Obviousness

US patent 8542705 relates to methods, systems, and computer-readable media for efficiently delivering media streams, particularly by immediately providing a key frame upon a channel change request, thereby avoiding the transmission of undecodable predictive frames. An analysis under 35 U.S.C. § 103 suggests that the independent claims (1, 8, and 15) would have been obvious to a person having ordinary skill in the art (PHOSITA) when the patent was filed.

The core problem addressed by US8542705—the delay in displaying an accurate picture during channel changes due to the initial receipt of predictive frames—was well-known in the art, and multiple prior art references explicitly sought to solve this.

Combination of Prior Art References and Motivation for Combination

The independent claims of US8542705 would be rendered obvious by combining the teachings of EP1523190A1 (Microsoft Corporation), US20030005455A1 (Bowers J. Rob), and common knowledge within the field of video encoding, specifically concerning MPEG and Group of Pictures (GOP) structures.

1. EP1523190A1 (Microsoft Corporation) - "Fast channel change"
This patent explicitly addresses the problem of channel change delay in digital television systems by reducing the time it takes for a client to display a decodable image. It teaches:

• A streaming server receiving media streams (referred to as "channel data") and maintaining them in buffers (a "stream buffer").
• Receiving a channel change request from a client device.
• Identifying the "next available I-frame" (key frame) for the requested channel.
• Transmitting this I-frame immediately to the client, allowing the client to "immediately decode and present video data from the new channel, reducing channel change time." This inherently means that no predictive frames are transmitted at the beginning of the channel change request, as the server prioritizes sending a decodable key frame.

2. US20030005455A1 (Bowers J. Rob) - "Aggregation of streaming media to improve network performance"
This reference teaches the concept of aggregating streaming media to improve network performance, implying the use of an "aggregation server." US8542705 itself depicts a streaming server receiving media from a "media aggregation server 361" (FIG. 3).

3. Common Knowledge of a Person Having Ordinary Skill in the Art (PHOSITA) regarding MPEG/GOP Structure
A PHOSITA in the field of video streaming at the time of the invention (2007) would have possessed common knowledge of:

• The Moving Picture Experts Group (MPEG) standard and its use of I-frames (key frames), P-frames, and B-frames (predictive frames). The patent US8542705 explicitly references MPEG, I-frames, B-frames, and P-frames in its background and description.
• The organization of video streams into Groups of Pictures (GOPs), where each GOP typically begins with an I-frame, and subsequent frames within the GOP are predictive frames referencing that I-frame.
• The fact that GOPs often have a defined, fixed length (N), meaning I-frames appear at regular intervals.
• Standard techniques for parsing video stream headers to extract parameters such as GOP size (N) or for inferring GOP boundaries and I-frame locations.

Obviousness Argument for Independent Claims 1, 8, and 15

All claims (Method, System, and Computer Readable Medium) require the following key elements:

• A streaming server receiving multiple media streams (e.g., first and second).
• These streams are maintained in buffers.
• The server is configured to receive the media stream from a media aggregation server.
• The buffers are allocated on a per Group of Pictures (GOP) basis.
• The server determines positions of key frames, specifically by "determining at the streaming server a group of pictures (GOP) size N and identifying Nth frames following the first key frame as subsequent key frames."
• Upon receiving a channel change request, the server identifies the next available key frame for the newly requested stream.
• This key frame is provided to the client, without transmitting any predictive frames at the beginning of the channel change request.

Obviousness Analysis:

1. Receiving multiple media streams at a streaming server and maintaining them in buffers; receiving from a media aggregation server: EP1523190A1 discloses a server managing "channel data" in a "stream buffer," implying the ability to handle multiple channels. The addition of receiving these streams from a "media aggregation server" is taught by US20030005455A1, which focuses on aggregating media for improved network performance. A PHOSITA would find it obvious to integrate an aggregation server into a streaming architecture, particularly when seeking to optimize media delivery.

2. Identifying a next available key frame and providing it without predictive frames on channel change: This core inventive concept is clearly taught by EP1523190A1. It directly states that upon a channel change, the server transmits the "next I-frame" to the client for immediate decoding, thereby reducing delay and preventing the client from receiving undecodable predictive frames first. Similar teachings are present in WO2006041784A2 ("sending an I-frame for the requested channel from the cache"), US20060143669A1 ("The media server sends one or more key frames... without necessarily waiting for the next key frame in the real-time stream"), and US20070107026A1 ("retrieving a key frame from a cache, transmitting the key frame to the client device").

3. Determining key frame positions by "determining GOP size N and identifying Nth frames following the first key frame as subsequent key frames": While not explicitly detailed in EP1523190A1, the need for the server to "identify" the next I-frame would lead a PHOSITA to employ known methods for this task. The structure of video streams, particularly MPEG, where I-frames typically mark the beginning of GOPs and often occur at predictable intervals (e.g., every N frames for a fixed GOP size), was common knowledge. The patent US8542705 itself states that "once a key frame is determined, the streaming server determines the GOP size N and identifies each Nth frame following a key frame as a subsequent key frame," implying this is a known "approach" among "a variety of approaches" for determining key frames. Therefore, a PHOSITA would find it obvious to use this well-understood method for key frame identification within a streaming server to achieve the goal of fast channel change.

4. Buffers allocated "on a per group of pictures (GOP) basis": To efficiently identify and deliver key frames (which are the start of GOPs), a PHOSITA would logically design or organize the server's buffers to be aware of GOP boundaries. This could involve physical segregation or logical indexing to facilitate rapid access to the beginning of any GOP. US20060143669A1 teaches a media server storing "at least one initial sequence of a video stream (i.e. starting from an I frame) for each channel." Storing an "initial sequence" beginning with an I-frame is functionally equivalent to being "GOP-aware" or "per GOP basis" buffering, as the I-frame initiates a GOP. This is an obvious engineering optimization for a system designed to serve I-frames promptly.

Motivation for Combination:
The overriding motivation for a PHOSITA to combine these elements is to directly address the acknowledged problem of delayed and distorted video during channel changes, as clearly articulated in EP1523190A1 and reiterated in US8542705. Faced with the task of implementing a "fast channel change" system (as taught by EP1523190A1), a PHOSITA would naturally draw upon their common knowledge of video encoding standards like MPEG. This knowledge would lead them to leverage the inherent GOP structure and the predictable recurrence of I-frames (at Nth frame intervals) as the most straightforward and efficient means for a streaming server to identify, store, and quickly retrieve these essential frames. Organizing buffers on a GOP-aware basis (or "per GOP basis") is a logical consequence of needing rapid access to I-frames that initiate GOPs. Incorporating a media aggregation server (as taught by US20030005455A1) would be an obvious architectural choice to manage content from multiple sources efficiently in such a streaming environment. The combination of these known elements and techniques achieves the desired functional outcome of fast, smooth channel switching, with a reasonable expectation of success.