Pollard, Ritsuko Hamahata

For radiowave propagation on earth-space communication links at high frequencies such as Ka-band, the effect of atmospheric gaseous absorption (mainly due to oxygen and water vapor) is the primary cause of attenuation. This thesis examines the applicability of the surface based Crane's model currently employed by the Advanced Communications Technology Satellite (ACTS) propagation experiment for estimation of attenuation due to atmospheric gaseous absorption (AGA), developed for Oklahoma, to sub-tropical climate regions such as Florida. The Microwave Propagation Model is used as a basis of comparison since it uses the direct atmospheric measurements (temperature, relative humidity, and pressure) made at different levels of the atmosphere with radiosonde instrumentation. The AGA was individually examined for oxygen and water vapor. Finally, accuracy of the Crane's model was verified by computing the attenuation results using real acquired data for both models and comparing their results in various ways for several months.

Due to a continuous growth of demand in voice and data communications for wireless systems, there is an ongoing challenge to design improved radiowave communication links. Polarization is one of the key properties of electromagnetic waves used in wireless communication and is the least studied in a scattering environment. A detailed understanding of how signals become de-polarized can improve the propagation models and can lead to more accurate propagation predictions and possibly new applications in the use of polarization. Potential beneficiaries of a system design using multi-dimensional polarization constellations to provide maximum polarization separation might include inter-satellite links, WLAN, LEO satellites [BSPP92] [IPLJ81] [SIKA98] and other systems operating in an environment where depolarization is insignificant. To examine the applicability of polarization in wireless communication systems, polarization field measurements were conducted, and the results and analysis are presented in this dissertation. Based on the analysis, a statistical model that characterizes polarization in relation to Poincare sphere is developed and presented. Design and implementation of an N-constellation diversity scheme that provides maximum polarization separation distance is presented. Furthermore, a decision-making algorithm is utilized for detection of the received electric field that selects the minimum Euclidian distance between the transmitter and receivers in Stokes space. The scenario is simulated for an N-constellation scheme for N = 2, 3, 4 based on the design scheme utilizing the developed statistical model and decision-making algorithm and is used to evaluate the detectability and performance comparison for various values of standard deviations.