PTAB

IPR2019-01292

Nokia Of America Corp v. Packet Intelligence LLC

Log In
Key Events
Petition
petition

1. Case Identification

2. Patent Overview

  • Title: Packet Monitoring for Conversational Flows
  • Brief Description: The ’646 patent relates to a packet monitor for examining packets passing through a connection point on a computer network to determine whether a packet belongs to a previously encountered "conversational flow," which may involve multiple, separate network connections related to a single user activity.

3. Grounds for Unpatentability

Ground 1: Obviousness over Riddle, Wakeman, Bruins, and Cheriton - Claims 1-3, 7, 16, and 18 are obvious over Riddle in view of Wakeman, Bruins, and Cheriton.

  • Prior Art Relied Upon: Riddle (Patent 6,412,000), Wakeman (Patent 5,740,175), Bruins (Patent 6,308,148), and Cheriton (Patent 6,091,725).
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner asserted that Riddle, the primary reference, taught a system for classifying network packet flows using information from multiple protocol layers. Riddle allegedly disclosed all major elements of the challenged claims, including a packet acquisition device, memory for storing flow data in "saved lists," and a processor for parsing packets and performing state operations. Critically, Petitioner argued that Riddle’s concept of "service aggregates"—which links disjointed connections for a single application like FTP (e.g., separate control and data channels)—satisfies the key "conversational flow" limitation of the ’646 patent. For claim limitations not explicitly in Riddle, Petitioner contended that Wakeman taught the claimed "cache subsystem" using a content-addressable memory (CAM) cache, and Cheriton taught the "input buffer memory" required by claim 7. Bruins was cited as teaching a formal database for storing flow data, which Petitioner argued would be an obvious alternative to Riddle's "saved lists."
    • Motivation to Combine: A person of ordinary skill in the art (POSITA) would combine Riddle with Wakeman's CAM-cache to improve the speed of flow lookups, a well-known benefit of caching. Similarly, a POSITA implementing Riddle's system in a router would have been motivated to include Cheriton's input buffer memory to prevent packet loss, a routine design choice for improving network performance.
    • Expectation of Success: Petitioner argued a POSITA would have had a reasonable expectation of success, as the combination involved applying conventional network components (caches, buffers) to a traffic classification system to achieve their predictable functions of increasing speed and reliability.

Ground 2: Obviousness over Riddle et al. and RFC 1945 - Claims 1-3, 7, 16, and 18 are obvious over Riddle, Wakeman, Bruins, and Cheriton, and further in view of RFC 1945.

  • Prior Art Relied Upon: Riddle (Patent 6,412,000), Wakeman (Patent 5,740,175), Bruins (Patent 6,308,148), Cheriton (Patent 6,091,725), and RFC 1945 (Hypertext Transfer Protocol -- HTTP/1.0).
  • Core Argument for this Ground:
    • Prior Art Mapping: This ground presented an alternative basis for finding the "conversational flow" limitation obvious. Petitioner argued that the base combination of Riddle, Wakeman, Bruins, and Cheriton provided the core packet monitoring system. RFC 1945, the specification for HTTP/1.0, was then added to teach another method of linking disjointed flows. Petitioner asserted that RFC 1945 disclosed the HTTP "Referer" header field, which indicates the source URL that led to a given request. Petitioner pointed to testimony from the Patent Owner's own expert in related litigation, who allegedly admitted that using the "Referer" field to correlate separate connection flows constitutes a "conversational flow."
    • Motivation to Combine: A POSITA would combine Riddle with RFC 1945 because Riddle explicitly taught classifying web traffic using HTTP header types (e.g., "Content-Type" and "User-Agent"). A POSITA seeking to enhance Riddle's web traffic classification capabilities would naturally consult the governing protocol standard, RFC 1945, and recognize that other header fields like "Referer" could be used to link related but separate HTTP flows, thereby improving classification accuracy.
    • Expectation of Success: Because Riddle already provided a framework for parsing HTTP headers to classify traffic, a POSITA would expect that extending this functionality to parse another well-defined header field from the same protocol would be straightforward and successful.

4. Key Claim Construction Positions

  • "Conversational Flow[s]": This term, present in all independent claims, was identified as the central construction issue. Petitioner noted that the Patent Owner previously agreed to a construction where a "conversational flow" involves a sequence of packets from an activity that may span more than one network connection (i.e., linking disjointed flows). Petitioner argued that its asserted prior art rendered the claims obvious under both the Patent Owner's construction and its own slightly modified proposal, as the prior art taught linking separate connections for a single application-level activity.

5. Key Technical Contentions (Beyond Claim Construction)

  • Satisfying the "Conversational Flow" Limitation: The petition's central technical argument was that, unlike a previously failed IPR petition by another party, its newly presented prior art successfully taught the concept of a "conversational flow." Petitioner dedicated significant argument to establishing three distinct technical bases for this limitation in the prior art:
    • Riddle’s "service aggregates" for linking separate control and data channels in FTP.
    • Riddle’s ability to classify all traffic related to the PointCast application as a single flow.
    • The combination of Riddle and RFC 1945, which taught using the HTTP "Referer" field to link related web requests.

6. Arguments Regarding Discretionary Denial

  • Petitioner argued against discretionary denial under §314(a) and §325(d). It contended that this petition was not cumulative to a prior IPR filed by Sandvine because it was based on entirely new prior art (Riddle) and arguments that directly addressed the deficiencies the Board identified in the Sandvine petition (i.e., the failure to show a "conversational flow"). Petitioner also argued that it had no involvement in the prior IPR and that parallel district court litigation would not resolve the invalidity issues on the same timeline or with the same prior art references.

7. Relief Requested

  • Petitioner requested the institution of an inter partes review and the cancellation of claims 1-3, 7, 16, and 18 of the ’646 patent as unpatentable.