 
Summary: The Center for Control, Dynamical Systems, and Computation
University of California at Santa Barbara
Spring 2009 Seminar Series
Presents
Information Theoretic Bounds for Distributed
Computation
Ola Ayaso
Electrical & Computer Engineering
Georgia Institute of Technology
Friday, April 03, 2009 3:00  4:00pm Webb 1101
Abstract:
We seek to understand the fundamental limitations, of distributed computation over networks, imposed
by limited communication capabilities and network topology. Computation in networks arises in various
contexts, such as wireless and sensor networks, consensus, and belief propagation. In each of these
situations, individual nodes have access to only part of the information they need and must communicate
via the network in order to execute required computations. In our formulation, each node has an initial
measurement and would like to estimate a given function that depends on the initial measurements of
multiple nodes. An underlying topology determines which nodes can communicate directly with each
other over noisy channels. A distributed computation algorithm specifies both the communications and
computations that nodes will execute to derive the desired function estimate. Using information theoretic
