SSP'05 -- Proceedings

	IEEE/SP 13th workshop on Statistical Signal Processing July, 17-20, 2005 - Bordeaux - France

Information regarding the paper

Title

On Optimal Parametric Field Estimation in Sensor Networks

Author(s)

Ke Liu	Ohio State University
Hesham El-Gamal	Ohio State University
Akbar Sayeed	University of Wisconsin-Madison

Get the paper in PDF format

To obtain Acrobat Reader (version 5 minimum required) necessary to his read.

Abstract

We develop a framework for field estimation using wireless sensor networks, subject to network power and communication channel constraints. The nodes communicate appropriate local statistics to a fusion center over a wireless MAC. We characterize the optimality conditions under which simple uncoded transmission yields the optimal (centralized) 1/k distortion scaling with the number of nodes. We proposed a universal parameter estimation framework based on local type/histogram statistics that achieves optimal scaling laws for finite alphabet source. However, it is shown that uncoded transmission generally breaks down under fading channels. We propose a simple coded strategy that achieves logarithmic scaling with total network power.