Highly accurate κ-μ approximation to sum of M independent non-identical Ricean variates

A scheme to approximate the distribution of the sum of M independent, non-identically distributed Ricean random variables by the κ-μ distribution is proposed. To this end, appropriate κ-μ distribution parameters are derived. The summands are assumed to have arbitrary mean powers and arbitrary fading parameters. The differences between exact and approximate distribution curves are almost imperceptible.