Stackable middleware services for advanced multimedia applications. Final report for period July 14, 1999 - July 14, 2001

Feng, Wu-chi ; Crawfis, Roger, Weide, Bruce

Ohio State University

关键词: Compression;    99 General And Miscellaneous//Mathematics, Computing, And Information Science;    Management;    Catalysts;    Algorithms;   

DOI  :  10.2172/771365
RP-ID  :  DOE/ER/25403--1
RP-ID  :  FG02-99ER25403
RP-ID  :  771365

美国|英语

来源: UNT Digital Library

【 摘 要 】

In this project, the authors propose the research, development, and distribution of a stackable component-based multimedia streaming protocol middleware service. The goals of this stackable middleware interface include: (1) The middleware service will provide application writers and scientists easy to use interfaces that support their visualization needs. (2) The middleware service will support a variety of image compression modes. Currently, many of the network adaptation protocols for video have been developed with DCT-based compression algorithms like H.261, MPEG-1, or MPEG-2 in mind. It is expected that with advanced scientific computing applications that the lossy compression of the image data will be unacceptable in certain instances. The middleware service will support several in-line lossless compression modes for error-sensitive scientific visualization data. (3) The middleware service will support two different types of streaming video modes: one for interactive collaboration of scientists and a stored video streaming mode for viewing prerecorded animations. The use of two different streaming types will allow the quality of the video delivered to the user to be maximized. Most importantly, this service will happen transparently to the user (with some basic controls exported to the user for domain specific tweaking). In the spirit of layered network protocols (like ISO and TCP/IP), application writers should not have to know a large amount about lower level network details. Currently, many example video streaming players have their congestion management techniques tightly integrated into the video player itself and are, for the most part, ''one-off'' applications. As more networked multimedia and video applications are written in the future, a larger percentage of these programmers and scientist will most likely know little about the underlying networking layer. By providing a simple, powerful, and semi-transparent middleware layer, the successful completion of this project will help serve as a catalyst to support future video-based applications, particularly those of advanced scientific computing applications.