Understanding the Weichselberger model: A detailed investigation

Abstract

The need for accurately characterizing the spatial domain of the wireless channel has become increasingly important with the growing use of wireless technology. To facilitate the development of signaling strategies and Multiple-Input Multiple-Output (MIMO) system architectures, many different channel modeling strategies have been employed. When it is desirable to create many different channel matrix realizations for a variety of environments, analytical models are typically used. One of the most popular analytical models, the Kronecker model, has been shown to be deficient in its representation of the channel. An alternate model, proposed by Weichselberger, has been shown to provide a better match than the Kronecker model in predicting a variety of channel metrics. As in the Kronecker case, this model requires knowledge of the link end correlation matrices. However, it also requires the additional knowledge of a power coupling matrix. In this paper, we use beamforming techniques to understand how the environment is represented in the Weichselberger channel model using measured MIMO channel matrices. We examine the model underpinnings and how the component parts capture the physics of the environment. Additionally, synthesized data is used to further examine the properties of the model's power coupling matrix.