Fractal and chaotic processes provide models for a remarkable range of natural and artificial phenomena. Random stable signals are modeled as solutions of stochastic differential equations driven by fractional Brownian motion. This talk is concerned with statistical methods of estimation of parameters of random stable signals.