Statistical analysis of co-channel interference in wireless communications systems

Co-channel interference is recognized as one of the major factors that limits the capacity and link quality of a wireless communications system. An appropriate understanding of the statistical behavior of the co-channel interference is therefore required when analyzing and designing techniques that mitigate its undesired effects. The total co-channel interference in a wireless communications system is usually modeled as the sum of lognormally distributed signals, and is generally assumed to be itself lognormally distributed. Based on this assumption, several methods for estimating the moments of the resulting lognormal distribution have been proposed. The accuracy of these methods has been studied in previous works, under the assumption of having all summand signals (individual interference signals) identically distributed. Such an assumption rarely holds in practical cases of emerging wireless communications systems, where co-channel interference may stem from far-away macrocells and nearby transmitters, causing the interference signals to have different moments. In this paper we present an analysis of the accuracy of two popular methods for computing the moments of a sum of lognormal random variables, namely Wilkinson's method and Schwartz and Yen's method, for the general case when the summands have different mean values and standard deviations in decibel units. We show that Schwartz and Yeh's method provides better accuracy than Wilkinson's method and is virtually invariant with the difference of the mean values and standard deviations of the summands.