Robust Optimization of an RFIC Isolation Problem Under Uncertainties

Modern electronics systems involved in communication and identification impose demanding constraints on both reliability and robustness of components. On the one hand, it results from the influence of manufacturing tolerances within the continuous down-scaling process into the output characteristics of electronic devices. On the other hand, the increasing integration process of various systems on a single die force a circuit designer to make some trade-offs in preventing interference issues and in compensating coupling effects. Thus, constraints in terms of statistical moments have come in a natural way into optimization formulations of electronics products under uncertainties. Therefore, in this paper, for the careful assessment of the propagation of uncertainties through a model of a device a type of Stochastic Collocation Method (SCM) with Polynomial Chaos (PC) was used. In this way a response surface model can be included in a stochastic, constrained optimization problem. We have illustrated our methodology on a Radio Frequency Integrated Circuit (RFIC) isolation problem. Achieved results for the optimization confirmed efficiency and robustness of the proposed methodology.