Reception reliability estimation techniques for mobile radio system planning

Abstract

System planners are faced with implementing wideband digital cellular radio systems that can provide the very high capacity and high quality communications required by cellular users at an affordable price. To assess the likelihood of achieving adequate system performance, system planners require relatively simple but effective techniques to estimate the performance of potential system configurations. These techniques should incorporate parameters that can easily be estimated by a simple appraisal of the propagation environment. In the literature there has been a parallel development between two general types of reception reliability estimation techniques, namely bit detection and statistical channel. This thesis presents a review of these two types of techniques and assesses their usefulness as tools for system planners. By linking the bit detection techniques to the statistical channel techniques, a consistent range of reception reliability estimation tools that system planners can use have been developed in this thesis. These new tools are called improved statistical channel techniques. Because future systems are envisaged to be digital and wideband, this research focuses on estimation techniques for digital wideband mobile radio systems, where the effects of power delay spread need to be considered. Wideband measurements have been performed as part of this research to determine the most appropriate parameters to use to describe power delay spread. Improved statistical channel techniques incorporating delay spread, noise, cochannel interference and diversity reception are presented. The results of the improved statistical channel techniques are compared to those produced by bit detection techniques. The results show that these improved techniques are practical for system design, requiring less than a thousandth of the computation time and closely predicting the results of bit detection techniques, given appropriate choice in parameter values. A method for predicting these parameter values for a wide range of environments and system technologies is presented. Recommendations for further development of the improved statistical channel techniques are discussed.