Quantum-Dot Cellular Automata Based On Trainable Associative Memory Neural Network for Implementing Reconfigurable Logic Gates

CMOS technology has reached its physical scaling limits, power consumption limits, and hence CMOS technology has to be replaced by other emerging technologies. Quantum-Dot Cellular Automata (QCA) has proven to be a better solution since it offers better scaling and low power consumption for processing digital signals. However, due to a lack of optimization support improvements, dynamical QCA paradigms require multi-layer designs and increased computations. Hence in this paper, a reconfigurable logic gate based on QCA paradigms is proposed. The design utilizes training of memory cells associated with QCA paradigms and hence can be switched between AND, OR, and MUX logic circuit designs dynamically. Compared to traditional dynamical QCA, the model is designed on a single-layer substrate with a minimum number of cells and size.