| Radio Frequency (RF) propagation models are an essential tool in the area of Wireless Sensor Networks (WSN). Accurate RF propagation models may be utilized to facilitate planning and deployment of sensor networks, increase battery efficiency of sensor nodes, and improve accuracy of localization and target-tracking applications that rely on the knowledge of the received signal power. Most of the RF propagation models that are currently used in WSN were originally developed to provide signal prediction in traditional high-power wireless systems such as satellite and personal communication systems. Such approach is not justified since WSN has different features than traditional wireless systems. These features must be taken into account for signal prediction in WSN to be accurate.;This dissertation aims at observing and characterizing RF propagation behavior in various outdoor deployment scenarios of WSN. This has been accomplished by performing a series of in-field studies and examinations of RF propagation in several potential deployment environments. The environments in this study include concrete surface, artificial turf ground, short natural grass, long natural grass, sand terrain, sparse tree, and dense tree. From the actual RF measurements, an empirical path loss model is derived for each environment. The parameters of these models include path loss exponent, path loss at a reference distance, and variation in the received signal level. Such important parameters were approximated through sufficiently large data-sets.;The comparisons between the path loss predicted by the developed empirical models and popular theoretical RF propagation models demonstrate the unsuitability of these models as a signal prediction tool for WSN deployed outdoors. In addition, path loss values predicted by different empirical path loss models were dissimilar. Such dissimilarity is due to the differences that exist in the wireless channel of each environment. Finally, physical effects of each of environment are studied based on the experimental evaluations. |
