Publication Type

Journal Article

Version

acceptedVersion

Publication Date

11-2015

Abstract

We develop a new class of event detection algorithms in Wireless Sensor Networks where the sensors are randomly deployed spatially. We formulate the detection problem as a binary hypothesis testing problem and design the optimal decision rules for two scenarios, namely the Poisson Point Process and Binomial Point Process random deployments. To calculate the intractable marginal likelihood density, we develop three types of series expansion methods which are based on an Askey-orthogonal polynomials. In addition, we develop a novel framework to provide guidance on which series expansion is most suitable (i.e., most accurate) to use for different system parameters. Extensive Monte Carlo simulations are carried out to illustrate the benefits of this framework as well as the quality of the series expansion methods, and the impacts that different parameters have on detection performance via the Receiver Operating Curves (ROC).

Keywords

Binomial point process, event detection, Poisson point process, series expansions, wireless sensor networks

Discipline

Computer Sciences | Software Engineering

Research Areas

Software and Cyber-Physical Systems

Publication

IEEE Transactions on Signal Processing

Volume

63

Issue

22

First Page

6122

Last Page

6135

ISSN

1053-587X

Identifier

10.1109/TSP.2015.2452218

Publisher

IEEE

Copyright Owner and License

Authors

Additional URL

https://doi.org/10.1109/TSP.2015.2452218

Share

COinS