PTAB

IPR2017-00261

Microsoft Corp v. Proxyconn Inc

Key Events
Petition
petition

1. Case Identification

2. Patent Overview

  • Title: System and Method for Data Access
  • Brief Description: The ’717 patent discloses methods for increasing data access efficiency in a network by reducing redundant data transmissions. The system uses a sender and receiver, where the sender transmits a "digital digest" (i.e., a fingerprint like an MD5 hash) of data, and the receiver checks its local cache for data with a matching digest before requesting the full data object.

3. Grounds for Unpatentability

Ground 1: Anticipation of Claim 16 under §102 by DRP

  • Prior Art Relied Upon: DRP (The HTTP Distribution and Replication Protocol, W3C Note, Aug. 1997).
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner argued that claim 16 depends from claim 12, which was previously found by the Board to be anticipated by DRP in a prior inter partes review (IPR). Petitioner asserted that issue preclusion prevents Patent Owner from re-litigating this finding. The sole new limitation in claim 16 is the requirement that the "digital digest creation" comprises the step of using the MD5 algorithm. DRP was argued to explicitly disclose using MD5 as its "checksum algorithm" for creating "content identifiers," thus teaching the final element of claim 16.

Ground 2: Obviousness of Claim 15 over DRP in view of Williams

  • Prior Art Relied Upon: DRP, and Williams (WO 96/25801).
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner contended that claim 15 depends from claim 12, which DRP anticipates. The only feature added by claim 15 is the use of a Cyclic Redundancy Check (CRC) for creating the digital digest. DRP taught using a "checksum algorithm," with MD5 being a preferred example, but also suggested any "well known checksum algorithm" would be acceptable.
    • Motivation to Combine: A POSITA would combine DRP with Williams because Williams explicitly taught CRC and MD5 as alternative "hash functions" for the same purpose of reducing redundant data transfer. Williams explained the tradeoff between the algorithms, noting that CRC could be faster in applications where cryptographic strength was not a primary concern. This provided a clear motivation to substitute the well-known CRC algorithm into the DRP framework to optimize for speed.
    • Expectation of Success: Because CRC was a well-known and standard checksum algorithm, a POSITA would have had a reasonable expectation of success in using it within the DRP system.

Ground 3: Obviousness of Claims 17, 19, and 20 over DRP in view of Mattis

  • Prior Art Relied Upon: DRP, and Mattis (Patent 6,292,880).
  • Core Argument for this Ground:
    • Prior Art Mapping: These claims relate to an "overlapping data" scenario where digests for a "principal" (new) version and "auxiliary" (old) version of data are used. Petitioner argued DRP disclosed the core method, including sending digests for different file versions and using a "Differential GET" request to retrieve only the changes. To the extent DRP lacked specific implementation details, Mattis was argued to supply them. Mattis disclosed a detailed web proxy cache system that stores and searches for objects using their MD5 digests, including obtaining digests from "permanent storage memory" (claim 20).
    • Motivation to Combine: A POSITA implementing DRP's protocol would be motivated to look to a reference like Mattis for established methods of building a high-performance, content-indexed cache. Both references solved the same problem (data redundancy) with the same tool (MD5 digests) in the same environment (HTTP networks), making their combination a natural and logical design choice.
    • Expectation of Success: Combining a high-level protocol (DRP) with a detailed implementation guide for a critical component like a cache (Mattis) was presented as a predictable integration with a high expectation of success.

Ground 4: Obviousness of Claims 21 and 25-27 over DRP in view of Mattis and Williams

  • Prior Art Relied Upon: DRP, Mattis, and Williams.
  • Core Argument for this Ground:
    • Prior Art Mapping: This ground extended the DRP and Mattis combination to cover additional claims. Claim 21 added the use of CRC to the method of claim 17. Claims 25-27 recited the same method from the receiver's perspective, including receiving a message with principal and auxiliary digests and searching a local "network cache memory" for matches. The combination of DRP and Mattis provided the base system, while Williams provided the teaching for using CRC and for sending digests for overlapping data in the same message.
    • Motivation to Combine: The motivation was cumulative of the prior grounds. A POSITA would combine DRP (protocol) and Mattis (cache implementation) for efficiency. That same POSITA would be further motivated by Williams to substitute CRC for MD5 to improve performance where security was not critical, and to adopt Williams's hierarchical approach to handling overlapping data.
    • Expectation of Success: Petitioner asserted this was a predictable combination of known elements from the prior art to yield the predictable result of an efficient data access system.

4. Key Claim Construction Positions

  • Petitioner stated its arguments rely on claim constructions previously adopted by the Board and the Federal Circuit in prior proceedings involving the ’717 patent. Key constructions included:
    • "sender/computer" and "receiver/computer": Construed broadly to mean computers that send and receive data, respectively, and can encompass intermediaries.
    • "searching for data with the same digital digest": Interpreted to require the ability to identify a specific data object from a set of potentially many objects stored in the entire network cache memory, not just on a single data object.

5. Relief Requested

  • Petitioner requested the institution of an inter partes review and the cancellation of claims 15-17, 19-21, and 25-27 of the ’717 patent as unpatentable on all asserted grounds.