Microsoft Corp v. Parallel Networks Licensing LLC

1. Case Identification

• Case #: IPR2015-00484
• Patent #: 5,894,554
• Filed: December 23, 2014
• Petitioner(s): Microsoft Corporation
• Patent Owner(s): Parallel Networks Licensing, LLC
• Challenged Claims: 12, 20-31, 33, 35-45, 47, and 49

2. Patent Overview

• Title: System For Managing Dynamic Web Page Generation Requests By Intercepting Request At Web Server And Routing To Page Server Thereby Releasing Web Server To Process Other Requests
• Brief Description: The ’554 patent relates to a system and method for managing web page requests to improve server performance. The invention describes intercepting a request at a primary web server and routing it to one of a plurality of "page servers" for processing, thereby releasing the primary web server to handle other incoming client requests.

3. Grounds for Unpatentability

Ground 1: Anticipation over SWEB 95 - Claims 12, 20-31, and 35-45 are anticipated by SWEB 95.

• Prior Art Relied Upon: SWEB 95 (Andresen et al., "SWEB: Towards A Scalable World Wide Web Server On Multicomputers," September 1995).
• Core Argument for this Ground:
  • Prior Art Mapping: Petitioner argued that SWEB 95, which describes a scalable web server architecture built on a cluster of workstations, discloses every limitation of the challenged independent claims. The SWEB 95 system receives an HTTP request at one of its servers (the claimed "Web server"), which then uses a distributed scheduler to determine the optimal server in the cluster to handle the request. If another server is selected (the claimed "page server"), the request is redirected to it. This act of receiving, analyzing, and redirecting the request meets the "intercepting" and "routing... to a dispatcher" limitations. The redirection frees the initial server's resources, meeting the "releasing said Web server" limitation. SWEB 95's scheduler examines dynamic system load information to select the optimal server, satisfying the dispatching limitations of claim 12. The reference’s disclosure of using Common Gateway Interface (CGI) programs to handle requests also meets the "dynamically generating a Web page" limitation.

Ground 2: Obviousness over SWEB 95 in view of Leaf - Claims 33 and 47 are obvious over SWEB 95 in view of Leaf.

• Prior Art Relied Upon: SWEB 95; Leaf (Patent 5,754,772).
• Core Argument for this Ground:
  • Prior Art Mapping: This ground addressed limitations in dependent claims 33 and 47 requiring the generated web page to be received back at the initial web server before being provided to the request source. SWEB 95 teaches redirecting the client to the selected page server, meaning the initial server is not in the return path. Leaf, however, explicitly discloses a hierarchical architecture where a front-end "Web Server" receives a request, routes it to a back-end "Transaction Processing System" that generates the dynamic page, and then receives the completed HTML document back from the back-end system to send to the client.
  • Motivation to Combine: Petitioner argued that a person of ordinary skill in the art (POSITA) would have been motivated to combine the teachings of Leaf with SWEB 95 to solve a known problem. The client-side redirection method in SWEB 95 creates overhead from an additional connect/request/respond cycle. A POSITA would recognize that implementing Leaf's front-end gateway architecture, where the initial server manages the entire transaction, would eliminate this overhead and create a more efficient, centralized system. Both references address the common problem of scaling web server performance.
  • Expectation of Success: The combination involved applying a known server architecture (Leaf) to a distributed system (SWEB 95) to achieve a predictable improvement in performance, ensuring a high expectation of success.

Ground 3: Obviousness over SWEB 95 in view of Bradley - Claims 29, 43, and 49 are obvious over SWEB 95 in view of Bradley.

• Prior Art Relied Upon: SWEB 95; Bradley (Bradley et al., "Web-based Access to an Online Atlas of Anatomy...", 1995).
• Core Argument for this Ground:
  • Prior Art Mapping: This ground targeted claim limitations requiring the server selection process to consider whether a requested web page is cached at a page server. While SWEB 95's scheduler considers factors like CPU load and file location, it does not explicitly mention cache status. Bradley, however, teaches a system that uses CGI programs to dynamically convert image files and then caches the generated output files to improve performance. When a request is received, Bradley's server first checks the cache to see if the dynamically generated file already exists, avoiding a costly re-conversion.
  • Motivation to Combine: A POSITA would combine Bradley's caching technique with the SWEB 95 system for the well-understood goal of improving performance. SWEB 95's scheduler already estimates the computational cost of fulfilling a request; incorporating cache status into this estimation is a logical and obvious extension. Caching the output of CGI operations, as taught by Bradley, would directly reduce the load on the SWEB 95 servers and was a known method for accelerating server response.
  • Expectation of Success: Applying the well-known technique of caching to reduce computational load in a distributed server environment like SWEB 95 would have been a straightforward modification with a predictable, positive outcome.
• Additional Grounds: Petitioner asserted additional obviousness challenges against numerous claims based on SWEB 95 alone, arguing that to the extent any element was not explicitly disclosed, it represented a simple application of a known technique for its intended purpose to achieve a predictable result in the context of load balancing.

4. Key Claim Construction Positions

• "intercepting said request at said Web server": Petitioner argued for the plain meaning of "to stop or interrupt," contending the term covers the process where a Web server receives a request and interrupts its own full processing of that request to route it elsewhere. This construction was central to mapping SWEB 95, where an initial server stops its own processing to invoke a scheduler that may redirect the request.
• "releasing said Web server to process other requests": Petitioner proposed that this phrase means freeing any of the Web server's hardware or software resources (e.g., CPU cycles, memory) so they become available for other tasks. This broad construction is met when a task is offloaded to another machine, as the initial server is no longer burdened by it, which is the fundamental purpose of the load-balancing system in SWEB 95.

5. Relief Requested

• Petitioner requests institution of an inter partes review and cancellation of claims 12, 20-31, 33, 35-45, 47, and 49 of Patent 5,894,554 as unpatentable.