Detection and classification of signals in impulsive noise modeled as an alpha-stable process

Studies the structure of the probability density function of random variables which are formed as the sum of two or more independent, symmetric stable random variables of different characteristic exponents. The authors present an asymptotic series expansion valid for large argument. Possible applications of the theory can be found in the problem of detection of impulsive stochastic transients superimposed over a background of Gaussian noise