DCT
2:24-cv-00353
Advanced Coding Tech LLC v. Google LLC
Key Events
Amended Complaint
I. Executive Summary and Procedural Information
- Parties & Counsel:
- Plaintiff: Advanced Coding Technologies LLC (Texas)
- Defendant: Google LLC (Delaware)
- Plaintiff’s Counsel: Fabricant LLP; McKool Smith, P.C.
- Case Identification: 2:24-cv-00353, E.D. Tex., 08/02/2024
- Venue Allegations: Plaintiff alleges venue is proper in the Eastern District of Texas based on Defendant’s regular and established places of business, including Google Global Cache (GGC) servers located in Sherman, Tyler, and Texarkana; Google Wi-Fi infrastructure at Starbucks locations; Google Fi telecommunications services; and Google Cloud Interconnect facilities in Denton County.
- Core Dispute: Plaintiff alleges that Defendant’s products and services that encode, decode, or deliver video using the AV1 codec, as well as products implementing 5G NR technology and network content delivery systems, infringe six patents related to video compression, media servers, and communication quality assessment.
- Technical Context: The technologies at issue relate to video coding/decoding (codecs) for efficient digital video transmission, content delivery networks (CDNs) for distributing media, and methods for ensuring wireless communication quality, all of which are foundational to modern internet streaming and mobile communication.
- Key Procedural History: The complaint asserts that Google was a founding member of the Alliance for Open Media (AOM), the entity that developed the accused AV1 video standard. It alleges that AOM conducted "patent due diligence" during AV1's development, which Plaintiff suggests put Google on notice of the patents-in-suit. The complaint also alleges notice of certain patents through prior litigations involving Google’s direct competitors and through citations made during the prosecution of Google’s own patents.
Case Timeline
| Date | Event |
|---|---|
| 2004-03-31 | ’891 Patent Priority Date |
| 2007-03-02 | ’101 Patent Priority Date |
| 2008-05-30 | ’025, ’995, and ’448 Patents Priority Date |
| 2010-09-28 | ’891 Patent Issue Date |
| 2012-01-03 | ’025 Patent Issue Date |
| 2012-07-24 | ’101 Patent Issue Date |
| 2013-06-03 | Alleged Notice Date for ’995 and ’448 Patents via prosecution of Google patent |
| 2014-03-31 | ’303 Patent Priority Date |
| 2015-05-26 | ’448 Patent Issue Date |
| 2015-09-01 | Alliance for Open Media (AOM) founded, with Google as a founding member |
| 2018-05-29 | ’303 Patent Issue Date |
| 2019-02-26 | ’995 Patent Issue Date |
| 2021-10-19 | Google Pixel 6 launch, an accused product |
| 2024-08-02 | First Amended Complaint Filing Date |
II. Technology and Patent(s)-in-Suit Analysis
U.S. Patent No. 8,090,025 - "Moving-Picture Coding Apparatus Method and Program, and Moving-Picture Decoding Apparatus, Method and Program," Issued January 3, 2012 (’025 Patent)
The Invention Explained
- Problem Addressed: The patent addresses the problem of "block distortion," a visual artifact in compressed video where the boundaries between adjacent processing blocks become visible, creating a discontinuous, blocky appearance (U.S. Patent No. 8,090,025, col. 1:45-51). This issue is particularly noticeable at low bitrates.
- The Patented Solution: The invention proposes a method to create a smoother predictive picture during the encoding process. Instead of just copying blocks from a reference frame, it uses the pixel values at the borders of a block as a "boundary condition" to mathematically generate an estimated video signal within the block that satisfies Poisson's Equation, a partial differential equation. This ensures continuity with adjacent blocks before the final residual signal is calculated and encoded (’025 Patent, col. 2:53-col. 3:2).
- Technical Importance: This approach aimed to improve video compression efficiency by creating a better predictive picture, which in turn reduces the amount of data needed to encode the residual error, leading to better visual quality at the same bitrate or a lower bitrate for the same quality (Compl. ¶79).
Key Claims at a Glance
- The complaint asserts at least independent claim 10 (’025 Patent, col. 36:51-38:3; Compl. ¶94).
- Claim 10 (A moving-picture decoding method) essential elements:
- demultiplexing coded data from an input signal based on a specific syntax structure;
- performing inverse-quantization and inverse-orthogonal transform to produce a decoded residual picture;
- defining a boundary condition of a border that corresponds to border motion-vector data from a reference picture;
- generating an estimated video signal in each rectangular zone that satisfies Poisson’s Equation, thus producing a first predictive picture;
- combining the first predictive picture and the decoded residual picture to generate a decoded moving-picture signal;
- storing the decoded moving-picture signal as a reference picture.
- The complaint does not explicitly reserve the right to assert dependent claims for this patent.
U.S. Patent No. 9,986,303 - "Video Image Coding Data Transmitter, Video Image Coding Data Transmission Method, Video Image Coding Data Receiver, and Video Image Coding Data Transmission and Reception System," Issued May 29, 2018 (’303 Patent)
The Invention Explained
- Problem Addressed: The patent addresses the need for efficient transmission and reception of video data that includes different resolutions or hierarchies, a common requirement for adaptive bitrate streaming over networks with variable bandwidth (U.S. Patent No. 9,986,303, col. 1:22-31).
- The Patented Solution: The invention describes a system that handles two streams: "basic video image coding data" and "supplementary video image coding data." The supplementary data includes pictures ("supplementary hierarchical picture") that are ordered earlier in both coding and display than pictures in the basic stream. This structure allows a receiver to reconstruct video by combining the basic data with supplementary data that may have arrived earlier, enabling flexible handling of different video layers or resolutions (’303 Patent, Abstract).
- Technical Importance: This technology facilitates adaptive streaming, where a service like YouTube can switch between lower-resolution (basic) and higher-resolution (supplementary) streams based on network conditions without interrupting playback (Compl. ¶80).
Key Claims at a Glance
- The complaint asserts at least independent claim 1 (’303 Patent, col. 18:41-65; Compl. ¶113).
- Claim 1 (A video image coding data receiver) essential elements:
- A processor and a memory unit;
- receiving basic video image coding data;
- decoding the basic data to reproduce a video image;
- receiving supplementary video image coding data, which includes a "supplementary hierarchical picture" whose coding and display order are earlier by a factor of a group of pictures than a "basic hierarchical picture";
- acquiring basic video image coding data received before the supplementary data;
- reconstructing video image coding data from the basic and supplementary data.
- The complaint does not explicitly reserve the right to assert dependent claims for this patent.
Multi-Patent Capsule: U.S. Patent No. 10,218,995 - "Moving Picture Encoding System, ... Moving Picture Reencoding Program," Issued February 26, 2019 (’995 Patent)
Technology Synopsis
The patent describes a hierarchical video decoding system that uses a "super-resolution enlarger" to create higher-resolution pictures from a standard-resolution input. The system employs multiple decoders and resolution converters to manage and combine standard-resolution and super-resolved reference frames, enabling an extension of the standard resolution (U.S. Patent No. 10,218,995, Abstract; Compl. ¶81).
Asserted Claims
At least Claim 2 is asserted (Compl. ¶130).
Accused Features
The complaint alleges that the super-resolution features in the AV1 and/or SVT-AV1 video standards, which involve upscaling and loop restoration operations, infringe the ’995 Patent (Compl. ¶¶131, 133).
Multi-Patent Capsule: U.S. Patent No. 9,042,448 - "Moving Picture Encoding System, ... and Moving Picture Reencoding Program," Issued May 26, 2015 (’448 Patent)
Technology Synopsis
Similar to the ’995 patent, the ’448 patent relates to a hierarchical video encoding system that generates higher-resolution video from standard-resolution content. It describes a system with multiple encoders and super-resolution enlargers that use different sets of reference pictures (some standard, some upscaled) to create a higher-resolution encoded bitstream (U.S. Patent No. 9,042,448, Abstract; Compl. ¶82).
Asserted Claims
At least Claim 1 is asserted (Compl. ¶147).
Accused Features
The complaint accuses the "libaom-av1" encoder, an implementation of the AV1 standard, which allows input video at a higher resolution to be coded at a lower resolution using upscaled or downscaled reference frames (Compl. ¶148).
Multi-Patent Capsule: U.S. Patent No. 8,230,101 - "Server Device for Media, Method for Controlling Server for Media, and Program," Issued July 24, 2012 (’101 Patent)
Technology Synopsis
The patent describes a media server system designed to manage large amounts of digital content by using both internal storage and a separate network storage device. A key feature is a "transfer control unit" that moves some content to network storage but presents a unified list of all content to the user, and a search unit that locates content in either storage location for streaming (U.S. Patent No. 8,230,101, Abstract; Compl. ¶83).
Asserted Claims
At least Claim 1 is asserted (Compl. ¶164).
Accused Features
The complaint alleges that Google's content delivery systems, including YouTube and the Google Cloud Content Delivery Network (CDN), infringe by using a combination of backend cloud storage (internal storage) and CDN caches (network storage) to deliver content to users (Compl. ¶¶163, 165).
Multi-Patent Capsule: U.S. Patent No. 7,804,891 - "Device and Method for Judging Communication Quality and Program Used for the Judgment," Issued September 28, 2010 (’891 Patent)
Technology Synopsis
The patent discloses a method for judging the quality of a communication channel. It involves transmitting data containing a "protected portion" and redundant bits with a predetermined value. The receiving device judges channel quality by identifying the number of redundant bits that have been corrupted during transmission (U.S. Patent No. 7,804,891, Abstract; Compl. ¶84).
Asserted Claims
At least Claim 1 is asserted (Compl. ¶178).
Accused Features
The complaint accuses Google Pixel smartphones that comply with the 5G NR standard. It alleges that the Low-density Parity-check (LPDC) based Cyclic Redundancy Check (CRC) method used in 5G NR for error detection infringes by using protected data and redundant bits (parity/CRC bits) to judge communication quality (Compl. ¶¶177, 179-180).
III. The Accused Instrumentality
Product Identification
The complaint names a broad suite of Google products and services, grouped by the technologies they allegedly implement:
- AV1 Codec Products: Google Pixel 6, 7, and 8 series smartphones, Google Chromebooks, Pixel Slate, Google Chrome, Chromecast, Android/Google TV, YouTube, Google Meet, and Google Duo. These products are accused of infringement to the extent they encode and/or decode digital video using an AV1 codec (Compl. ¶¶86, 93, 112, 129, 146).
- Network Content Delivery Systems: YouTube, Google Cloud Content Delivery Network (CDN), and Google Smart Home are accused of infringing the ’101 Patent on media servers (Compl. ¶163).
- 5G NR Products: Google Pixel smartphone products compliant with 5G NR technology (e.g., Pixel 4a, 5, 6, 7, 8 series) are accused of infringing the ’891 Patent on communication quality (Compl. ¶177).
Functionality and Market Context
- The complaint alleges that the accused AV1 codec products utilize specific video compression techniques such as predictive coding, motion estimation, and block partitioning to efficiently encode and decode video streams (Compl. ¶¶95-98). The complaint provides a block diagram of a typical hybrid video encoder to illustrate this process (Compl. p. 40, FIGURE 1).
- The Google Cloud CDN is described as a system that accelerates web and video content delivery by using a global network to bring content closer to users, caching content from origin servers to reduce latency (Compl. ¶¶90-91, 165).
- The 5G NR-compliant Pixel phones are alleged to use CRC error detection methods to judge the quality of the transmission channel and correct errors, which is fundamental to reliable high-speed mobile data communication (Compl. ¶¶179-181).
IV. Analysis of Infringement Allegations
U.S. Patent No. 8,090,025 Infringement Allegations
| Claim Element (from Independent Claim 10) | Alleged Infringing Functionality | Complaint Citation | Patent Citation |
|---|---|---|---|
| a moving-picture decoding method comprising the steps of: demultiplexing coded data from an input signal based on a specific syntax structure... | The Accused Products demultiplex coded data from an input signal, which is obtained by predictive coding. This is shown in a diagram of a hybrid video encoder. | ¶95; p. 40, FIGURE 1 | col. 37:28-34 |
| performing inverse-quantization to the post-quantization data... performing inverse-orthogonal transform to the post-quantization orthogonal transform coefficients data to produce a decoded residual picture... | The Accused Products perform inverse-quantization and inverse-orthogonal transform to produce a decoded residual picture of a video area. | ¶100 | col. 38:2-7 |
| defining a boundary condition of a border that corresponds to the border motion-vector data... | The Accused Products use motion estimation to find matching pixel values at the border between blocks, generating border motion-vector data when boundary conditions match. | ¶97 | col. 37:55-65 |
| and generate an estimated video signal in each rectangular zone in the picture to be coded, that satisfies Poisson’s Equation, thus producing a first predictive picture; | The Accused Products use smoothing algorithms in Overlapped Block Motion Compensation ("OMBC"), which allegedly satisfies Poisson's Equation to produce a predictive picture from an estimated video signal. | ¶98 | col. 37:66-38:1 |
| combining the first predictive picture and the decoded residual picture to generate a decoded moving-picture signal; | The Accused Products combine the predictive picture and the decoded residual picture to generate the final decoded moving-picture signal. | ¶102 | col. 38:8-10 |
| and storing the decoded moving-picture signal for at least one picture as a reference picture. | The Accused Products store the decoded signal as a reference picture by updating the set of reference frames for use in decoding subsequent pictures. | ¶103 | col. 38:11-13 |
Identified Points of Contention (’025 Patent):
- Technical Question: A central question is whether the "smoothing algorithms in Overlapped Block Motion Compensation ('OMBC')" used in the AV1 standard, as alleged by the complaint, do in fact generate a signal "that satisfies Poisson’s Equation" as required by the claim (Compl. ¶98). The dispute may focus on whether the function performed by OMBC is mathematically equivalent to solving Poisson's equation or merely achieves a similar outcome (smoothing block boundaries).
- Scope Question: The analysis may question whether finding a "border... having a boundary condition that matches" via motion-vector search, as claimed, reads on the block motion estimation algorithms used in AV1 (Compl. ¶97).
U.S. Patent No. 9,986,303 Infringement Allegations
| Claim Element (from Independent Claim 1) | Alleged Infringing Functionality | Complaint Citation | Patent Citation |
|---|---|---|---|
| a video image coding data receiver comprising a processor and a memory unit... | The Accused Products (e.g., Pixel phones) contain processors (e.g., Google Tensor G3) and memory. | ¶114 | col. 18:42-45 |
| receiving basic video image coding data; [and] decoding the received basic video image coding data so as to reproduce a video image; | The Accused Products are configured to receive and decode a basic video stream, such as a 720p resolution video, to reproduce an image. | ¶115 | col. 18:46-48 |
| receiving supplementary video image coding data including a supplementary hierarchical picture whose coding order and display order are earlier by a factor of a group of pictures...than those of a basic hierarchical picture... | The Accused Products receive supplementary data, such as a 1080p stream. The complaint alleges that AV1's use of an "S frame" to switch between frame rates or resolutions provides a picture whose coding/display order is earlier than the basic pictures that are still in the buffer. The complaint provides a diagram of a decoder buffer to illustrate how frames are handled over time. | ¶¶116-117, 119; p. 51, Figure | col. 18:49-59 |
| acquiring basic video image coding data received before supplementary video image coding data that has been received at the moment; | The Accused Products are configured to acquire the basic video data from the buffer, which was received before the supplementary data (e.g., the S frame) that triggered the resolution switch. | ¶120 | col. 18:60-62 |
| and reconstructing video image coding data from the basic video image coding data and the supplementary video image coding data. | The Accused Products reconstruct the final video image by combining the basic and supplementary data, for example, after a resolution switch. | ¶121 | col. 18:63-65 |
Identified Points of Contention (’303 Patent):
- Scope Question: A key issue will be whether the AV1 standard's mechanism for adaptive bitrate streaming, which uses "S frames" (switch frames), maps onto the patent's specific definitions of "basic hierarchical picture" and "supplementary hierarchical picture" (Compl. ¶¶116-117). The court may need to determine if an S frame and the frames in the buffer meet the claim's precise requirement that the supplementary picture's coding and display order "are earlier by a factor of a group of pictures" than the basic picture.
- Technical Question: The analysis raises the question of how the received data is technically "stored in a buffer before decoding" in the accused products and whether this process aligns with the patent's implicit model of acquiring previously-received basic data "at the moment" a switch to supplementary data occurs (Compl. ¶¶119-120).
V. Key Claim Terms for Construction
For the ’025 Patent:
- The Term: "that satisfies Poisson’s Equation"
- Context and Importance: This term is central to infringement, as it defines the mathematical property of the generated "estimated video signal." Practitioners may focus on this term because the plaintiff's case hinges on proving that the accused AV1 smoothing algorithm (OMBC) is not merely similar to, but actually "satisfies," this specific partial differential equation. The defendant may argue that its algorithm is technically different and does not meet this strict mathematical constraint.
- Intrinsic Evidence for Interpretation:
- Evidence for a Broader Interpretation: The specification describes the purpose as producing a predictive picture that "maintaining continuity of a video signal" and avoids a "discontinuous state between blocks" (’025 Patent, col. 2:32-35). An argument could be made that any mathematical model achieving this stated goal falls within the scope.
- Evidence for a Narrower Interpretation: The patent repeatedly and explicitly names "Poisson's Equation" throughout the detailed description and claims (’025 Patent, col. 2:53-54, Abstract). The detailed description provides specific mathematical formulas for generating the signal, which could be used to argue that only systems implementing these or mathematically equivalent formulas can "satisfy" the equation.
For the ’303 Patent:
- The Term: "supplementary hierarchical picture whose coding order and display order are earlier by a factor of a group of pictures ... than those of a basic hierarchical picture"
- Context and Importance: This term defines the specific temporal and structural relationship between the two types of video data. The infringement case depends on whether AV1's S-frames and buffering mechanism for switching resolutions can be construed as creating this exact arrangement. The defendant may argue that AV1's architecture is distinct and does not create a "supplementary hierarchical picture" with this precise "earlier" ordering relative to a "basic hierarchical picture."
- Intrinsic Evidence for Interpretation:
- Evidence for a Broader Interpretation: The specification describes the invention in the context of transmitting both "an overview video image" (basic) and "an extension video image" (supplementary) (’303 Patent, col. 1:44-46). This broader purpose could support an interpretation where any system that combines a base layer with an enhancement layer for adaptive streaming fits the claim.
- Evidence for a Narrower Interpretation: The claim language is highly specific, requiring the supplementary picture to be earlier by "a factor of a group of pictures," with both hierarchies being "set in units of the group of pictures." This suggests a specific, structured relationship that may be argued to be narrower than the more flexible frame-switching mechanisms used in modern codecs like AV1.
VI. Other Allegations
- Indirect Infringement: The complaint alleges inducement of infringement across all asserted patents, stating that Google provides "instructions, documentation, and other information to customers and end-users suggesting that they use the... Accused Products in an infringing manner," citing technical support, marketing, and product manuals (Compl. ¶¶105, 122, 138, 156). Contributory infringement is also alleged, on the basis that accused software components are material to the inventions, are not staple articles of commerce, and are known by Google to be especially made for infringement (Compl. ¶¶106, 123, 139, 157).
- Willful Infringement: The complaint alleges willfulness based on two primary grounds. First, it alleges Google was a "founding member of the Alliance for Open Media, the organization that publishes the AV1 Specification," and that AOM conducted "patent due diligence," which Plaintiff claims either did or should have uncovered the patents-in-suit (Compl. ¶¶107, 124, 140, 158). Second, it alleges actual knowledge from prior litigations against "direct competitors of Defendant" and from citations of the patents or their family members during the prosecution of Google's own patents (Compl. ¶¶87, 88, 108, 125, 142, 159).
VII. Analyst’s Conclusion: Key Questions for the Case
- A core issue will be one of technical satisfaction: does the Overlapped Block Motion Compensation (OBMC) algorithm used in the accused AV1 standard mathematically "satisfy Poisson's Equation" as strictly required by Claim 10 of the ’025 patent, or does it represent a technically distinct method for smoothing block boundaries?
- A key question of definitional scope will be whether the AV1 standard's use of S-frames and reference picture buffering for adaptive resolution switching can be construed to meet the specific structural and temporal ordering requirements of "basic" and "supplementary" hierarchical pictures as laid out in Claim 1 of the ’303 patent.
- A central question for willfulness and damages will be one of knowledge and intent: what level of knowledge or willful blindness can be established from Google's role as a founding member of the Alliance for Open Media and its alleged "patent due diligence" during the development of the accused AV1 standard, particularly in light of alleged notice from prior litigations and patent prosecution?