Identification of the Hurst index of a step fractional Brownian motion (Q5933671)

The paper proposes the construction of a step fractional Brownian motion by using wavelets. More explicitly, the authors start from the father and the mother of the Lemarié-Meyer wavelet basis, using the mother, they define a family of functions \(\Phi_{\lambda}(t,t)\) which are localised in the vicinity of \(t=\lambda\), and loosely speaking, they define the step fractional Brownian motion as a linear combination of i.i.d. standard Gaussian random variables with \(\Phi_{\lambda}(t,\lambda)\) as coefficients. Some properties are displayed, and the problem of identifying the scale function is considered. The authors claim that they are pioneers of the introduction of multifractional Brownian motion, but this is not quite exact. Well before them Mandelbrot introduced this concept in his studies on mathematical finance.