Electrical computers and digital processing systems: multicomput – Computer-to-computer protocol implementing – Computer-to-computer data streaming
Reexamination Certificate
1998-08-24
2004-02-24
Etienne, Ario (Department: 2157)
Electrical computers and digital processing systems: multicomput
Computer-to-computer protocol implementing
Computer-to-computer data streaming
C709S205000, C345S473000
Reexamination Certificate
active
06697869
ABSTRACT:
FIELD OF THE INVENTION
The invention relates to streaming multimedia files via a network. The invention relates in particular to enabling the emulation of streaming graphics or video animation over the Internet within a broadcast context.
BACKGROUND ART
The term “streaming” refers to transferring data from a server to a client so that it can be processed as a steady and continuous stream at the receiving end. Streaming technologies are becoming increasingly important with the growth of the Internet because most users do not have fast enough access to download large multimedia files comprising, e.g., graphics animation, audio, video, or a combination thereof, etc. Streaming, however, enables the client's browser or plug-in to start processing the data before the entire file has been received. For streaming to work, the client side receiving the file must be able to collect the data and send it as a steady stream to the application that is processing the data. This means that if the client receives the data faster than required, the excess data needs to be buffered. If the data does not arrive in time, on the other hand, the presentation of the data will not be smooth.
The term “file” is used herein to indicate an entity of related data items available to a data processing and capable of being processed as an entity. Within the context of the invention, the term “file” may refer to data generated in real-time as well as data retrieved from storage.
Among the technologies that are currently available or under development for the communication of graphics data via the Internet are VRML 97 and MPEG-4. VRML 97 stands for “Virtual Reality Modeling Language”, and is an International Standard (ISO/IEC 14772) file format for describing interactive 3D multimedia content on the Internet. MPEG-4 is an ISO/IEC standard being developed by MPEG (Moving Picture Experts Group). In both standards, the graphical content is structured in a so-called scene graph. A scene graph is a family tree of coordinate systems and shapes, that collectively-describe a graphics world. The top-most item in the scene family tree is the world coordinate system. The world coordinate system acts as the parent for one or more child coordinate systems and shapes. Those child coordinate systems are, in turn, parents to further child coordinate systems and shapes, and so on.
VRML is a file format for describing objects. VRML defines a set of objects useful for doing 3D graphics, multi-media, and interactive object/world building. These objects are-called nodes, and contain elemental data which is stored in fields and events. Typically, the scene graph comprises structural nodes, leaf nodes, interpolation nodes and sensor nodes. The structural nodes define the spatial relationship of objects within a scene. The leaf nodes define the physical appearance of the objects. The interpolation nodes define animations. The sensor nodes define user interaction for particular user input modalities. VRML does not directly support streaming of data from a server into a client. Facilities such as synchronization between streams and time stamping that are essential in streaming do not exist in VRML. However, VRML has a mechanism that allows external programs to interact with VRML clients. This has been used in sports applications to load animation data into the client. See, for example, “VirtuaLive Soccer” of Orad Hi-Tec Systems, Ltd at <http://www.virtualive.com>. This web document discusses a process for producing realistic, animated, three-dimensional graphic clips that simulate actual soccer match highlights for being sent via the Internet. The system generates content that complements television sports coverage with multimedia-rich Web pages in near real time. In this example, the process works in two steps. First the graphics models of the stadium and of the soccer players are downloaded along with an external program, in this case a Java Applet. The user can then interact with the external program to request a particular animation. The data for this animation is then downloaded into the client and interacted with by the user. In terms of node type, this process first downloads the structural and leaf nodes, and thereupon the interpolation nodes. By changing the set of interpolation nodes, it is possible to run a different animation sequence. The process used in this example is somewhat equivalent to a single step process in which the user can choose the complete VRML file that contains all the models (structural nodes) and all the animation data (interpolator nodes). This approach leads to long download times before any content can be played on the client. This is experienced as a frustrating experience, especially if compared to TV broadcast where content is available instantly.
The other technology introduced above, MPEG-4, defines a binary description format for scenes (BIFS) that has a wide overlap with VRML 97. MPEG-4, on the other hand, has been designed to support streaming of graphics as well as for video. MPEG-4 defines two server/client protocols for updating and animating scenes: BIFS-Update and BIFS-Anim. Some of the advantages of MPEG-4 over VRML are the coding of the scene description and of the animation data as well as the built-in streaming capability. The user does not have to wait for the complete download of the animation data. For example, in the soccer match broadcast application mentioned earlier the animation an start as soon as the models of the players and the stadium are downloaded. MPEG-4 further has the advantage that it more efficient owing to its BIFS transport protocol that uses a compressed binary format.
Within the context of streaming, the known technologies mentioned above have several limitations with regard to bandwidth usage, packet-loss concealment or recovery and multi-user interactivity, especially in a broadcast to large numbers of clients.
As to bandwidth, the complete animation is generated at the server. This results in a large amount of data that needs to be transported over the network, e.g., the Internet, connecting the client to the observer. For example, in the soccer broadcast application mentioned above, the 22 soccer players need to be animated. Each animation data point per individual player comprises a position in 3D space and a set of say, 15 joint rotations to model the player's posture. This represents 63 floating-point values. If it is assumed that the animation update rate is 15 data points per seconds, a bit-rate of 665 Kbps is required. This bit-rate can be reduced through compression. Typically, using BIFS reduces the bit-rate by a factor of 20, giving a bit-rate of about 33 Kbps. However, this number has not taken into account overhead required for the Internet protocols (RTP, UDP and IP) and for additional data types, such as audio. However, typical modems currently commercially available on the consumer market have a capacity of 28.8 Kbps or 33.6 Kpbs. It is clear that streaming animation causes a problem at the end user due to bandwidth limitations. In the case of a broadcast to a large number of clients, say 100,000 clients, the data stream will need to be duplicated at several routers. A router on the Internet determines the next network point to which a packet should be forwarded on its way toward its final destination. The router decides which way to send each information packet based on its current understanding of the state of the networks it is connected to. A router is located at any juncture of networks or gateway, including each Internet point-of-presence. It is clear that the broadcast could lead to an unmanageable data explosion across the Internet. To prevent that from happening, the actual bandwidth needs to be limited to much lower than 28.8 Kbps.
As to packet loss concealment, VRML-based systems utilize reliable protocols (TCP). Packet losses are not an issue here. In the case of MPEG-4, BIFS uses RTP/UDP/IP. A packet loss recovery mechanism is therefore required. In a point-to-point application, re-transmission of lost packets ca
Mallart Raoul
Sinha Atul
Etienne Ario
Koninklijke Philips Electronics , N.V.
Salad Abdullahi E.
Ure Michael J.
LandOfFree
Emulation of streaming over the internet in a broadcast... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Emulation of streaming over the internet in a broadcast..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Emulation of streaming over the internet in a broadcast... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3339354