On the Design and Optimization of Distributed Signal Detection and Parameter Estimation Systems.
Abstract
In this report, the problems of hypothesis testing and parameter estimation in a distributed framework are considered. First, hypothesis testing in distributed systems with data fusion is treated. The approach can easily be applied to decentralized systems without data fusion. Optimal decision rules at the detectors and optimal fusions rules are derived for the distributed hypothesis testing problems using the Neyman-Pearson criterion, the general Bayesian criterion and the minimum equivocation criterion. Correspondence between information theory and detection theory is established. Decentralized postdetection integration problems are also considered and optimum fusion rules, as well as optimum decision rules at the individual detectors are obtained for two proposed schemes. Next, decentralized Bayesian parameter estimation is considered and optimum estimation rules at the local estimators and optimum combining rules are obtained for the minimum mean square error criterion, the absolute error criterion and the uniform cost function criterion.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 1987
- Accession Number
- ADA193934
Entities
People
- Imad Y. Hoballah
- Pramod Varshney
Organizations
- Syracuse University