Research and Development of Parameter Extraction Approaches for Memristor Models

Area of Science:

• * Materials Science and Engineering
• * Electrical Engineering and Computer Science

Background:

• * Memristors are crucial for advanced computing, enabling neural processors and non-volatile memory.
• * Accurate memristor modeling is essential for circuit design, relying heavily on parameter extraction from experimental data.
• * Compact models are widely used for memristor simulation, but their precision depends on effective parameter extraction.

Purpose of the Study:

• * To review existing parameter extraction methods for memristor compact models.
• * To propose novel parameter extraction algorithms tailored for adaptive memristor models.
• * To validate the effectiveness of the developed algorithms using experimental data.

Main Methods:

• * Comprehensive literature review of current memristor parameter extraction techniques.
• * Development of new algorithms for adaptive compact memristor model parameter extraction.
• * Experimental validation using the current-voltage characteristics of a TiN/Hf_xAl_1-xO_y/HfO₂/TiN vertical memristor structure.

Main Results:

• * Identified limitations in existing parameter extraction methods.
• * Proposed and detailed new algorithms for adaptive compact memristor models.
• * Demonstrated the improved accuracy of the developed methods for memristor simulation.

Conclusions:

• * The proposed parameter extraction algorithms offer enhanced accuracy for adaptive memristor models.
• * These advancements are critical for the reliable design of neural processors and non-volatile memory.
• * The validated methods provide a more precise approach to memristor circuit simulation.