The University of British Columbia

We have determined the form of the distributions that best describe the manner in which polarization states disperse across the Poincare sphere over time, i.e., the second- order statistics of dispersion, and how the parameters of these distributions are related to the Ricean K-factors and the cross- correlation coefficient that describe the diversity channel. When sampled over time, the trajectory taken by the polarization states resembles a random walk across the sphere. Of particular note is our observation that the form of the turning angle distribution provides a mechanism that constrains the size of the polarization cluster when the Ricean K-factors and cross-correlation between diversity branches are both high. The results will be of interest to those engaged in the design of polarization adaptive antennas and methods for simulating polarization dispersion.