Oscillatory fractional Brownian motion (Q385587)

The authors ``introduce oscillatory analogues of fractional Brownian motion [(fBm)], subfractional Brownian motion [(sfBm)] and other related long range dependent Gaussian processes.'' According to them, the oscillatory fractional Brownian motion (ofBm) is a centered Gaussian process \(\xi ^{H}\), with parameter \(H\in (1/2,1)\) and covariance function \[ \operatorname{E}\xi _{s}^{H}\xi _{t}^{H}=\frac{1}{2}\left( s^{2H}h_{s}+t^{2H}h_{t}-\left| s-t\right| ^{2H}h_{\left| s-t\right| }\right) ,\qquad s,t\geq 0, \] and the oscillatory sub-fractional Brownian motion (osfBm) is a centered Gaussian process \(\zeta ^{H}\), with parameter \(H\in (1/2,1)\) and covariance function \[ \operatorname{E}\zeta _{s}^{H}\zeta _{t}^{H}=s^{2H}h_{s}+t^{2H}h_{t}-\frac{1}{2}\left[ \left( s+t\right) ^{2H}h_{s+t}+\left| s-t\right| ^{2H}h_{\left| s-t\right| }\right] ,\qquad s,t\geq 0, \] where \((h_{t})_{t>0}\) is a periodic function in logarithmic scale, i.e., \( h_{t}=h_{at}\), \(t>0\), for some constant \(0<a<1\). The function \(h\) oscillates between two positive values, slower as \(t\rightarrow \infty \) and faster as \( t\rightarrow 0\). The processes fBm, sfBm and other related ones have been obtained from occupation time fluctuation limits of systems of particles, with or without branching, performing symmetric stable processes in the Euclidean space. The present authors show that the oscillatory analogues are obtained when instead of systems of stable processes one considers systems of hierarchical random walks. They also study the properties of these oscillatory processes in depth.