PTAB

IPR2025-00969

Google LLC v. Sandpiper CDN LLC

Key Events
Petition
petition

1. Case Identification

2. Patent Overview

  • Title: Content Delivery System
  • Brief Description: The ’903 patent discloses a content delivery system that uses shared "repeater servers" to replicate and deliver resources from multiple content providers' origin servers to client machines, thereby off-loading request processing from the origin servers. The system uses tables with alias names to manage and direct content requests.

3. Grounds for Unpatentability

Ground 1: Claims 1, 2, 22-24, 26, 28-32, 35, 37, 38, 40, 41, and 43-45 are obvious over Kenner.

  • Prior Art Relied Upon: Kenner (WO 1996/041285).
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner argued that Kenner, which discloses a distributed video clip delivery system, taught every limitation of the challenged claims. Kenner’s network of geographically dispersed "Index Managers" (IMs) that store and retrieve video clips were equated to the ’903 patent’s "shared repeater servers." These IMs replicate resources from a "source IM" (equated to an origin server) and are shared because they deliver content from multiple source IMs associated with different content providers. Petitioner asserted that Kenner’s "clip database," which stores information for each video clip including a unique "video ID," met the "table" and "alias name" limitations. This database is used to identify the source IM (content provider) associated with a requested clip. For dependent claims, Petitioner argued Kenner taught features like embedded objects, HTTP-based resources (via its use of URLs), and selecting servers based on network cost, load, and location.
    • Key Aspects: This ground formed the foundation for all other challenges, establishing Kenner as a base system for content delivery that allegedly included all core features of the ’903 patent, such as replication, shared servers, and the use of a database with unique identifiers to manage content from multiple providers.

Ground 2: Claim 33 is obvious over Kenner in view of Vetter.

  • Prior Art Relied Upon: Kenner (WO 1996/041285) and Vetter (a 1994 journal article titled "Mosaic and the World-Wide Web").
  • Core Argument for this Ground:
    • Prior Art Mapping: This ground focused on claim 33, which requires that the delivered resources comprise a File Transfer Protocol (FTP)-based resource. Petitioner contended that while Kenner taught delivering HTTP-based content, Vetter explicitly disclosed that web browsers like Mosaic could access content using various protocols, including FTP.
    • Motivation to Combine: A POSITA would combine Kenner and Vetter to expand the types of resources available through Kenner’s established content delivery network. Since both references addressed delivering content via a web browser using URLs, incorporating Vetter’s teaching of FTP was presented as a simple, predictable substitution of one known protocol for another within the URL scheme to enhance system functionality.
    • Expectation of Success: The combination was argued to be straightforward because swapping protocol schemes (e.g., 'http://' for 'ftp://') in a URL was a well-known technique.

Ground 3: Claim 34 is obvious over Kenner in view of Rekimoto.

  • Prior Art Relied Upon: Kenner (WO 1996/041285) and Rekimoto (European Patent Pub. No. EP 0753836).
  • Core Argument for this Ground:
    • Prior Art Mapping: This ground addressed claim 34, which recites using an alias name that is a "domain name" and a "name server" that provides Domain Name System (DNS) resolution. Petitioner argued that Rekimoto explicitly taught using a mapping server with DNS resolution to identify the closest server to a user to reduce latency.
    • Motivation to Combine: A POSITA would be motivated to integrate Rekimoto’s DNS capabilities into Kenner’s system to improve performance. Kenner already sought to reduce latency by prioritizing servers based on proximity, and Rekimoto provided a known method (DNS) for achieving this more efficiently. The combination would use Kenner’s "video ID" as the domain name for resolution by Rekimoto’s DNS server.
    • Expectation of Success: Success was expected because DNS was a well-known, standard technology for resolving names to IP addresses to select servers, and its application to Kenner's geographically distributed network was a predictable improvement.

Ground 4: Claims 27, 36, 39, 42, and 46 are obvious over Kenner in view of Boyles.

  • Prior Art Relied Upon: Kenner (WO 1996/041285) and Boyles (Patent 5,511,208).
  • Core Argument for this Ground:
    • Prior Art Mapping: This ground targeted claims requiring the repeater server to "reject the client request" if the alias name is not included in the table. Petitioner asserted that Kenner suggested this functionality by describing a "reason for the denial" being transmitted if database checks fail. Boyles was introduced because it explicitly taught returning a "negative reply" to a requesting node if a target resource is not known or available in its network of cache servers.
    • Motivation to Combine: A POSITA would combine these teachings to improve user experience. While Kenner acknowledged the need to advise a user of a request failure, Boyles provided a more detailed and express implementation of this feature. Adding Boyles's explicit "negative reply" mechanism to Kenner’s system was argued to be an obvious design choice to provide more intuitive feedback.
    • Expectation of Success: The combination was presented as a straightforward implementation of a known error-handling technique into a similar system, as both systems already handled requests by searching local and then remote servers.

4. Arguments Regarding Discretionary Denial

  • Petitioner argued that discretionary denial under §325(d) was improper because the prior art and combinations asserted in the petition were never applied by the Examiner during prosecution. While Kenner was cited in an IDS, there was no evidence it was substantively considered.
  • Petitioner further argued that denial under the Fintiv factors was improper. It contended that the district court trial date was scheduled for September 2026, which would occur after the Board's anticipated Final Written Decision (November 2026), making the timing neutral. Moreover, few resources had been expended in the parallel litigation, and Petitioner stipulated it would not pursue the same grounds in district court if the IPR was instituted, mitigating concerns of duplicative efforts.

5. Relief Requested

  • Petitioner requested institution of an inter partes review and cancellation of claims 1, 2, 22-24, and 26-46 of the ’903 patent as unpatentable.