Stochastic modelling of Rice fading channels with temporal, spatial and spectral correlation

Abstract

A statistical characterisation of a small-scale fading mobile radio channel is presented and a fading simulator for the simulation of multiple input–multiple output multicarrier systems is introduced. The derivation of the equivalent lowpass channel as a function of time, transmit and receive antenna positions and subcarrier frequency, leads to a tapped delay-line model with time, space and frequency-selective taps. The taps are modelled as a sum of a deterministic line-of-sight or dominant scattered path and a zero-mean Gaussian part composed of a number of unresolvable scattered paths, and are therefore Rice fading. The Gaussian parts can have the desired temporal and spatiospectral correlations generated by time-correlation shaping filtering and a space–frequency correlation transformation, respectively. The structure and details of the simulator are presented, and the simulator is shown to achieve good accuracy while retaining a reasonable computational complexity.