Slow relaxation of memristive networks

Memristors are a new breed of passive device which can store memory. For the case of a generic network of memristive components, one can write a linear system of the equations using the nodal analysis. We show that for the case of HP memristors, one can obtain a linear relationship between the internal parameters of the devices representing the memory, implying strong correlation. Due to the non-linear properties of memristors, one can show using general arguments that the relaxation of a purely memristive system is not exponential, but indeed follows a power law. We moreover analyze correlations between memristors in the specific case of a regular lattice of memristors.