Energy-efficient devices for intelligent and optimized nanoelectronics parameters

The rapid pace of nanoelectronic systems' development creates a need for efficient, high-performance circuit designs that minimize energy consumption, propagation delay, and device reliability issues. However, it is challenging to optimize these parameters by using conventional optimization techniques, which are normally performed by manual tuning and fixed design rules. In this paper, an intelligent optimization framework for optimizing Carbon Nanotube Field-Effect Transistor (CNTFET) nanoelectronic devices is presented. A new Feed-Forward Polar Fox Network (FFPFN) has been created for characterizing the nonlinear relationship between CNTFET parameters. The prediction ability of this neural network is utilized to optimize CNTFET parameters by using the Polar Fox Optimization Algorithm (PFOA) to explore the high-dimensional parameter space for optimal configurations that result in minimum power consumption, propagation delay, and maximum energy efficiency. The suggested framework is implemented by optimizing CNTFET-based 8-input priority encoder in MATLAB using the Stanford CNTFET model and achieves significant improvements in terms of power consumption, propagation delay, and power-delay product in comparison to conventional CNTFET designs. The suggested neural network model predicts with 99.98% accuracy and reduces the value of RMSE to 0.08%, thus proving that it is efficient in comparison to existing techniques. As shown in the results, there are considerable improvements in terms of energy efficiency and operational performance, which emphasize the framework's ability to facilitate intelligent high-performance, energy-conscious nanoelectronic circuits.