Modeling and simulation of non-Gaussian correlated clutter

A performance investigation of mobile cellular radio systems using MRC in K-distribution multipath fading environment

“…Once the vector Ai in Eq. 2 is assumed to follow K-distribution, its characteristic probability density function which is known to be generalized chidistribution can be written as [12], [13], [14], [15], Therefore, the probability density function of the random variable ai == u; may readily be found as, is written by the product of a non-negative random variable ui(i == 0,1,2, ... , L) and a zero-mean complex Gaussian random vector Zi (i == 0, 1, 2, ... , L) as,…”

“…When each interferer has equal mean power (Pi == P) the probability density function of the summation 9 == L~1 Yi may be written by [10], [11], [19], _ gL-l exp(_1js) > pg(g)~f(L)PL ,g~0 (15) In this section, we assume that the fading environment is such that the characteristic PDF of the faded interferers defined in Eq. 2 are strongly correlated [6], i.e., all the interfering signals are modulated by the same non-negative random variables Ui == U (i == 1, 2, 3, ... , L).…”