Structural and Parametric Identification of Knowm Memristors.

Area of Science:

• Solid-state physics and materials science
• Electronic device modeling and simulation
• Non-volatile memory technologies

Background:

• Accurate modeling of memristive devices is crucial for their integration into next-generation electronics.
• Existing models often fail to capture essential characteristics like low-current behavior and cycle-to-cycle variability.
• Knowm memristors serve as a relevant case study due to their unique properties.

Purpose of the Study:

• To propose a novel and generalized identification method for memristive devices.
• To develop a new mathematical model that accounts for low-current effects and cycle-to-cycle variability in memristors.
• To demonstrate the model's capability in capturing complex memristor dynamics, including snapforward/snapback effects and frequency properties.

Main Methods:

• Experimental characterization of Knowm memristors to obtain intrinsic current-voltage (I-V) curves.
• Development of a parametric identification technique based on acquired measurement data.
• Formulation of a new mathematical model where the switching threshold is a function of state variables.

Main Results:

• A novel mathematical model for memristive devices was successfully developed and identified.
• The model accurately represents low-current effects and addresses cycle-to-cycle variability in memristor behavior.
• The model's ability to capture snapforward/snapback effects, frequency dependence, and switching variability was demonstrated.

Conclusions:

• The proposed identification method provides a generalized approach applicable to various memristive elements.
• The developed mathematical model offers enhanced accuracy in describing memristor characteristics, paving the way for more reliable device simulations.
• Further research into the model's limitations and potential extensions is warranted for broader applications.