DCT

1:19-cv-01945

VoiceAge EVS LLC v. HMD Global Oy

Log In

I. Executive Summary and Procedural Information

  • Parties & Counsel:
  • Case Identification: [VoiceAge EVS LLC](https://ai-lab.exparte.com/party/voiceage-evs-llc) v. [HMD Global Oy](https://ai-lab.exparte.com/party/hmd-global-oy), 1:19-cv-01945, D. Del., 10/15/2019
  • Venue Allegations: Plaintiff alleges that venue is proper because Defendant is a foreign corporation that may be sued in any judicial district.
  • Core Dispute: Plaintiff alleges that Defendant’s mobile devices, by implementing the industry-standard Enhanced Voice Services (EVS) audio codec, infringe five patents related to digital speech and audio compression technology.
  • Technical Context: The EVS codec is a foundational technology for high-definition voice services (VoLTE) on modern 4G/LTE mobile networks, designed to deliver higher quality audio more efficiently than previous standards.
  • Key Procedural History: The complaint alleges that all asserted patents have been declared essential to the 3GPP EVS standard. Plaintiff also alleges it sent letters to HMD on August 21, 2019, and September 16, 2019, inviting HMD to license the patent portfolio, which may be relevant to allegations of willful infringement.

Case Timeline

Date Event
2002-05-31 ’710 Patent Priority Date
2006-10-24 ’843 Patent Priority Date
2007-06-22 ’073 Patent Priority Date
2010-04-06 ’710 Patent Issue Date
2011-05-11 ’475 Patent Priority Date
2013-03-19 ’843 Patent Issue Date
2014-01-01 3GPP adopts the EVS codec standard (approx. date)
2014-04-17 ’741 Patent Priority Date
2014-09-02 ’475 Patent Issue Date
2015-03-24 ’073 Patent Issue Date
2016-01-01 T-Mobile becomes first U.S. carrier to support EVS (approx. date)
2017-12-26 ’741 Patent Issue Date
2019-08-21 Plaintiff sends first pre-suit letter to Defendant
2019-09-16 Plaintiff sends second pre-suit letter to Defendant
2019-10-15 Complaint Filing Date

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

U.S. Patent No. 7,693,710 - "Method and device for efficient frame erasure concealment in linear predictive based speech codecs"

  • Patent Identification: U.S. Patent No. 7,693,710, "Method and device for efficient frame erasure concealment in linear predictive based speech codecs," issued April 6, 2010 (Compl. ¶47).

The Invention Explained

  • Problem Addressed: The patent addresses the degradation of synthesized speech quality caused by the erasure or loss of data frames during transmission over wireless or packet-switched networks (Compl. ¶49; ’710 Patent, col. 11:18-21). This loss can desynchronize the decoder from the encoder, causing audible artifacts, especially when low bit-rate encoders rely on pitch prediction from previous frames (’710 Patent, col. 11:46-57).
  • The Patented Solution: The invention proposes a method where the encoder analyzes the original sound signal, classifies it (e.g., as unvoiced, voiced, voiced transition), and determines "concealment/recovery parameters" such as signal energy and phase information (Compl. ¶52-53). These parameters are transmitted to the decoder, which then uses this information from the last good frame to more effectively conceal the lost frame and accelerate recovery once good frames are received again (’710 Patent, col. 2:58-63; Compl. ¶54). This allows the concealment strategy to adapt to the specific type of speech signal where the error occurred (’710 Patent, col. 11:58-12:5).
  • Technical Importance: This technique provides more robust and higher-quality voice communication over unreliable networks by intelligently managing the impact of packet loss (Compl. ¶48).

Key Claims at a Glance

  • The complaint asserts independent device claim 15 (Compl. ¶98, 100).
  • The essential elements of independent claim 15 are:
    • A device for conducting concealment of frame erasure.
    • In the encoder, a determiner of concealment/recovery parameters selected from a group including a signal classification parameter, an energy information parameter, and a phase information parameter.
    • A communication link for transmitting these parameters to the decoder.
    • The decoder conducts frame erasure concealment and recovery in response to the received parameters.
    • The parameters include the phase information parameter, which is determined by a searcher that finds the position of a first glottal pulse and a quantizer that encodes that position.

U.S. Patent No. 8,401,843 - "Method and device for coding transition frames in speech signals"

  • Patent Identification: U.S. Patent No. 8,401,843, "Method and device for coding transition frames in speech signals," issued March 19, 2013 (Compl. ¶56).

The Invention Explained

  • Problem Addressed: In Code-Excited Linear Prediction (CELP) codecs, the heavy reliance on prediction from past frames means that a single transmission error (an erased frame) can propagate through subsequent frames, desynchronizing the decoder and causing a "very annoying" perceptual impact (Compl. ¶59; ’843 Patent, col. 2:10-17).
  • The Patented Solution: The patent discloses a "transition mode (TM) encoding technique." For specific frames, such as the onset of a voiced sound, this technique replaces the standard "adaptive codebook," which is dependent on past excitation, with a new "glottal-shape codebook" (Compl. ¶60). This new codebook is fixed and independent of past excitation, allowing the encoder and decoder to use the same excitation information immediately after a frame erasure is over, which enables rapid "convergence to clean-channel synthesis" (’843 Patent, col. 5:59-6:5).
  • Technical Importance: This approach improves the robustness of CELP-based codecs against transmission errors by eliminating error propagation without adding delay, complexity, or increasing the bit rate (Compl. ¶61).

Key Claims at a Glance

  • The complaint asserts independent device claim 11 (Compl. ¶109-110).
  • The essential elements of independent claim 11 are:
    • A device for generating a transition mode excitation.
    • A generator of a codebook search target signal.
    • A transition mode codebook for generating a set of codevectors independent from past excitation.
    • The transition mode codebook comprises a codebook of glottal impulse shapes.
    • A searcher of the transition mode codebook for finding the codevector that optimally corresponds to the codebook search target signal.

U.S. Patent No. 8,990,073 - "Method and device for sound activity detection and sound signal classification"

  • Patent Identification: U.S. Patent No. 8,990,073, "Method and device for sound activity detection and sound signal classification," issued March 24, 2015 (Compl. ¶63).
  • Technology Synopsis: The patent addresses the technical problem of sound activity detection (SAD), background noise estimation, and sound signal classification (Compl. ¶64). It discloses using "tonal stability estimation" to improve the performance of SAD algorithms, particularly in the presence of music, and to better discriminate between unvoiced sounds and music signals (Compl. ¶65-66).
  • Asserted Claims: Independent claim 31 and dependent claim 36 (Compl. ¶120).
  • Accused Features: The accused products' alleged capability to perform sound activity detection by estimating the tonal stability of a sound signal using its frequency spectrum, in compliance with the EVS Standard (Compl. ¶120-122).

U.S. Patent No. 8,825,475 - "Transform-domain codebook in a CELP coder and decoder"

  • Patent Identification: U.S. Patent No. 8,825,475, "Transform-domain codebook in a CELP coder and decoder," issued September 2, 2014 (Compl. ¶68).
  • Technology Synopsis: The patent addresses the challenge of improving encoded speech quality in ACELP models at higher bitrates, where simply increasing codebook size is inefficient (Compl. ¶69-70). The solution involves modifying the CELP model by adding a "transform-domain codebook stage" to the existing time-domain codebook stages, allowing for better control over how additional bits are allocated across different frequency components (Compl. ¶71).
  • Asserted Claims: Independent claim 1 and dependent claim 3 (Compl. ¶131).
  • Accused Features: The accused products' alleged use of a CELP codebook structure that includes both an innovative codebook stage and a transform-domain codebook stage, along with a selector that can change the order of these stages based on characteristics of the input signal and the codec's bit rate (Compl. ¶132-133).

U.S. Patent No. 9,852,741 - "Methods, encoder and decoder for linear predictive encoding and decoding of sound signals upon transition between frames having different sampling rates"

  • Patent Identification: U.S. Patent No. 9,852,741, "Methods, encoder and decoder for linear predictive encoding and decoding of sound signals upon transition between frames having different sampling rates," issued December 26, 2017 (Compl. ¶74).
  • Technology Synopsis: The patent addresses the technical problem of switching between different bit rates that use different internal sampling rates in a multi-rate coder without introducing audible artifacts (Compl. ¶76). The invention provides a computationally efficient method for converting linear predictive (LP) filter parameters between different sampling rates by operating on the power spectrum of the filter, which avoids the need to re-sample the entire past synthesis signal (Compl. ¶77-78).
  • Asserted Claims: Independent claim 17 and dependent claim 20 (Compl. ¶142).
  • Accused Features: The accused products' alleged ability to switch between frames using different internal sampling rates (S1 and S2) by computing the power spectrum of the LP synthesis filter at S1, modifying it (by extending or truncating), and then inverse transforming it to compute the LP filter parameters at S2 (Compl. ¶143-144).

III. The Accused Instrumentality

Product Identification

  • HMD’s mobile devices and other devices with EVS codec capabilities, referred to as "HMD's EVS Products," with the Nokia 6.1 cited as a specific example (Compl. ¶80).

Functionality and Market Context

  • The accused products are mobile devices that include hardware and software to implement the EVS codec as defined by the 3GPP EVS Standard (Compl. ¶82). This functionality enables high-quality voice calls, marketed as "Enhanced HD Voice" or "Ultra HD Voice," over modern wireless networks (Compl. ¶9, 82). The complaint alleges that because the accused products are compliant with the EVS Standard, they necessarily infringe the VoiceAge patents, which are asserted to be essential to that standard (Compl. ¶83-84). The complaint includes a chart comparing the spectral bandwidth of the EVS codec to older standards, illustrating its technical superiority (Compl. ¶7).

IV. Analysis of Infringement Allegations

’710 Patent Infringement Allegations

Claim Element (from Independent Claim 15) Alleged Infringing Functionality Complaint Citation Patent Citation
a device for conducting concealment of frame erasure...comprising: in the encoder, a determiner of concealment/recovery parameters selected from the group consisting of a signal classification parameter, an energy information parameter and a phase information parameter related to the sound signal... HMD’s EVS Products implement the EVS codec, which determines concealment parameters as described in 3GPP TS 26.445 §§ 4.1, 4.4, and 5.5. ¶98, 100 col. 2:58-65
the concealment/recovery parameters include the phase information parameter; to determine the phase information parameter, the determiner comprises a searcher of a position of a first glottal pulse in a frame of the encoded sound signal... The EVS codec determines a phase information parameter that includes searching for the position of a glottal pulse, as described in 3GPP TS 26.447 § 5.3. ¶100 col. 22:46-51
and the searcher measures a sample of maximum amplitude within a pitch period as the first glottal pulse, and the determiner comprises a quantizer of the position of the sample of maximum amplitude within the pitch period. The EVS codec measures the maximum amplitude and quantizes its position within the pitch period as part of determining the phase information parameter, per 3GPP TS 26.447 § 5.3. ¶100 col. 22:52-65
  • Identified Points of Contention:
    • Scope Questions: The infringement theory hinges on the patents being essential to the EVS standard. A primary point of contention may be whether compliance with the cited sections of the 3GPP standards necessarily results in a device that practices every limitation of the asserted claims, or if non-infringing alternatives could exist within the standard.
    • Technical Questions: A potential dispute may arise over whether the specific technical implementation of "phase information" determination in the EVS standard is functionally identical to the claimed "searcher," "measure," and "quantizer" of a "first glottal pulse."

’843 Patent Infringement Allegations

Claim Element (from Independent Claim 11) Alleged Infringing Functionality Complaint Citation Patent Citation
a device for generating a transition mode excitation...comprising: a transition mode codebook for generating a set of codevectors independent from past excitation... HMD's EVS Products implement a transition mode that replaces an adaptive codebook excitation with one that is independent of past excitation, per 3GPP TS 26.445 §§ 4.1, 4.4, 5.1, and 5.2. ¶109, 110 col. 2:56-62
wherein the transition mode codebook comprises a codebook of glottal impulse shapes... The transition mode codebook used in the EVS-compliant products is alleged to be a codebook of glottal impulse shapes as described in the standard. ¶110 col. 5:59-6:5
a searcher of the transition mode codebook for finding the codevector of said set corresponding to the transition mode excitation optimally corresponding to the codebook search target signal. HMD's EVS Products are alleged to search the transition mode codebook to find the optimal codevector for the target signal, as required by the EVS standard. ¶110 col. 4:45-50
  • Identified Points of Contention:
    • Scope Questions: The case may turn on whether the term "codebook of glottal impulse shapes," as defined and used in the ’843 Patent, has the same scope and meaning as the specific codebook structures implemented in the EVS standard for transition frames.
    • Technical Questions: A factual question may be whether the search process for finding an "optimally corresponding" codevector in the accused devices performs the same function in the same way to achieve the same result as the "searcher" required by the claim.

V. Key Claim Terms for Construction

’710 Patent

  • The Term: "concealment/recovery parameters"
  • Context and Importance: This term defines the core information that must be determined by the encoder and transmitted to the decoder to practice the invention. The scope of this term is central to the infringement analysis, as the data transmitted by the accused EVS codec must fall within its definition.
  • Intrinsic Evidence for Interpretation:
    • Evidence for a Broader Interpretation: Claim 15 introduces the terms "signal classification parameter, an energy information parameter and a phase information parameter" with the phrase "selected from the group consisting of," which may suggest the list is exemplary, not exhaustive (’710 Patent, claim 15).
    • Evidence for a Narrower Interpretation: The specification repeatedly details these specific parameters, and the complaint highlights a statement from the prosecution history that emphasizes the novelty of "classifying successive frames as unvoiced, unvoiced transition, voiced transition, voiced, or onset, and calculating an energy information parameter" (Compl. ¶55). This could support an interpretation limiting the term to parameters that perform these specific functions.

’843 Patent

  • The Term: "codebook of glottal impulse shapes"
  • Context and Importance: This term defines the specific type of codebook that replaces the conventional adaptive codebook in transition frames. Infringement depends on whether the codebook used in the EVS standard for transition frames meets this definition.
  • Intrinsic Evidence for Interpretation:
    • Evidence for a Broader Interpretation: The specification describes the invention as replacing the adaptive codebook with a "non-predictive, for example glottal-shape, codebook search" (’843 Patent, col. 2:58-59). This phrasing might suggest that a "glottal-shape" codebook is merely an example of a broader category of "non-predictive" codebooks.
    • Evidence for a Narrower Interpretation: The patent's abstract, title, and detailed description consistently and specifically refer to a "codebook of glottal impulse shapes," suggesting it is a specific and essential feature of the invention, not just an example (’843 Patent, Abstract; col. 5:59-6:5). The complaint also cites the prosecution history, which explicitly mentions that the prior art did not teach a transition mode codebook comprising a "codebook of glottal impulse shapes" (Compl. ¶62).

VI. Other Allegations

  • Indirect Infringement: The complaint alleges inducement of infringement against HMD (Compl. ¶86-87). The factual basis for this allegation is that HMD provides its customers with EVS-capable mobile devices and user guides with instructions on how to make phone calls, knowing that when these devices are used on an EVS-enabled network, the end-users will directly infringe the patents (Compl. ¶91-94).
  • Willful Infringement: The complaint alleges willful infringement based on pre-suit knowledge (Compl. ¶101-102, 112-113). The alleged bases for knowledge are: (1) HMD’s awareness, as a sophisticated telecommunications company, that the asserted patents were declared essential to the EVS standard (Compl. ¶41); and (2) HMD’s receipt of letters from VoiceAge EVS on August 21, 2019, and September 16, 2019, which provided notice of the patent portfolio (Compl. ¶88).

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

  • A central issue will be one of standard-essentiality and claim scope: does compliance with the cited provisions of the 3GPP EVS standard necessarily require practicing every element of the asserted claims? The dispute will likely focus on whether non-infringing design alternatives could exist within the standard's framework, which would challenge the allegation that the patents are truly essential.
  • A key technical question will be one of definitional and functional equivalence: do the specific algorithms and data structures defined in the EVS standard operate in a manner that is functionally identical to the methods and structures claimed in the patents? This will require a detailed comparison between the standard's implementation of concepts like "transition mode" and "frame erasure concealment" and the specific claim language, such as "codebook of glottal impulse shapes" and "concealment/recovery parameters."