PTAB

IPR2017-01544

LG Electronics Inc v. Broadcom Corp

Key Events

Petition

petition

Title: System and Method for Enhancing Performance of Personal Video Recording (PVR) Functions on HITS Digital Video Streams
Brief Description: The ’967 patent discloses a method for improving the performance of trick play functions, such as rewind, on Headend In The Sky (HITS) digital video streams. The method reduces processing load by selectively decoding only the portions of a video picture necessary to build a reference picture, specifically omitting the decoding of slices below a picture’s last intracoded slice.

Prior Art Relied Upon: MacInnis (Application # 2002/0061183).
Core Argument for this Ground:
- Prior Art Mapping: Petitioner argued that MacInnis discloses every limitation of claims 1-5. MacInnis describes a system for enabling trick modes (e.g., rewind) on progressively refreshed video streams that lack full I-pictures, which is the same context as the ’967 patent. Crucially, to save processing time and increase decoding speed, MacInnis teaches that "not all slices [in every picture] must be decoded" and that "only the necessary slices are decoded." Petitioner asserted this teaching of partial decoding for efficiency directly corresponds to the core limitation of claim 1: decoding a portion of a picture "without having decoding the at least one slice below the intracoded slice." MacInnis explains this is achieved by decoding from the top of a picture only down to the end of the necessary I-slice, thus inherently omitting slices below it. The dependent claims were also argued to be disclosed, including predicting slices from a prior picture (claim 3), top-to-bottom decoding in raster order (claim 4), and operating on a P-picture from a HITS stream (claim 5).

Prior Art Relied Upon: Chen ’677 (Application # 2001/0026677).
Core Argument for this Ground:
- Prior Art Mapping: Petitioner contended that Chen ’677, which describes transcoding progressive I-slice MPEG streams to enable trick play, discloses the limitations of claims 1-4. Chen ’677 explicitly teaches that "P-frames need only be partially decoded in order to recover the I-slices." Petitioner argued this partial decoding to extract necessary data for subsequent frame prediction is the same as the ’967 patent’s claimed invention. To predict a slice in a subsequent P-frame (e.g., P7), Chen ’677 decodes the necessary reference slices (P-slice and I-slice) from a prior P-frame (e.g., P4). This process inherently omits decoding unnecessary slices below the reference I-slice in P4, thus meeting the key limitation of claim 1. Petitioner mapped other elements, such as prediction between frames and raster-order decoding, to Chen ’677's disclosure of slice-by-slice processing in a top-to-bottom manner across frames.
- Motivation to Combine (for §103 grounds): For claim 5, Petitioner argued that HITS streams are a well-known type of the progressive I-slice streams described in Chen ’677. A POSITA would have found it obvious to apply the decoding efficiency techniques of Chen ’677 to a HITS stream as a straightforward substitution to achieve predictable results.

Prior Art Relied Upon: MacInnis (Application # 2002/0061183) and ISO 13818-2 (the MPEG-2 video standard).
Core Argument for this Ground:
- Prior Art Mapping: This ground was presented as an alternative to the anticipation argument for claim 4. Petitioner argued that while MacInnis teaches a top-to-bottom decoding order that constitutes raster order, if the Board were to find this disclosure insufficient, the limitation is rendered obvious by the MPEG-2 standard itself (ISO 13818-2).
- Motivation to Combine: MacInnis explicitly states that its system is built upon and its terms (like "rows and slices") are defined by the MPEG-2 standard. A POSITA implementing the system taught in MacInnis would have been directly motivated to consult ISO 13818-2. That standard explicitly directs that slices within a picture "shall occur in the bitstream in the order in which they are encountered, starting at the upper-left of the picture and proceeding by raster-scan order."
- Expectation of Success: Combining the teachings would be straightforward and yield predictable results, as raster-scan is the standard, most direct method for decoding MPEG data.
Additional Grounds: Petitioner asserted an alternative obviousness challenge for claims 1-5 based on Chen ’677 in view of ISO 13818-2, arguing that if Chen ’677 was found not to explicitly teach "transport packets" or "raster order," the MPEG-2 standard would have provided these conventional elements to a POSITA.

"instructions" (claim 1): Petitioner argued this term means "commands, separate from video or audio data, that specify trick play functions to be performed." This construction was based on the specification’s differentiation between data packets and embedded command packets, distinguishing it from the video syntax itself.
"without having decoding the at least one slice below the intracoded slice" (claim 1): Petitioner proposed this means "omitting decoding the at least one slice below the intracoded slice." This was based on the patent’s own description that omitting the decoding of this portion reduces processing requirements.

Priority Date Entitlement: A central contention of the petition was that the challenged claims of the ’967 patent were not entitled to the filing date of their parent application. Petitioner argued the parent application lacked any disclosure of the key limitation "decoding at least a portion of another picture... without having decoding the at least one slice below the intracoded slice." Consequently, the claims were only entitled to the filing date of the continuation-in-part application (December 11, 2002), which made both MacInnis and Chen ’677 available as prior art under at least §102.

Petitioner requested the institution of an inter partes review and the cancellation of claims 1-5 of the ’967 patent as unpatentable under 35 U.S.C. §§ 102 and 103.