Characterization of randomly time variant linear channel-an application to directional radio channels

Summary form only given. Next generation wireless communication systems will exploit emerging technologies such as smart antennas, aiming at capacity enhancement, cell range and coverage extension, and interference reduction. The current activities in measurement and modeling of mobile radio channels have focused on directional channels in order to facilitate the angular domain characterization of the channel, and the spatio-temporal processing of the received signal. The characterization of the mobile radio channel requires the analytic and thorough investigation of three (3) propagation phenomena, namely: time dispersion (or frequency selectivity), frequency dispersion (or time variance), and angular dispersion (or spatial variance). The use of antenna arrays at the receiver of the measurement equipment, such as vector channel sounders, enables us to investigate the spatial dispersion of the channel, and insert two more parameters in the system functions. These are the direction cosine of the arriving waves, and a spatial parameter that is related to the antenna array. Especially for linear arrays it has been proved that a single Fourier transform connects the spatial parameter with the azimuth. Hence a more comprehensive description of the channel requires a set of six (6) parameters, resulting in eight (8) basic functions. This paper presents the system characteristic functions needed to describe the deterministic channel, and furthermore the autocorrelation functions used to statistically characterize the random channel. The relationship between the aforementioned functions is depicted, and a consistent analysis is presented for specific categories of channels. Hence ergodicity, stationarity, and uncorrelated scattering are issues addressed.