Browse Prior Art Database

Multi-Rate Video Transmission Scheme

IP.com Disclosure Number: IPCOM000116969D
Original Publication Date: 1995-Dec-01
Included in the Prior Art Database: 2005-Mar-31
Document File: 4 page(s) / 134K

Publishing Venue

IBM

Related People

Kandlur, DD: AUTHOR [+2]

Abstract

Certain multimedia applications (e.g., videoconferencing, video distribution) may require large amounts of Bandwidth (BW) when sent across a network. In a heterogeneous network environment, not all stations will have the same processing power and bandwidth access to handle an equivalent amount of communication and presentation throughput. Consequently, it is highly desirable that the multimedia streams generated by any given source may scale to allow different destinations to receive the streams at a rate which they can handle. A method of scaling the multimedia stream to allow multiple stations with varying capabilities to receive varying rates (qualities) of the stream is needed.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 48% of the total text.

Multi-Rate Video Transmission Scheme

      Certain multimedia applications (e.g., videoconferencing, video
distribution) may require large amounts of Bandwidth (BW) when sent
across a network.  In a heterogeneous network environment, not all
stations will have the same processing power and bandwidth access to
handle an equivalent amount of communication and presentation
throughput.  Consequently, it is highly desirable that the multimedia
streams generated by any given source may scale to allow different
destinations to receive the streams at a rate which they can handle.
A method of scaling the multimedia stream to allow multiple stations
with varying capabilities to receive varying rates (qualities) of the
stream is needed.

      A hierarchical compressed video generation and transmission
scheme is proposed which will allow multiple destinations to receive
different components of a video stream depending on their
communication and processing capabilities.  The scheme is based on
the segmentation of a video stream into a hierarchy of components
conforming to the video compression standards to which they apply.
The notion is illustrated in Fig. 1, where a video stream generated
by station A, is segmented into three components C1, C2, and C3,
which may be received separately or in combinations by stations B
through F, depending on their desired quality or communication and
processing capabilities.  Component C1 may be a basic rate video,
while components C2 and C3 carry higher rate information.  In some
cases the components may be entirely independent and be combined in
different ways to achieve desired rates.  In other cases, the
components
have dependencies and only certain combinations are meaningful.

      For example, divide a 30 frame per second (fps) video stream
into five consecutive sequences of six frames (Fig. 2).  Of each six
frame sequence (referred to as a set), the first frame is sent as
component C1, the 3rd and 5th are sent as component C2, and the 2nd,
4th, and 6th are sent as component C3.  As a result any station
receiving C1 will obtain a 5 fps video stream.  Any station receiving
C1 and C2 will obtain a 10 fps stream and by receiving C3 will obtain
a 15 fps stream.  To obtain the full 30 fps a station must receive
all three components.

      The above scenario is perfectly suited to a Motion-JPEG stream,
where each video frame is independent of any other, and the bandwidth
required for each component is linearly dependent on the number of
frames per second of that component.  In an MPEG or px64 video
encoding format, the partitioning is not as simple.  These video
streams are encoded using interframe encoding techniques, which helps
reduce the required bandwidth, but imposes a dependency between
frames.  Hence, each component cannot be treated fully independently
of all other components.

      Once again, divide the video stream into sequences (sets) of
six frames, where the first fra...