A new method for fractional Brownian motion interpolation

This article presents a new method for interpolating long fractional Brownian motion (fBm) sequences called incremental Fourier interpolation (IFI). Instead of computing the interpolated samples directly, as is the case with existing algorithms, IFI computes the first-order increments between the original and interpolated samples. For long sequences, these increments can be computed using the computationally efficient fast Fourier transform. Estimators for the fBm parameters are also incorporated into the algorithm. Simulations are presented for both known and unknown parameter cases that demonstrate the accuracy of IFI even for relatively short length sequences.