PTAB

IPR2017-00167

Oracle America Inc v. Realtime Data LLC

Log In
Key Events
Petition
petition Intelligence

1. Case Identification

2. Patent Overview

  • Title: System and Method for Lossless Data Compression and Decompression
  • Brief Description: The ’812 patent discloses a data compression method that combines two well-known techniques: run-length encoding for repeated data blocks and dictionary encoding for non-repeated data blocks into a single compression algorithm.

3. Grounds for Unpatentability

Ground 1: Obviousness over O’Brien and Nelson - Claims 1 and 28 are obvious over O’Brien in view of Nelson.

  • Prior Art Relied Upon: O’Brien (Patent 4,929,946) and Nelson (The Data Compression Book, 1992).
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner argued that O’Brien taught a complete data compression system that combined a run-length encoder with what it termed a "reference value encoder" or "string encoder" to produce a single output stream. This string encoder implemented the signature features of an LZW compression algorithm. The Nelson textbook was cited to establish that a person of ordinary skill in the art (POSA) would have understood that LZW-based string compression, like that in O'Brien, is a form of dictionary-based encoding. O’Brien disclosed detecting runs of three or more bytes (the "detecting" step), outputting an encoded sequence for the run (the "outputting an encoded run-length sequence" step), and building, searching, and outputting codes for non-run data strings using its string encoder. Nelson clarified that O’Brien’s "reference values" and "string table" constituted the "code words" and "dictionary" recited in the claims. For system claim 28, Petitioner contended O’Brien’s run-length control codes were a type of "control code word" used by the run-length encoder.
    • Motivation to Combine: A POSA would combine the teachings because O’Brien itself stated that its "adaptive data compression algorithm" was "well known" and did not disclose it in detail, thereby encouraging a POSA to consult an authoritative source like Nelson. Nelson explicitly described various LZW implementations as "dictionary encoders." Therefore, a POSITA would not be "combining" two systems, but rather using Nelson as a guide to understand and properly characterize the dictionary-based nature of the LZW algorithm already present in O'Brien's integrated system.
    • Expectation of Success: The result would have been predictable because the combination merely involved applying the standard terminology and context from Nelson to the well-understood LZW algorithm disclosed in O’Brien. No technical modification was required.

Ground 2: Obviousness over O’Brien, Nelson, and Welch - Claim 14 is obvious over O’Brien in view of Nelson and the Welch Patent.

  • Prior Art Relied Upon: O’Brien (Patent 4,929,946), Nelson (The Data Compression Book, 1992), and Welch (Patent 4,558,302).
  • Core Argument for this Ground:
    • Prior Art Mapping: This ground challenged claim 14, which recites a "program storage device readable by a machine" containing instructions to perform the method of claim 1. The unpatentability of the underlying method steps was based on the same O'Brien and Nelson combination from Ground 1. The Welch patent, which describes the foundational LZW algorithm, was introduced to show that it was well-known and obvious to implement such compression algorithms in either hardware or software. Welch explicitly described both hardware (Figs. 2-5) and software (Figs. 6-9) implementations of its LZW encoder.
    • Motivation to Combine: Petitioner argued a POSA would have known that the choice between a hardware or software implementation was a routine design choice based on performance and cost requirements. Welch provided specific evidence that this was true for the exact type of LZW compression used in O'Brien's system. Therefore, it would have been obvious to create a software implementation of O’Brien’s method and store it on a program storage device as recited in claim 14.
    • Expectation of Success: Success was expected because Welch demonstrated that LZW algorithms were routinely and successfully implemented in software.

Ground 3: Obviousness over O’Brien, Nelson, and Heath - Claim 9 is obvious over O’Brien in view of Nelson and Heath.

  • Prior Art Relied Upon: O’Brien (Patent 4,929,946), Nelson (The Data Compression Book, 1992), and Heath (Patent 5,973,630).

  • Core Argument for this Ground:

    • Prior Art Mapping: This ground challenged dependent claim 9, which added the limitation of "bit-packing encoded run-length sequences and code words that are output." While O'Brien taught outputting variable-length codes, Heath was introduced for its express teaching of bit-packing. Heath disclosed that code words from a compression algorithm are "shifted and packed into [a] compressed data stream" to improve efficiency.
    • Motivation to Combine: A POSA would have been motivated to incorporate Heath's bit-packing into O'Brien's system to further O'Brien's stated goal of "speeding both transmission and storage of the data." O'Brien's use of variable-length reference values would make bit-packing a particularly desirable and known optimization technique to reduce the average number of bits in the output stream. Both references addressed improving LZW-style dictionary compression by incorporating run-length encoding, making the combination logical and intuitive.
    • Expectation of Success: A POSA would have had a reasonable expectation of success as Heath's packing method could have been easily incorporated into O'Brien's existing byte assembly module.

  • Additional Grounds: Petitioner asserted an additional obviousness challenge against claim 22 (dependent on claim 14) over O'Brien, Nelson, Welch, and Heath, which combined the reasoning for claims 9 and 14.

4. Key Claim Construction Positions

  • "control code word" (claim 28): Petitioner proposed this term be construed as "a dictionary index corresponding to a signal to a decoder." This construction was based on the ’812 patent’s specification, which described control codes (e.g., for dictionary reset, run-length indication) as specific dictionary indices that signal the decoder to perform certain actions.

5. Relief Requested

  • Petitioner requests institution of an IPR and cancellation of claims 1, 9, 14, 22, and 28 of the ’812 patent as unpatentable under 35 U.S.C. §103.