Abstract
In wireless communications, sums of lognormal random variables occur in many problems because signal shadowing is well modeled by the lognormal distribution. Since the lognormal sum distribution is known to have no analytic expression, a lognormal sum is usually approximated by a single lognormal random variable. In this paper, we provides two methods of lognormal sum approximation. One is based on the probability density function of lognormal random variables, and the other is a combination of Fenton-Wilkinson method and Schwartz-Yeh method. Numerical examples are provided to compare these approximations.