On the Performance of Multi-Antenna Techniques for Spatially and Temporally Correlated Wireless Channels

Abstract

As the demand for advanced wireless services continues to grow, system designers must employ innovative signal processing techniques to increase data throughput and maintain reliablity under adverse channel conditions. Multi-antenna techniques, such as space-time coding and beamforming, have shown promise in realizing these goals. As these and other techniques are introduced, understanding their performance in realistic scattering environments is of paramount importance.

This thesis contributes to the field of wireless communications by determining the performance of multi-antenna techniques for spatially and temporally correlated wireless channels. First, we propose a general space-time covariance model that is applicable to arbitrary scatterer geometry, arbitrary array geometry at the base station and the mobile, and includes Doppler effects due to mobile motion. We then apply this model, in conjunction with a two-dimensional Gaussian scatterer model based on recent field measurements, to evaluate the exact pairwise error probability for arbitrary space-time block codes and determine an upper bound on the probability of a block error. In addition, we derive exact closed-form expressions for the symbol error probability for orthogonal space-time block coding, maximum ratio transmission, and beamsteering for spatially correlated quasi-static wireless channels. Finally, we present extensive numerical results that illustrate the performance of these techniques for varying degrees of spatial and temporal correlation. We also provide a comparative performance assessment of beamforming and orthogonal space-time block coding and determine the channel conditions for which one technique is favored over the other.