Superdiffusion and time-dependent long-range interactions on networks.

This work addresses the motion of a random walker on a discrete structure with the presence of long-range interactions, the strength of which can change in time. No restriction is made on the nature of the substrate, which may range from ordered lattice to complex network. The presence of long-range interactions on regular lattices with nearest-neighbor hops enhances the speed of the diffusive motion, but the dependence of the mean square displacement (MSD) traveled by the walker has been found to still increase linearly with time. In contrast to this, our results for a model where strength of long range interactions increases with time reveals a non linear dependence of MSD with respect to time. The time duration of such a dependence, as well as the value of power law exponent, depend on the way the interactions change in time. The model on the cycle graph is amenable to exact analytical treatment within the Markov chain framework. Numerical simulations are in excellent agreement with the theoretical predictions, and can be easily extended to complex structures where analytical approaches are not possible.