Abstract
Large Scale WSN deployment is considered to be a very complex operation in terms of both efficiency and cost. Sensor cost has never been low, not even today, though at the same time the large areas to be covered call for a huge number of wireless sensors. The complexity escalates when it comes to forest areas, as such areas consist of an assortment of different landscapes and terrains and therefore, they cannot be modeled in detail. In this paper a first approach to forest area classification is attempted, so that simulation models can describe the attenuation of the alarming signal and the corresponding connectivity in the classified area. A number of models produced by Shawn simulator are provided, for forest areas with dense and scant plantation of trees. The model that best converges to the real experiment metrics per area is determined and is therefore referenced as “the model” for the specific classified area. Provided the fact that the simulation model per classified area is granted, a step to the optimal sensor topology is performed, as the connectivity and path loss problem is mainly and in most cases linearly confined to distance variations. Recent Patents are also covered. The simulation models along with optimal sensor topologies can become a guide in Large Scale WSN deployment primarily for the areas classified as “dense” and “scant” in terms of plantation, with the aim to deploy a reliable WSN for environmental purposes.
Keywords: WSN simulation, large scale, forest areas, Shawn simulator, connectivity, topologies, path loss, Near-Earth Propagation Models, quality factor, WSN deployments, prototype areas, spatial features, Moderate Resolution Imaging Spectroradiometer (MODIS), TOSSIM
10