Abstract
A memristor is a passive circuit element that has numerous applications ranging from storage to processing. The attributes of memristor, such as low power, fast switching speed, and non-volatility, make it a promising candidate for computing applications. To deploy the memristors at the circuit level, a versatile, computationally efficient, and generic model is required to analyze the circuit performance. This article presents a computationally inexpensive, generic, and accurate phenomenological model of the memristor. The proposed model predicts an accurate current-voltage ( I - V ) relationship based on the current conduction mechanism. The presented modeling technique can be incorporated into any practical memristor behavior by optimizing the fitting parameters. The model has the capability to optimize its accuracy and computational efficiency by calibrating the model parameters. The results are compared with the characterization data of numerous memristors and noteworthy models of memristors to validate the proposed approach. The results depict that the proposed model is more flexible, computationally efficient, versatile, and generic. It improves the simulation run time up to 24.47\% with the relative root mean squared error of 0.4142\% . The model exhibits remarkable results when compared with titanium dioxide-based devices. The proposed model can be deployed to various applications, such as logic design, memory design, and neuromorphic computing.