Influence of measurement conditions on the evaluation of some radio channel parameters

Communications with mobile stations undergo perturbations due to the necessary radio link. Digital systems are particularly sensitive to multipath effects and frequency selectivity. The wide-band characteristics of the channel are complex and usually summarized by parameters like the delay spread, delay window, or correlation bandwidth. The design of systems is based on statistical studies of these parameters extracted from intensive experimental campaigns. The theoretical definitions of these parameters are widely accepted. However, we have observed that some differences may arise in the practical computation method used. We first show in this paper that the evaluation method and the measurement conditions may lead to differences in parameter estimates which cannot be neglected. We used a large number of complex impulse responses measured in a UHF urban cellular environment as a test set. Theoretical relation were developed to predict the effects of some experimental conditions. These formulas show a good match with the mains changes observed in the measurements. From this analysis, some recommendations are proposed to make the statistical evaluation of these parameters more relevant. Relationships between different channel parameters are also investigated.