PTAB

IPR2019-00477

Unified Patents Inc. v. Hertl Media, LLC

Log In

1. Case Identification

2. Patent Overview

  • Title: Data Processor with Parallel Queues for Multi-Language Multimedia Streams
  • Brief Description: The ’365 patent describes a system for processing multimedia data streams to reduce latency when a user switches between different language audio tracks. The system uses a set of parallel FIFO (First-In-First-Out) queues, where each queue corresponds to a specific language and stores only references pointing to the language-specific content located in a main input buffer.

3. Grounds for Unpatentability

Ground 1: Claims 1-17 are obvious over Gordon in view of Yamada.

  • Prior Art Relied Upon: Gordon (Application # 2008/0101421) and Yamada (Patent 5,610,914).
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner argued that Gordon disclosed a system for processing multimedia streams (e.g., from a DVD) with parallel FIFO queues to buffer separate language-specific audio streams, enabling rapid switching between them. However, Gordon’s queues stored the actual audio data. Petitioner asserted that Yamada, which addresses the general problem of multi-stream buffering, taught a more efficient architecture. Yamada disclosed using a single, large shared buffer memory to store the actual data packets ("cells") and a plurality of smaller, parallel FIFO queues that store only address pointers referencing the location of the corresponding data in the shared buffer. Petitioner contended that combining Gordon's multi-language application with Yamada's reference-based buffering architecture rendered the key limitations of independent claims 1, 11, and 15 obvious. This included the input buffer (Yamada's shared buffer), the parallel queuing buffer storing only references (Yamada's parallel FIFO pointer queues), and a feeder to extract the references (Yamada's memory control).
    • Motivation to Combine: A POSITA would combine Yamada's teachings with Gordon's system to achieve predictable improvements. The primary motivations were to reduce system cost and improve performance. By replacing Gordon’s multiple large data buffers with Yamada's single large data buffer and multiple small reference queues, a POSITA would reduce memory requirements and implementation costs. This modification was a known technique to improve such a device, as storing smaller references is more efficient than storing large data segments in multiple queues.
    • Expectation of Success: A POSITA would have had a reasonable expectation of success because the combination involved applying a known, general-purpose buffering solution (Yamada) to a specific, analogous application (Gordon's multi-language playback). The implementation would not upset Gordon’s core function of switching between audio streams but would simply perform it more efficiently. Petitioner argued this required only routine engineering, as Yamada’s solution was applicable to any technical application buffering multiple data streams.

Ground 2: Claims 1-17 are obvious over Gordon in view of Brown.

  • Prior Art Relied Upon: Gordon (Application # 2008/0101421) and Brown (Patent 6,868,225).
  • Core Argument for this Ground:
    • Prior Art Mapping: This ground presented an alternative combination to argue the same central point. As in Ground 1, Gordon was cited for its teaching of parallel queues for different language audio streams. Petitioner combined this with Brown, a patent assigned to TiVo for a digital video recorder (DVR). Brown disclosed a system where a received data stream is stored in a single linear buffer on a storage device (the claimed input buffer). To manage playback, Brown used an "Event Buffer" containing "logical segments" that functioned as references. Each logical segment was a pointer containing a timestamp and memory address for a corresponding data segment in the main storage, allowing the system to access and play content without parsing the entire immense data stream. Petitioner argued that incorporating Brown's reference-based indexing into Gordon's parallel queue structure would result in the claimed invention, where Gordon's parallel FIFOs would store Brown's logical segments (references) instead of the actual audio data.
    • Motivation to Combine: A POSITA would be motivated to modify Gordon's system with Brown's reference-based indexing to gain known performance benefits, such as reduced memory and processor bandwidth usage. Since both DVD playback systems (Gordon) and DVRs (Brown) face the challenge of managing large multimedia streams for playback, a POSITA would have looked to solutions in this analogous art. The use of an input buffer for DVD playback was well-known, and combining it with Brown’s efficient indexing method was presented as an obvious path to improving Gordon's design.
    • Expectation of Success: The combination involved integrating known elements from analogous fields to achieve the predictable result of a more efficient system. Modifying Gordon’s parallel queues to store Brown’s logical segments was argued to be a straightforward task for a POSITA, involving routine engineering with a high likelihood of success. The resulting system would perform the same function as Gordon (language switching) in substantially the same way, just with a more efficient, well-understood data management technique.

4. Relief Requested

  • Petitioner requests institution of an inter partes review and cancellation of claims 1-17 of Patent 9,324,365 as unpatentable under 35 U.S.C. §103.