A tool for simulation and visualization of distributed estimation in wireless sensor networks

Abstract

A Java-based tool has been designed and implemented to simulate distributed estimation on a wireless sensor network (WSN) platform. The tool allows users to vary several system parameters of the WSN such as the parameter to be estimated; the amount of and type of noise in the system; types of channel impairments; algorithms for data processing at the sensors; and types of estimators. The simulator has been developed by fulfilling the requirements of usability principles in visual software design, and it produces outputs that have relevance to various sensor network applications such as localization, synchronization, and fault detection. Educational opportunities introduced by the tool are two-fold. It can be used in senior-level undergraduate digital signal processing, communications, and digital controls classes to introduce students basic statistical signal processing concepts such as noise and sample estimation in noise. The current version of the tool has been evaluated via student feedback in a workshop. Preliminary results are promising, and will allow us to make modifications to the simulation tool as necessary.