Recursive Threshold Logic-A Bioinspired Reconfigurable Dynamic Logic System With Crossbar Arrays.

Area of Science:

• Neuromorphic Engineering
• Computer Science

Background:

• Neuron firing principles and weighted charge accumulation inspire computational models.
• Biological neurons exhibit dynamic feedback and feedforward responses.
• Threshold logic design applies threshold functions to weighted input sums.

Purpose of the Study:

• To present a recursive threshold logic unit emulating Boolean expressions time-sequentially.
• To implement Boolean expressions using memristive crossbars and CMOS comparators.
• To leverage reliable memristor programming in 1T1R configurations for larger crossbar sizes.

Main Methods:

• Utilizes output feedback from threshold logic gates for recursive emulation.
• Employs analog resistive dividers in memristive crossbars for Boolean expression implementation.
• Designs a hard-threshold function with a CMOS comparator for OR and AND operations.
• Employs 1T1R memristive crossbar configuration to suppress sneak path currents.
• Tunes reference threshold voltage for AND/OR logic with constant R_ON memristor resistance.

Main Results:

• Achieves reliable programming of memristors, enabling larger crossbar sizes and more inputs.
• Demonstrates circuit resilience to memristor variability and aging.
• Shows the proposed recursive logic uses fewer cross-points.
• Exhibits lower power dissipation compared to existing memristive and CMOS logic.

Conclusions:

• The recursive threshold logic unit offers an efficient and resilient approach to Boolean expression emulation.
• Memristive crossbars with 1T1R configuration provide a scalable solution for complex logic.
• The proposed design presents advantages in terms of component count and power efficiency.