Prior art

An analysis of the prior art cited by U.S. Patent 11,991,234 is detailed below. The analysis focuses on references that could be considered relevant to the key inventive concepts of segmenting media content into "streamlets," creating multiple versions of these streamlets at different bitrates, and distributing the encoding process.

Analysis of Cited Prior Art for US 11,991,234

Based on the patent's description, the core invention lies in the server-side method of preparing media for adaptive streaming. This involves:

1. Segmentation: Dividing the content into small, time-indexed chunks ("streamlets").
2. Multi-bitrate Encoding: Creating a set of these streamlets for each time index, with each streamlet in the set having a unique bitrate.
3. Distributed Encoding (Master/Host): A system for assigning these encoding tasks to multiple computers based on a bidding process to optimize for speed, enabling near-real-time encoding of live events.

The following prior art was cited during the prosecution of this patent and its parent applications. This analysis evaluates their relevance to the claims as understood from the provided patent text.

Key Prior Art References and Potential Anticipation

1. U.S. Patent No. 7,032,042 B1 (filed Apr. 2, 2001) - "Method and system for distributing streaming media presentations from a network of servers" (Assignee: iBEAM Broadcasting Corp.)

• Publication/Filing Date: Filed April 2, 2001; Published April 18, 2006.
• Brief Description: This patent describes a system for distributing streaming media from a network of servers, which may be geographically dispersed. It focuses on how a client can be directed to an optimal server and how different components of a media presentation (like video, audio, and slides) can be synchronized and delivered from potentially different servers. It mentions delivering media streams at different bitrates to accommodate varying client connection speeds.
• Potential Anticipation of Claim(s):
  • This reference teaches the concept of providing multiple streams at different bitrates to accommodate client bandwidth, a foundational element of adaptive streaming. As such, it could be seen as anticipating the basic concept within the system and method claims of creating multiple quality levels of a stream. However, it does not appear to describe the specific process of segmenting the content into discrete, independently addressable files ("streamlets") that can be requested via standard protocols like HTTP. The '042 patent appears to focus more on traditional streaming protocols and server selection rather than the file-based segmentation approach of US 11,991,234.

2. U.S. Patent Application Publication No. 2002/0194358 A1 (filed Jun. 15, 2001) - "System and method for providing content on demand" (Assignee: Digital Fountain, Inc.)

• Publication/Filing Date: Filed June 15, 2001; Published December 19, 2002.
• Brief Description: This publication details a method for delivering content by encoding it into a "fountain" of data packets, where a client can reconstruct the original content by receiving a sufficient number of these packets, regardless of which specific packets are received. This is a form of forward error correction. The system can adapt by sending more or fewer packets based on network conditions. It also discusses encoding content at different resolutions and frame rates.
• Potential Anticipation of Claim(s):
  • While focused on a different technical mechanism (fountain codes), this reference discloses the idea of encoding content into multiple versions with different characteristics (e.g., resolution, frame rate) to adapt to client conditions. This touches upon the multi-bitrate aspect of the claims in US 11,991,234. However, it does not teach the specific segmentation into time-indexed "streamlets" that are then encoded into different bitrate versions as discrete files. The '358 publication's method is about packet-level redundancy, not file-based segment switching.

3. U.S. Patent Application Publication No. 2002/0069288 A1 (filed Nov. 21, 2000) - "Method and apparatus for playing a composite presentation from a plurality of sources" (Assignee: Microsoft Corporation)

• Publication/Filing Date: Filed November 21, 2000; Published June 6, 2002.
• Brief Description: This reference describes a system for playing a composite media presentation where different media elements (e.g., video, audio, images) are retrieved from multiple, independent sources and synchronized for playback on a client. It mentions selecting a stream from a source that offers multiple bitrates. A "metafile" or playlist directs the client on how to assemble the presentation.
• Potential Anticipation of Claim(s):
  • This publication anticipates the general concept of a client retrieving media from multiple sources and selecting from available bitrates. This is relevant to the system claims of US 11,991,234 which describe a client retrieving streamlets from various web servers (System 100). The use of a metafile is also analogous to the "Virtual Timeline (VT)" and "Quantum Media Extension (QMX)" described in the '234 patent (FIGS. 6a-6c). However, it does not explicitly describe the server-side process of pre-segmenting the entire media file into small, uniform-duration chunks, and then creating discrete, alternative bitrate versions of each chunk.

4. U.S. Patent Application Publication No. 2002/0152332 A1 (filed Apr. 12, 2001) - "Dynamic content delivery" (Assignee: Inktomi Corporation)

• Publication/Filing Date: Filed April 12, 2001; Published October 17, 2002.
• Brief Description: This patent application describes a network of content delivery nodes that can dynamically alter content based on client requests or network conditions. It discusses transcoding content into different formats or bitrates at edge servers closer to the user. This allows for adaptation without having to store all versions of the content at the origin server.
• Potential Anticipation of Claim(s):
  • This reference discloses the concept of delivering content at different bitrates based on network conditions, which is central to the '234 patent. The idea of transcoding at an edge server is a form of generating different bitrate versions of content. However, the '234 patent's claims are focused on a specific pre-processing architecture: segmenting into streamlets and creating sets of these streamlets at the origin before distribution. The '332 publication's dynamic, on-the-fly transcoding at an edge location is a different architectural approach and does not appear to anticipate the specific segmentation and encoding method claimed. Furthermore, it does not disclose the master/host distributed encoding system based on completion bids.

Conclusion on Prior Art

The prior art cited against US 11,991,234 and its parent applications establishes that the general concepts of adaptive bitrate streaming, delivering content from multiple servers, and using metafiles to describe media presentations were known in the art prior to the patent's priority date of April 30, 2004.

However, the specific combination of features appears to be what lends patentability to the claims of US 11,991,234. The key distinctions are:

• Discrete File-Based Segmentation: The method of creating small, independent, time-indexed media files ("streamlets") that can be served over standard HTTP from any web server is a cornerstone of the invention. The prior art tends to discuss continuous streams or dynamic transcoding rather than this specific pre-segmentation approach.
• Distributed Encoding System: The architecture of a "master module" assigning encoding jobs for individual streamlets to a plurality of "host computing modules" based on "job completion bids" (as described in FIG. 5a) is a specific implementation detail that is not clearly disclosed in the cited references. This system is particularly advantageous for encoding live events in near-real-time, which the patent highlights as a key benefit.

Therefore, while the cited art anticipates the general goal of adaptive streaming, it does not appear to explicitly teach or suggest the specific apparatus, system, and method for achieving it as claimed in US 11,991,234, particularly concerning the streamlet segmentation and the master/host bidding-based encoding system.