An iterative approach to memristor

The history dependence of conductance variation enables memristive characteristics. Here, we propose an iterative approach to the analytical description of the memristor. The model analytically reproduces the fingerprints of the memristor, e.g., the frequency dependence of the normalized current–voltage loop area, the on–off ratio, and the synaptic weight changes of the biological plasticity. The analytical descriptions of frequency-dependent loop area and on–off ratio also yield estimations of the optimal scanning frequency for enhancing the memory effect. The proposed model matches well with the numerical simulation, experimental observations, and published results. This iterative insight into the history- and frequency-dependent memristive features provides a compact analytical description of the memristor, which should be useful for designing and integrating memristor circuits and neuromorphic computing systems.