Statistical Simulation and Modeling of Nano-scale CMOS VCO Using Artificial Neural Network

The variation of intra-die process parameters play a significant role in determining the yield of an analog/RF circuit. This paper presents statistical results demonstrating the effect of variations of process parameters on a nano-scale CMOS voltage controlled oscillator circuit. A statistical model relating the process parameter variations and the performance variations has been constructed using artificial neural network. The constructed model shows accuracy similar to that obtained though Monte Carlo analysis technique, however, consuming much less time.