Abstract

In order to correctly investigate and understand wireless sensor network (WSN) protocols such as data collection, dissemination, and IPv6 integration, real-world experiments need to be executed in a comparable and systematic way. Current WSN sites are built around specific scenarios and environments. In that sense, each site is unique with regard to its network properties, topological and hardware characteristics. This coupled with the real vagaries of the wireless communication very often leads to significantly different performance results of evaluation of protocols or applications within and across sites [5], [4]. Furthermore, the choice of experimentation sites which suit best a particular evaluation follows the more is better approach, rather than being systematically made. A systematic selection of experimentation sites is needed in order to ensure exposing the system under test to significantly