Abstract:
A wireless sensor network (WSN) is a self-organizing network, consisting of tiny wireless nodes which carry out a sensing and transferring task in collaboration. WSNs are getting popular because of their ease of deployment, self-organizing capability, low cost and their wide range of applications. This wide spectrum of applications raises most of the time application specific research problems in WSN protocol stack and algorithm design. However, the performance of the most proposed models in literature, have been evaluated on planar surfaces, assuming a distance based sensing and 2D freespace communication model, while assuming a random deployment scheme which commonly takes place in 3D inaccessible terrains. In this thesis, we investigated the problem of incorporating a realistic modeling environment into sensor networks. Our motivation has been the non-realistic and contradictive assumptions in WSN performance evaluations, where the formations of the topographic surface that would normally block the communication and sensing task are not taken into account. We incorporated a 3D terrain model into the performance evaluation of a collaborative target tracking application. The evaluation is done on various artificially generated but realistic terrains, on the performance metrics of mean error which is a measure of tracking accuracy, number of communicating sensor pairs and number of detecting sensors. Our simulations show that the performance predictions could be misleading on the paper design, due to non-realistic assumptions with regards to the WSN deployment region.
