PTAB

IPR2024-00895

Juniper Networks Inc v. Orckit Corp

Key Events
Petition
petition

1. Case Identification

2. Patent Overview

  • Title: Deep Packet Inspection in a Software-Defined Network
  • Brief Description: The ’111 patent discloses a method for deep packet inspection (DPI) within a Software-Defined Network (SDN). The method involves a network node receiving instructions and a criterion from an external controller, checking if a received packet satisfies the criterion, and if so, sending the packet to a different entity specified in the instruction for inspection.

3. Grounds for Unpatentability

Ground 1: Obviousness over Lefebvre and Chua - Claims 32-36 and 39-54 are obvious over Lefebvre in view of Chua.

  • Prior Art Relied Upon: Lefebvre (Patent 10,097,452) and Chua (Patent 9,264,301).
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner argued that Lefebvre discloses a centralized controller in an SDN that steers network traffic through various "inline services" (such as DPI) based on predefined policies. Lefebvre’s controller instructs network switches on how to route packets through a "service chain." However, Lefebvre’s services are dedicated appliances separate from the controller. Chua was argued to cure this deficiency by disclosing an SDN controller that can itself host and perform Layer 4-7 services, including DPI, eliminating the need for separate, expensive hardware. The combination, therefore, taught a network node receiving instructions from an external controller (Lefebvre) that is configured to perform DPI (Chua), checking a packet against a criterion (Lefebvre's classification policies), and, if the criterion is met, sending the packet to an entity other than its original destination (i.e., the controller itself for inspection, as taught by Chua).
    • Motivation to Combine: A POSITA would combine Chua's teaching of an integrated DPI controller with Lefebvre's traffic-steering framework to reduce the cost, complexity, and potential points of failure associated with the separate, dedicated service appliances described in Lefebvre. Both references operate in the same field of SDN, use OpenFlow protocols, and address the common goal of efficiently managing network traffic.
    • Expectation of Success: Petitioner asserted a high expectation of success because Lefebvre's controller is already a programmable, software-based component. Integrating DPI functionality, as explicitly taught by Chua (which also discloses a software development kit for creating controller applications), would be a straightforward software modification for a POSITA.

Ground 2: Obviousness over Lefebvre, Chua, and Rash - Claims 37 and 38 are obvious over Lefebvre in view of Chua and Rash.

  • Prior Art Relied Upon: Lefebvre (Patent 10,097,452), Chua (Patent 9,264,301), and Rash (Patent 9,813,447).
  • Core Argument for this Ground:
    • Prior Art Mapping: This ground builds on Ground 1 to address the limitations of claims 37 and 38, which require sending only a portion of the packet to the controller. The combination of Lefebvre and Chua teaches sending the entire packet for inspection. Rash was argued to explicitly teach a system that mirrors only a portion of a packet (e.g., "the first L bytes") to an "application identification engine." This targeted mirroring minimizes the amount of data transmitted for analysis, thereby reducing network load and improving efficiency.
    • Motivation to Combine: A POSITA implementing the system of Lefebvre and Chua would be motivated to incorporate Rash's teachings as a logical optimization. Since the goal is packet inspection, sending only the necessary portion of the packet to the controller (as taught by Rash) instead of the entire packet would reduce network congestion and the processing burden on the controller, which directly aligns with the general goal of improving network performance.
    • Expectation of Success: Petitioner contended a POSITA would have a high expectation of success, as Rash presents the mirroring of packet portions as a known technique for network analysis. Modifying the redirect/mirror function of the Lefebvre/Chua combination to handle only a portion of the packet was argued to be a routine implementation detail within the skill of a POSITA.

4. Key Claim Construction Positions

  • "controller": Petitioner proposed this term should be construed as "an entity configured to perform deep packet inspection on packets." This construction was argued to be supported by the patent's specification, which repeatedly describes the central controller as the entity performing DPI. This construction is critical for mapping Chua's teachings, where the controller itself performs DPI, onto the claims.
  • "criterion": Petitioner proposed this term be construed as "the conditions, or basis, used to identify packets for actions, such as inspection." As the term does not appear in the specification, this construction was based on its plain and ordinary meaning in the context of the claims. This interpretation is broad enough to encompass the various subscriber-based, application-based, and flow-based traffic classification policies disclosed in Lefebvre.

5. Arguments Regarding Discretionary Denial

  • Petitioner argued that discretionary denial under §325(d) is unwarranted because the specific combination of Lefebvre, Chua, and Rash, along with the accompanying expert analysis, presents art and arguments that are materially different from those considered by the examiner during prosecution.
  • Petitioner further argued that discretionary denial under §314(a) and the Fintiv factors is inappropriate. The parallel district court litigation was stayed less than four months after the complaint was filed, resulting in minimal investment by the parties and the court. Furthermore, Petitioner stipulated that it would not pursue in the district court any grounds raised or that reasonably could have been raised in the IPR petition, a factor that weighs heavily against discretionary denial.

6. Relief Requested

  • Petitioner requests institution of an inter partes review and cancellation of claims 32-54 of Patent 10,652,111 as unpatentable.