DST-based combining scheme outperforms maximal-ratio combining

A novel combining scheme based on the Dempster-Shafer theory (DST) is presented. The proposed Dempster-Shafer combining (DSC) uniquely assigns `mass` to a set consisting of the adjacent constellation points instead of probability to characterise uncertainties in decision-making, and then transfers the masses to the corresponding `belief interval`. Taking advantage of multiple information sources, more reliable decision statistics can be achieved by reducing the belief interval. It is beneficial to initially assume that the transmitted signal is included in a set containing diverse decision candidates, because it lacks convincing evidence to make an unambiguous decision directly. Uncertainties processing of the proposed DSC makes it possible for DSC to outperform the conventional maximal-ratio combining.