In my talk at this workshop last year, I discussed a new source coding
problem, in which a finite number of nodes collect local measurements
of a wave in space, and deliver an encoding of these readings to a
decoder whose task is to produce an approximation of the spatial wave.
I referred to this problem as the ``multiterminal source coding problem
for spatial waves.''  

This year I will present a new result on the problem of coding waves:
the entropy of densely sampled spatial waves does not grow unbound with
the sampling density.  The interest in this result stems from the fact
that, for other more classical data models, a different conclusion holds
(e.g., the entropy of oversampled bandlimited fields does grow unbound
with sampling density), and therefore this has implications in terms
of the ability/inability of a dense sensor network to transport all of
the data it gathers.

Joint work with Joseph M. Rosenblatt (UIUC/Math).