Analysis and Mitigation of Tropospheric Effects on Ka Band Satellite Signals and Estimation of Ergodic Capacity and Outage Probability for Terrestrial Links

The first part of this work covers the effect of the troposphere on Ka band (20-30 GHz) satellite signals. The second part deals with the estimation of the capacity and outage probability for terrestrial links when constrained to quadrature amplitude modulations.;The desire for higher data rates and the need for available bandwidth has pushed satellite communications into the Ka band (20--30 GHz). At these higher carrier frequencies the effects of scintillation and rain attenuation are increased. In regards to the effects of scintillation, the first part of this work quantifies, through the use of a multiple phase screen simulation model, the benefits of using two receive antennas to mitigate tropospheric-induced scintillation on Ka band satellite downlinks. Two representative turbulence profiles are considered, and cumulative distribution curves for scintillation-induced attenuation are generated for selection and maximal ratio combining schemes and compared to those for a single antenna. The results indicate that there can be significant diversity gains achieved by combining two antennas separated by only a short distance. Also, a comparison of simulation results with the results predicted by the basic Rytov approximation shows that at elevation angles greater than 10 degrees, Rytov theory can accurately predict performance benefits of antenna combining, but at elevation angles less than 10 degrees it is better to use multiple phase screen simulations to make performance predictions. In addition, the effects of scintillation-induced phase perturbations on the output power of large aperture antennas is examined. It is found that the output power degradation due to scintillation-induced phase perturbations is generally negligible and can be countered by the simple means of antenna tracking if necessary.;In regards to rain attenuation, this work developed simple methods for estimating the outage probability and outage capacity and ergodic capacity of satellite links due to rain fades. The rain-induced fades of a satellite link are often modeled with a log-log-normal distribution. Researchers have determined methods for calculating the outage probability for Shannon capacity for log-log-normal channels. However, in practical communications systems, the input signal is constrained to a discrete signalling set such as finite-size quadrature amplitude modulations. Under these conditions the outage probability with regards to the constrained capacity is a more accurate measure. A method is detailed in this work for tightly estimating the outage probability and outage capacity of satellite links with quadrature amplitude modulations. In addition this work derives a lower bound for the ergodic constrained capacity of log-log-normal channels. To date, no other method for calculating the outage probability, outage capacity, or a lower bound for the ergodic capacity for a log-log-normal channel with a finite-size quadrature amplitude modulation has been published. Also, this portion of the work quantifies the benefit of using receive diversity to mitigate rain fades, providing the gains in outage capacity due to the use of diversity for a tropical region and a fairly dry region under the constraint that practical constellations are transmitted. The above information and analysis methods provide useful tools for satellite system planners.;The second part of this work examines terrestrial communication links, which can suffer greatly from channel fading or shadowing. Two common statistical models for channels are the Rayleigh distribution and the log-normal distribution. The goal of this second part of the work was to develop a simple method for tightly estimating the ergodic capacity and outage probability of these two channel types when used with quadrature amplitude modulated signalling sets. Specifically an innovative method was developed for estimating the ergodic constrained capacity for Rayleigh and log-normal channels with and without antenna combining. The expressions facilitate straightforward computation of outage probability as well. Researchers have determined methods for calculating the ergodic Shannon capacity for log-normal and Rayleigh channels for single and multiple receive antenna systems. However, in practical communications systems, the input signal is constrained to a discrete signalling set such as finite-size quadrature amplitude modulation constellations. Under these conditions the ergodic constrained capacity is a more accurate measure. The method detailed in this work provides a uniform expression for computing the ergodic capacity, both Shannon and constrained, of Rayleigh and log-normal channels with and without antenna combining. The expressions facilitate straightforward computation of outage probability as well. Both the noise-limited and interference-limited cases are studied. To date, no other method for estimating the outage probabilities for the constrained capacity of Rayleigh or log-normal channels has been published for either the noise-limited case or interference-limited case. Also, no method for estimating the ergodic constrained capacity of a log-normal channel or of an interference-limited Rayleigh channel has appeared in the literature. The analysis methods and information for terrestrial links developed in the second part of this work provide useful tools for the designers of wireless communication systems in general and have particular application to cellular mobile and ultra-wideband systems.