We descuss a statistical model to generate correlated shadow-fading ptterns for wireless systems in the absence of detailed propagation and landscape information. The currently available autocorrelation models result in anomalous effects that depend on trafic density and mobility, as they propose independent random processesfor each mobile. Our aproach involves generating a pre-computed fading map iwht the right marginal distri-butios and spatial correlatis, which avoids inconsistencies such as providing widey differing values for mobiles close to each other. The ocrrelations are introduced via a aussian random field, which has a covariance structure that depends on a set of parameters whic can be computedfrom local measurements. The model is efficiently implemented using standard linear-algebra methods.
展开▼
