DCT

2:19-cv-00041

Implicit LLC v. Sandvine Corp

Log In

I. Executive Summary and Procedural Information

  • Parties & Counsel:
    • Plaintiff: Implicit, LLC (Washington limited liability company with principal place of business in Texas)
    • Defendant: Sandvine Corporation (Canadian corporation with principal place of business in Ontario, Canada)
    • Plaintiff’s Counsel: The Davis Firm, PC
  • Case Identification: 2:19-cv-00041, E.D. Tex., 03/19/2019
  • Venue Allegations: Plaintiff alleges venue is proper in the Eastern District of Texas because Defendant conducts business in the state, places accused products into the stream of commerce with the expectation of purchase by consumers in the district, and derives revenue from products provided to individuals in the district.
  • Core Dispute: Plaintiff alleges that Defendant’s network policy control products and platforms infringe nine U.S. patents related to server request management and dynamic data processing in a network environment.
  • Technical Context: The dispute centers on technologies for deep packet inspection (DPI), network traffic management, and policy enforcement, which are critical for network operators to monitor, manage, and monetize data traffic.
  • Key Procedural History: The complaint alleges that Defendant had knowledge of certain patents prior to the suit. Notice of U.S. Patent No. 8,694,683 and the application leading to U.S. Patent No. 9,270,790 was allegedly provided to Procera Networks, Inc., which was later acquired by Sandvine. Notice of U.S. Patent No. 9,591,104 was allegedly provided in a prior litigation, [Implicit, LLC](https://ai-lab.exparte.com/party/implicit-llc) v. Sandvine Corp. No. 2:18-cv-54 (E.D. Tex.). Plaintiff asserts vicarious liability theories against Sandvine for the conduct of Procera based on the 2017 merger of the two companies.

Case Timeline

Date Event
1999-12-29 Earliest Priority Date for ’683, ’790, ’104, ’780, ’839, ’378 Patents
2003-08-06 Earliest Priority Date for ’075 Patent
2011-11-08 U.S. Patent No. 8,056,075 Issues
2014-04-08 U.S. Patent No. 8,694,683 Issues
2014-12-19 Plaintiff allegedly provides written notice of ’683 Patent to Procera Networks
2015-06-01 Plaintiff allegedly provides written notice of application for ’790 Patent to Procera
2016-02-23 U.S. Patent No. 9,270,790 Issues
2016-04-26 U.S. Patent No. 9,325,740 Issues
2017-03-07 U.S. Patent No. 9,591,104 Issues
2017-09-01 Sandvine and Procera become a "combined company" (approx. date)
2018-03-27 Plaintiff serves Sandvine in Sandvine I, providing alleged notice of ’104 Patent
2018-07-17 U.S. Patent No. 10,027,780 Issues
2018-07-24 U.S. Patent No. 10,033,839 Issues
2019-03-05 U.S. Patent No. 10,225,378 Issues
2019-03-19 First Amended Complaint Filed

II. Technology and Patent(s)-in-Suit Analysis

U.S. Patent No. 8,056,075 - Server Request Management

Issued November 8, 2011

The Invention Explained

  • Problem Addressed: The patent describes the inefficiency and security risks of client-based systems for executing downloaded applications ("applets"), where each client machine must have its own local resources to verify and compile code, making enterprise-wide security management difficult (ʼ075 Patent, col. 1:11-2:11).
  • The Patented Solution: The invention proposes a server-based architecture that centralizes the management of applet requests. This "applet server" can receive a request, build the requested applet from local resources (e.g., source code modules and compilers), perform transformations like security verification or optimization on the server, and then deliver the resulting applet to the client computer in a requested format (ʼ075 Patent, Abstract; col. 2:19-48; Fig. 1). This shifts the security and compilation burden from the client to a managed server environment.
  • Technical Importance: This server-side processing model aimed to enhance security and performance in distributed application environments by centralizing control, a departure from client-heavy models prevalent in the early internet era.

Key Claims at a Glance

  • The complaint asserts at least independent claim 1 (Compl. ¶45).
  • Claim 1 requires a method for delivering applets comprising the steps of:
    • Configuring an applet server manager at a server computer to manage requests from client computers.
    • Receiving a request at the applet server manager.
    • Passing the request from the applet server manager to one or more networks.
    • Receiving one or more applets at the applet server manager from the networks.
    • Processing the one or more applets at the server, where processing includes at least one of compressing, optimizing, or verifying the applets.
    • Sending the processed applets from the applet server manager to the client computers.

U.S. Patent No. 8,694,683 - Method and System for Data Demultiplexing

Issued April 8, 2014

The Invention Explained

  • Problem Addressed: The patent addresses the challenge of processing data that can arrive in many different formats (e.g., compressed, encrypted, different network protocols). Statically configuring a system for every possible sequence of data conversion routines is described as highly inefficient, particularly when formats may be unknown until the data is received (ʼ683 Patent, col. 1:24-2:21).
  • The Patented Solution: The invention provides a system for dynamically identifying a sequence of "message handlers" (i.e., conversion routines) to process a data packet. When a packet is received, the system identifies its source and target formats and assembles a "path"—a sequence of processing steps—to perform the necessary conversions. This path and its associated state information are stored and reused for subsequent packets of the same message, effectively "demultiplexing" different data streams to their correct processing paths (ʼ683 Patent, Abstract; col. 2:51-3:11).
  • Technical Importance: This dynamic, path-based approach to data processing provides a flexible framework for handling complex, multi-protocol data streams without requiring pre-configuration for every possible data type.

Key Claims at a Glance

  • The complaint asserts at least independent claim 1 (Compl. ¶63).
  • Claim 1 requires a method comprising:
    • Receiving a packet of a message.
    • Determining a key value for the packet based on one or more headers.
    • Using the key value to determine if a "path" is currently stored for that key.
    • In response to determining no path is stored:
      • Identifying one or more routines for processing the packet, including a TCP routine to convert packets with a TCP format.
      • Creating a path using the identified routines.
    • Processing the packet using the created path.

U.S. Patent No. 9,270,790 - Method and System for Data Demultiplexing

Issued February 23, 2016

  • Technology Synopsis: This patent, related to the ’683 patent, describes a system for dynamically identifying and applying a sequence of processing routines (a "path") to convert data packets from a source format to a target format based on packet headers (ʼ790 Patent, Abstract).
  • Asserted Claims: At least independent claim 1 (Compl. ¶69).
  • Accused Features: The complaint alleges infringement by products that implement flow-based processing and inspect application data on TCP traffic (Compl. ¶¶ 35, 66, 72).

U.S. Patent No. 10,027,780 - Method and System for Data Demultiplexing

Issued July 17, 2018

  • Technology Synopsis: This patent, also from the same family, describes a method for dynamically creating a processing path for data packets by using a key derived from packet headers to identify a sequence of conversion routines, including a TCP routine ('780 Patent, Abstract).
  • Asserted Claims: At least independent claim 1 (Compl. ¶81).
  • Accused Features: The complaint accuses products that implement "flow-based processing" and have the "ability to inspect application data on TCP traffic" (Compl. ¶¶ 35, 66).

U.S. Patent No. 10,033,839 - Method and System for Data Demultiplexing

Issued July 24, 2018

  • Technology Synopsis: This patent continues the technology of dynamically creating and applying a sequence of protocol-handling routines to data packets based on a key value derived from packet headers, thereby demultiplexing different data streams ('839 Patent, Abstract).
  • Asserted Claims: At least independent claim 1 (Compl. ¶87).
  • Accused Features: The accused features are Defendant's products that implement "flow-based processing and the ability to inspect application data on TCP traffic" (Compl. ¶¶ 35, 66).

U.S. Patent No. 10,225,378 - Method and System for Data Demultiplexing

Issued March 5, 2019

  • Technology Synopsis: This patent describes a similar system for receiving a data packet, determining a key value from its headers, and using that key to identify or create a "path" of processing routines, including a TCP routine, to handle the packet ('378 Patent, Abstract).
  • Asserted Claims: At least independent claim 1 (Compl. ¶93).
  • Accused Features: The complaint accuses products that perform "flow-based processing" and inspect TCP traffic (Compl. ¶¶ 35, 66).

The complaint also asserts U.S. Patent Nos. 8,856,779; 9,325,740; and 9,591,104, but does not provide sufficient detail for individual analysis.

III. The Accused Instrumentality

Product Identification

The accused products include Sandvine’s PacketLogic platforms and software, Policy Traffic Switch, TCP accelerator, Traffic Steering Engine, and Maestro Engine, among others (Compl. ¶35).

Functionality and Market Context

The complaint alleges these products are used by network operators to "monitor, manage and monetize their network traffic" (Compl. ¶101). The products are described as implementing "flow-based processing and the ability to inspect application data on TCP traffic," which is consistent with industry-standard deep packet inspection (DPI) and network policy control technology (Compl. ¶66). These platforms allow network providers to identify different types of data traffic (e.g., video streaming, file sharing) and apply specific rules or policies to that traffic. No probative visual evidence provided in complaint.

IV. Analysis of Infringement Allegations

The complaint does not provide claim charts or detailed infringement theories mapping specific product features to claim elements. It makes general allegations that the Accused Products are "covered by at least claim 1" of each asserted patent (Compl. ¶¶ 45, 51, 57, 63, 69, 75, 81, 87, 93). As such, a detailed claim chart summary cannot be constructed. The analysis below identifies potential points of contention based on the general allegations and the patent claims.

’075 Patent Points of Contention

  • Scope Questions: A central question may be whether Defendant's network traffic management system, which operates on data packets in transit, constitutes an "applet server manager" that "deliver[s] one or more applets" as recited in claim 1. The dispute may focus on whether classifying and applying policies to network flows is equivalent to the claimed method of receiving, processing, and sending discrete, executable "applets" to client computers.
  • Technical Questions: The complaint does not specify how the accused products perform the claimed "processing" step, which requires at least one of "compressing," "optimizing," or "verifying" the "applets." The infringement analysis will require evidence of the accused products performing one of these specific functions on what can be defined as an "applet."

’683 Patent Points of Contention

  • Scope Questions: Claim 1 requires "creating a path" of processing routines if one is not already stored for a given key value. A key issue may be whether applying a pre-configured, static policy rule in the accused DPI system constitutes "creating a path" in the dynamic sense described by the patent. The analysis may turn on the definition of "creating" versus "applying."
  • Technical Questions: What evidence does the complaint provide that the accused products dynamically "identify" a sequence of routines and "create a path" for a new data flow, as opposed to matching the flow to an existing, pre-defined policy in a ruleset? The infringement analysis will likely require a detailed examination of how the accused software architecture handles new, unclassified data streams.

V. Key Claim Terms for Construction

Term from the ’075 Patent: "applet"

  • Context and Importance: This term is fundamental to claim 1 of the ’075 Patent. Its construction will be critical because the accused products are network policy systems that manage data packets, not traditional application servers that deliver executable programs. Practitioners may focus on this term because Defendant will likely argue that its system does not handle "applets" in the sense contemplated by the patent.
  • Intrinsic Evidence for Interpretation:
    • Evidence for a Broader Interpretation: The specification suggests an "applet" can be any "form of program instructions, whether in binary, source or intermediate format" and can be a "self contained program, or it can be a code fragment associated with a larger application" (ʼ075 Patent, col. 3:19-24). This could support a broad reading beyond traditional Java applets.
    • Evidence for a Narrower Interpretation: The patent’s background section is heavily rooted in the context of "architecture neutral programming languages" like Java, where programs are "downloaded from a server computer to a client computer to be interpreted and executed locally" (ʼ075 Patent, col. 1:32-38). This context may support a narrower definition tied to executable code intended for a client machine.

Term from the ’683 Patent: "creating a path"

  • Context and Importance: This active step is a core limitation of claim 1 of the '683 patent and its asserted relatives. The infringement case may depend on whether applying a pre-set rule in a DPI engine constitutes "creating a path." Practitioners may focus on this term because the dynamic, on-the-fly nature of "creating" may distinguish the invention from static policy enforcement systems.
  • Intrinsic Evidence for Interpretation:
    • Evidence for a Broader Interpretation: The patent describes "demultiplexing" as identifying a sequence of conversion routines and controlling the processing of a message by that sequence (ʼ683 Patent, col. 3:5-11). This could be interpreted broadly to include selecting and linking pre-existing handlers into a logical chain for a specific data flow.
    • Evidence for a Narrower Interpretation: The detailed description explains that the system "searches for and identifies a sequence of conversion routines" and "stores the identified sequence so that the sequence can be quickly found (without searching) when the next packet... is received" (ʼ683 Patent, col. 2:51-67). This suggests an initial, one-time act of "creating" a new path data structure for a previously unseen type of message, which may be narrower than simply applying an existing rule.

VI. Other Allegations

Indirect Infringement

The complaint alleges induced infringement, stating that Defendant provides "instructional manuals or videos" that instruct users to operate the accused products in an infringing manner (e.g., Compl. ¶47, ¶65). It also alleges contributory infringement, stating that components of the accused products are especially made or adapted to practice the inventions and are not staple articles of commerce (e.g., Compl. ¶48, ¶66).

Willful Infringement

Willfulness is alleged for all asserted patents. For the ’683 and ’790 Patents, willfulness is based on alleged pre-suit knowledge stemming from notice letters sent to Procera Networks in 2014 and 2015, prior to its acquisition by Sandvine (Compl. ¶¶ 38-40, 67, 73). For the ’104 Patent, willfulness is based on alleged knowledge from the prior Sandvine I litigation (Compl. ¶41, ¶79). For the remaining patents, willfulness is based on knowledge acquired from the filing of the original and amended complaints in the present action (e.g., Compl. ¶¶ 49, 55, 61).

Vicarious Liability

The complaint alleges that Sandvine is liable for any infringing conduct of Procera under the doctrines of single business enterprise and alter ego, based on the companies holding themselves out as a "combined company" and "one Sandvine team" after their 2017 merger (Compl. ¶¶ 99-103).

VII. Analyst’s Conclusion: Key Questions for the Case

  1. A core issue will be one of definitional scope: can terms rooted in the context of early 2000s application delivery and data conversion, such as "applet" ('075 patent) and dynamically "creating a path" of conversion routines ('683 patent family), be construed to read on the functionality of a modern, high-performance deep packet inspection and network policy enforcement system?

  2. A key evidentiary question will be one of technical operation: what evidence will Plaintiff present to demonstrate that Sandvine's products perform the specific, dynamic steps claimed in the data demultiplexing patents—such as identifying routines and "creating" a new path data structure for a previously unseen data flow—as opposed to merely applying pre-configured classification rules from an existing policy engine?

  3. A significant legal question will concern damages and willfulness: if infringement is found, to what extent can Plaintiff establish pre-suit knowledge and willful infringement for patents noticed to Procera before its acquisition by Sandvine, and how will the court address the vicarious liability claims central to that theory?