Apple Inc v. California Institute Of Technology

1. Case Identification

• Case #: IPR2017-00297
• Patent #: 7,916,781
• Filed: December 12, 2016
• Petitioner(s): Apple Inc.
• Patent Owner(s): California Institute of Technology
• Challenged Claims: 3-12, 19-21

2. Patent Overview

• Title: Serial Concatenation of Interleaved Convolutional Codes Forming Turbo-Like Codes
• Brief Description: The ’781 patent describes methods for encoding signals using error-correcting codes, specifically focusing on a class of codes known as irregular repeat-accumulate (IRA) codes. The core invention involves a two-step encoding process where an outer coder first performs a linear transform operation (e.g., repeating information bits) and an inner coder then performs an accumulation operation on the transformed bits to generate parity bits.

3. Grounds for Unpatentability

Ground 1: Obviousness over Ping in view of Divsalar - Claims 3, 5-8, 10, and 12 are obvious over Ping in view of Divsalar.

• Prior Art Relied Upon: Ping (a 1999 journal article on low-density parity check codes) and Divsalar (a 1998 conference paper on "turbo-like" codes).
• Core Argument for this Ground:
  • Prior Art Mapping: Petitioner argued that Ping teaches a two-stage encoding process that meets the limitations of the challenged claims. Ping’s first stage involves computing summations of information bits using a generator matrix, which constitutes the claimed "linear transform operation." The second stage involves accumulating these summations to generate the final parity bits, which constitutes the claimed "accumulation operation." Although Ping teaches this process, Divsalar provides a clear and explicit framework for a two-step "repeat and accumulate" (RA) encoder, reinforcing the structure taught in Ping.
  • Motivation to Combine: A person of ordinary skill in the art (POSITA) would combine Ping and Divsalar because both references are directed to the same field of improving error-correcting codes by reducing encoding complexity. Divsalar’s explicit teaching of a two-step encoder with distinct blocks for repetition and accumulation would have provided a clear and advantageous implementation for Ping’s more mathematically described method. Using Divsalar’s structure would simplify the encoder’s implementation and reduce memory requirements, a goal explicitly stated in Ping.
  • Expectation of Success: A POSITA would have a reasonable expectation of success because both references use similar underlying principles, such as mod-2 summing for accumulation, and operate in the same technical domain. The combination represents a straightforward application of a known encoding architecture (from Divsalar) to a specific encoding algorithm (from Ping).

Ground 2: Anticipation by Ping - Claims 19-21 are anticipated by Ping.

• Prior Art Relied Upon: Ping (a 1999 journal article on low-density parity check codes).
• Core Argument for this Ground:
  • Prior Art Mapping: Petitioner asserted that Ping, standing alone, discloses every element of independent claim 19 and its dependent claims 20 and 21. Ping was argued to teach a method of encoding that includes an "accumulation of mod-2 or exclusive-OR sums of bits in subsets of the information bits." This is shown in Ping’s Equation (4), which defines how parity bits are recursively calculated from summations of information bits. Petitioner further argued that Ping’s disclosed matrix structure, which has a column weight of ‘t=4’, inherently teaches the limitation that "at least two of the information bits appear in three subsets of the information bits," as every information bit appears in exactly four subsets.

Ground 3: Obviousness over Ping, Divsalar, and MacKay - Claim 9 is obvious over Ping in view of Divsalar, and further in view of MacKay.

• Prior Art Relied Upon: Ping, Divsalar, and MacKay (a 1999 journal article comparing constructions of irregular codes).

• Core Argument for this Ground:

  • Prior Art Mapping: This ground builds upon Ground 1 by adding MacKay to address the specific limitation of claim 9, which requires an irregular code where "the information bits appear in a variable number of subsets." The combination of Ping and Divsalar teaches a regular code structure, where each information bit contributes to the same number of parity bits. MacKay explicitly teaches that the performance of the exact type of codes discussed in Ping (Gallager/LDPC codes) is significantly improved by making them irregular. MacKay achieves this by using a "nonuniform weight per column" in the parity-check matrix, which directly corresponds to information bits appearing in a variable number of subsets.
  • Motivation to Combine: A POSITA would be motivated to apply MacKay’s teachings to the Ping/Divsalar combination to achieve the known and desirable benefit of superior performance. Ping and MacKay both discuss improving the same class of codes and use identical terminology (e.g., "column weight"). The modification would have been straightforward: simply altering the generator matrix in Ping to have non-uniform column weights as explicitly taught by MacKay.
  • Expectation of Success: Success was expected because MacKay provides a clear roadmap for converting a regular LDPC code into a higher-performing irregular one without changing the fundamental encoding architecture.

• Additional Grounds: Petitioner asserted an additional obviousness challenge for claims 4 and 11 based on Ping, Divsalar, and Coombes (Patent 4,271,520). This ground relied on Coombes to teach the specific limitation of outputting parity bits after the information bits.

4. Key Claim Construction Positions

• "codeword": Petitioner proposed this term be construed to mean "a discrete encoded sequence of data elements," consistent with its common meaning to a POSITA and a construction adopted in related district court litigation.
• "linear transformation": Petitioner proposed this term be construed as "one that obeys the laws of linear algebra including distributive and associative properties." This construction was argued to be consistent with the understanding of a POSITA and was adopted by the Board in a prior IPR proceeding involving the ’781 patent.

5. Relief Requested

• Petitioner requests institution of an inter partes review and cancellation of claims 3-12 and 19-21 of the ’781 patent as unpatentable under 35 U.S.C. §§ 102 and 103.