URSI/Remote Sensing Club of Finland/IEEE XXIII National Convention on Radio Science and Remote Sensing Symposium

In the FW approach [1, 2, 3] the mean ¡ and the standard deviation ¢ σ of the approximation (Gaussian, dB) are derived by matching the first two moments of the power sum with the first two moments of the sum of the individual contributions. The FW approach is applicable when the standard deviations of the lognormal components are lower than 4 dB for uncorrelated signal components. For higher deviation values, this approximation tends to underestimate the mean and overestimate the variance of the sum distribution. When there is correlation between the components, the FW approximation is quite accurate for higher deviation values (up to 12 dB) also [1]. Furthermore, in several applications the main interest is in the tails of the cumulative distribution functions. The FW approximation is reasonably accurate in the tail distributions. It is very simple and thus attempting choice to be used in the outage evaluations.