Presented by:
Young-Han Kim
Date:
Monday 23rd July 2018 - 14:45 to 15:30
Venue:
INI Seminar Room 1
Abstract:
Many
interesting source and channel models have memory. When the temporal dependence
fades away sufficiently fast (or more precisely, if the source or noise process
is ergodic), the standard coding techniques developed in classical Shannon
theory can be extended beyond i.i.d. memoryless cases, resulting in limiting
expressions for rate--distortion and capacity. The key idea behind this
extension is the ergodic decomposition of stationary processes, which was
utilized earlier by Gallager for rate--distortion theory of ergodic sources and
by Kim for capacity of ergodic channels with or without capacity. Such
interplay between information theory and ergodic theory is expected to play an
important role in problems other than point-to-point source and channel coding.
Some technical background can be found in
The video for this talk should appear here if JavaScript is enabled.
If it doesn't, something may have gone wrong with our embedded player.
We'll get it fixed as soon as possible.
If it doesn't, something may have gone wrong with our embedded player.
We'll get it fixed as soon as possible.