco against related patents, and a declaratory judgment action filed by Cisco concerning the patent-in-suit on the same day it issued, suggesting a high degree of pre-suit awareness of the technology and infringement contentions.

Case Timeline

Date	Event
2007-06-13	’998 Patent Priority Date
2014-11-05	First related patent suit filed by Ramot against Cisco
2019-06-12	Second related patent suit filed by Ramot against Cisco
2021-02-26	Ramot sues Cisco subsidiary Acacia over related patents
2021-03-01	Cisco completes acquisition of Acacia Communications, Inc.
2022-05-24	U.S. Patent No. 11,342,998 Issues
2022-05-24	Cisco files declaratory judgment action on the ’998 Patent
2022-08-15	Complaint Filing Date

II. Technology and Patent(s)-in-Suit Analysis

U.S. Patent No. 11,342,998 - "Linearized Optical Digital-to-Analog Modulator"

• Patent Identification: U.S. Patent No. 11,342,998, issued May 24, 2022.

The Invention Explained

• Problem Addressed: The patent’s background section explains that widely used optical modulators, such as the Mach-Zehnder Interferometer (MZI), suffer from an "inherent non-linear response" (’998 Patent, col. 1:63-64). When converting a digital signal to an analog optical signal, this non-linearity produces a distorted, "cosine-shaped" output instead of a desired linear one, which limits the performance, dynamic range, and resolution of high-speed communication systems (’998 Patent, col. 2:1-6, 20-22).
• The Patented Solution: The invention proposes a system that corrects this non-linearity in the digital domain before the signal becomes optical. Instead of a direct one-to-one mapping of input data bits to modulator electrodes, it employs a "digital-to-digital converter" (DDC). This DDC intelligently maps an N-bit input data word to a different, M-bit output "actuation vector" (where M is typically greater than N) that drives the modulator’s electrodes (’998 Patent, Abstract). The mapping is not trivial; it provides the "freedom to choose the electrode actuation pattern which best approximates a desired ideal output for the given input," effectively pre-compensating for the modulator's non-linear behavior (’998 Patent, col. 7:31-34).
• Technical Importance: This digital pre-distortion technique was designed to enable "high-performance and large bandwidth digital to analog signal conversion," a key requirement for advancing "multi-GHz mixed-signal systems" in fields like wireless communications, defense radar, and high-speed data networks (’998 Patent, col. 1:38-48).

Key Claims at a Glance

• The complaint asserts independent claim 1 and notes infringement of "one or more claims" (Compl. ¶63, ¶85).
• The essential elements of independent claim 1 include:
  • An optical modulation system comprising an input for N digital input data bits, an input optical signal, and a modulator.
  • A "digital-to-digital mapping" that converts the N input bits to a set of M digital output bits, which are associated with a plurality of voltage values used to modulate the input optical signal.
  • The mapping is defined by a specific mathematical property: for a "first subset" of successively increasing input values, the incremental changes ("deltas") between corresponding output values "decrease."
  • For a "second subset" of successively increasing input values, the deltas between corresponding output values "increase."
• The complaint states that Cisco infringes "one or more claims of the Asserted Patent, including without limitation, claim 1," which reserves the right to assert other claims (Compl. ¶63).

III. The Accused Instrumentality

Product Identification

The complaint identifies two broad categories of accused products: "Acacia Accused Products" and "Cisco Accused Products" (Compl. ¶42, ¶57, fn 1). These include Cisco's coherent optical transceiver modules (e.g., AC1200, CIM8, QSFP-DD), standalone Digital Signal Processor (DSP) ASICs (e.g., Jannu, Greylock, Pico), and Silicon Photonic Integrated Circuits (PICs) developed by its subsidiary Acacia. The complaint also accuses Cisco's broader portfolio of routers and switches (e.g., Nexus and NCS series) that incorporate these components or similar functionality (Compl. ¶25-27, ¶43, ¶45-47).

Functionality and Market Context

The accused products are components for high-speed (100Gbps to over 1.2Tbps) optical networks used in data centers, metropolitan, and long-haul applications (Compl. ¶25, ¶28). The complaint alleges that a central feature of these products is their use of sophisticated DSPs to perform "digital pre-equalization, pre-distortion, shaping, and non-linearity compensation" (Compl. ¶33, ¶42, ¶80). This functionality is marketed with terms like "3D Shaping" and "Non-linear compensation" and is alleged to be necessary to "overcome fiber transmission impairments" and compensate for distortions from analog and optical components at high speeds (Compl. ¶33, ¶39, ¶41). The complaint includes a block diagram of the accused AC1200 module, which shows a "PICO DSP Engine" processing data before it is sent to the optical modulator driver (Compl. ¶30, p. 12, Fig. 7).

IV. Analysis of Infringement Allegations

’998 Patent Infringement Allegations

Claim Element (from Independent Claim 1)	Alleged Infringing Functionality	Complaint Citation	Patent Citation
an input for a plurality of N digital input data bits	The accused DSP ASICs and transceiver modules accept parallel digital data, such as multi-bit symbols or words, from a host interface.	¶65	col. 2:47
an input optical signal	The accused Silicon Photonic ICs incorporate an unmodulated laser source whose light is fed into the waveguide branches of a Mach-Zehnder modulator.	¶68	col. 2:49
a modulator for modulating the input optical signal...to output a modulation of the input optical signal	The accused products employ optical modulators, such as Mach-Zehnder modulators, within their Silicon Photonic ICs to modulate the input laser signal and generate a modulated optical output for transmission over fiber.	¶70	col. 2:48-52
a digital-to-digital mapping maps the plurality of N digital input data bits to a set of M digital output data bits associated with a plurality of voltage values	The DSP ASICs in the accused products perform digital signal processing to pre-equalize, pre-distort, or shape signals, which the complaint alleges constitutes the claimed "digital mapping." This is marketed as "3D Shaping" and "Non-linear compensation."	¶75, ¶80	col. 7:6-8
for a first subset of successively increasing digital input values... deltas between numerical values of successive digital outputs... decrease	The complaint alleges that the digital signal processing used to alter the modulator's optical response results in mapping subsets where the deltas between successive mapped digital outputs decrease.	¶83	col. 17:41-48
for a second subset of successively increasing digital input values... deltas between numerical values of successive digital outputs... increase	The complaint alleges that the same digital signal processing for altering the modulator's optical response results in other mapping subsets where the deltas between successive mapped digital outputs increase.	¶84	col. 17:50-57

Identified Points of Contention

• Scope Questions: The case may turn on the construction of "digital-to-digital mapping." A central question is whether this term, which the patent specification exemplifies with structures akin to a look-up table (’998 Patent, col. 9:61-62), can be interpreted to cover the real-time, algorithmic "digital signal processing" for "non-linear compensation" and "3D-Shaping" allegedly performed by Cisco's DSPs (Compl. ¶39, ¶41).
• Technical Questions: A key evidentiary hurdle will be demonstrating that the accused functionality meets the specific mathematical limitations of the claim. What evidence does the complaint provide that Cisco's generalized "non-linear compensation" algorithms necessarily result in the specific claimed behavior of creating subsets with "decreasing" and "increasing" deltas between successive output values? The complaint alleges this "on information and belief" as a consequence of linearizing the optical response but does not present direct technical proof such as simulations or measurements (Compl. ¶83, ¶84). The complaint's included visual shows a block diagram for a coherent transceiver where the DSP performs "Impairment equalization," a feature described as for "Equalization of imperfections," which may or may not map directly to the specific linearization technique claimed (Compl. ¶32, p. 14).

V. Key Claim Terms for Construction

• The Term: "digital-to-digital mapping"

  • Context and Importance: This term defines the core inventive concept. Its scope is critical because if it is construed narrowly, Cisco's DSP-based algorithmic compensation might fall outside the claims. Practitioners may focus on this term because the patent's description of a "digital-to-digital converter" that can be implemented as an ASIC or look-up table must be reconciled with the accused products' use of complex, real-time processing engines.
  • Intrinsic Evidence for Interpretation:
    • Evidence for a Broader Interpretation: The specification describes the DDC's function as providing "freedom to choose the electrode actuation pattern which best approximates a desired ideal output," a functional description that could encompass real-time algorithms (’998 Patent, col. 7:31-34). It also contemplates implementation in a high-speed "Application Specific Integrated Circuit ('ASIC')" (’998 Patent, col. 8:8-12), which is what the complaint accuses.
    • Evidence for a Narrower Interpretation: The patent also describes the DDC in terms that suggest a more static, pre-calculated relationship, referring to it as potentially a "look-up table" and stating that for each digital input vector Dᵢ, the DDC "produces a corresponding binary vector Bᵢ" (’998 Patent, col. 9:61-62, col. 10:48-51). This could support an argument that the term requires a fixed mapping rather than dynamic, on-the-fly calculation.

• The Term: "deltas between numerical values of successive digital outputs ... decrease [and] increase"

  • Context and Importance: These limitations quantify the "linearization" function of the mapping. Infringement requires proof that the accused products' output exhibits these specific mathematical properties.
  • Intrinsic Evidence for Interpretation:
    • Evidence for a Broader Interpretation: The patent explains this behavior as the mechanism for correcting a "cosine-shaped output variation" (’998 Patent, col. 2:5-6). Plaintiff may argue that any effective linearization of such a curve will inherently produce regions of decreasing and increasing slope (deltas), making these limitations a natural and necessary result of the "non-linear compensation" function that Cisco advertises (Compl. ¶41, ¶81).
    • Evidence for a Narrower Interpretation: The claim language requires discrete "subsets" where these properties hold. A defendant could argue that its compensation algorithms do not operate in a way that creates such cleanly defined subsets or that the property is not consistently met, thereby avoiding literal infringement. The patent itself shows idealized graphs and tables, which might be used to argue for a strict, ordered interpretation of this limitation.

VI. Other Allegations

• Indirect Infringement: The complaint alleges inducement by citing Cisco's public-facing marketing materials, including "Blog posts, Presentations, Whitepapers, and Videos," which allegedly instruct and encourage customers on how to use the infringing features of the accused products (Compl. ¶87). Contributory infringement is alleged on the basis that components like the DSP ASICs and PICs are material to the invention, are not staple articles of commerce, and are especially made or adapted for use in an infringing manner (Compl. ¶88).
• Willful Infringement: The willfulness claim is supported by extensive allegations of pre-suit knowledge. The complaint asserts that Cisco was aware of Ramot's related patents and infringement allegations since at least 2014, engaged in prior litigation over the technology family, unsuccessfully challenged related patents in IPRs, and filed a declaratory judgment action on the ’998 patent on the very day it issued (Compl. ¶89-94). A marketing video screenshot included in the complaint highlights "Non-linear compensation" as a key benefit of its DSP technology, which may be used as evidence of intent (Compl. ¶41, p. 22).

VII. Analyst’s Conclusion: Key Questions for the Case

• A core issue will be one of definitional scope: can the term "digital-to-digital mapping," which the patent illustrates with look-up tables and fixed vector selections, be construed to cover the dynamic, algorithmic "non-linear compensation" and "3D-Shaping" performed by the accused DSPs?
• A key evidentiary question will be one of functional proof: can the plaintiff demonstrate that the generalized "impairment equalization" performed by the accused products results in the specific mathematical behavior required by Claim 1—namely, the creation of distinct input subsets where the deltas between corresponding output values systematically decrease and increase?
• The extensive history of litigation and Cisco's proactive filing of a declaratory judgment action on the day the patent issued raises a significant question regarding willfulness: given the documented awareness, the court will have to determine whether Cisco's continued conduct constitutes the type of "egregious" behavior that warrants enhanced damages should infringement be found.