Abstract :
We provide a compact characterization of outage rates for a wideband communication system whose parameters are chosen to model an outdoor cellular downlink. The base station transmitter is equipped with an antenna array, while the mobile receiver has a single antenna. Our analysis quantifies the effects of frequency and spatial diversity for measurement-based channel models available in the literature. Design prescriptions based on our framework would apply, for example, to fourth-generation cellular systems using orthogonal frequency-division multiplexing. Our information-theoretic computations yield the following findings. 1) Complex models typically employed in simulations can be replaced by simple, bandwidth-dependent, tap-delay-line models without loss of accuracy. 2) The spectral efficiency (i.e., the achievable rate, divided by the bandwidth) is well approximated as a Gaussian random variable, so that it is only necessary to specify its mean and variance in order to compute the outage rates. We provide analytical formulas for the means and variance as a function of the space–time channel model, and verify that the resulting outage rates match closely with simulation. 3) For a wide class of outdoor channels, the mean spectral efficiency depends only on the spatial diversity, while the variance depends on the spatial and frequency diversity via a product. Our definitions of frequency and spatial diversity have physically motivated interpretations, and do not rely on high signal-to-noise ratio asymptotics, as in prior work.
