Multivariate Scale Invariance
Scale invariance, or scaling, is now recognized as an ubiquitous property in a variety of realworld applications which are very different in nature. The sonamed scale invariance paradigm is based on the assumption that temporal dynamics in data are not driven by one, or a few, representative time scales, but by a large continuum of them. Selfsimilar stochastic processes provide the basal mathematical framework for the modeling of scaling phenomena. In essence, selfsimilarity states that a signal cannot be distinguished from any of its dilated copies. Fractional Brownian motion (fBm)  the only Gaussian, selfsimilar, stationary increment process – has massively been used as a reference process in the modeling of scaling properties in univariate realworld signals.
Notwithstanding its theoretical and practical importance, fBm falls short of providing an encompassing modeling framework for scaling because most modern contexts of application involve the recording of multivariate time series that hence need to be jointly analyzed. The construction of a comprehensive multivariate estimation paradigm is still an open problem in the literature. The sonamed Operator fractional Brownian motion (OfBm) is a natural extension of fBm. It was recently been defined as the only Gaussian, multivariate selfsimilar process with stationary increment.
In this project, our contribution is threefold.
Related Publications

Nonlinear Regression for Bivariate Sefsimilarity Identification  Application to Anomaly Detection in Internet Traffic Based on a Joint Scaling Analysis of Packet and Byte Counts,International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2016.

NonLinear Wavelet Regression and Branch and Bound Optimization for the Full Identification of Bivariate Operator Fractional Brownian Motion,IEEE Transactions on Signal Processing, 2016.