Long range prediction techniques for mobile wireless communication channels

A ray tracing algorithm and simulation tool is developed using Microsoft Visual C++. This tool is capable of computing the channel model as seen by a moving mobile receiver in the presence of a configurable geometry of reflectors. The tool provides valuable information, such as angle of arrival, phase and magnitude for each electromagnetic field unit vector for all the rays (including diffracted rays) that contribute to the channel model.; Next, certain stationary time invariant autoregressive prediction techniques are described and evaluated for several test cases from the ray tracing simulation tool. These test cases include Ricean and Rayleigh fading, as well as Rayleigh fading including diffracted rays, and several very nonstationary test cases, where the number of reflectors contributing to the channel model changes abruptly.; The information provided by the ray tracing tool is used to understand the time varying nature of the mobile communication channel. A nonlinear phase in the field from each ray, which causes the communication channel to be nonstationary, is analyzed and quantified for a variety of rays from different channels. It is concluded that this must be accounted for if reliable predictions are to be developed.; Several time varying autoregressive techniques, which are implemented in MATLABRTM are developed and tested. The prediction results from these techniques provide more reliable predictions further into the future than the stationary time invariant autoregressive prediction techniques.