A systematic approach to statistical simulation of complex analog integrated circuits

It is well known that during the manufacturing process integrated circuits are subject to process parameter variations which result in the inherent performance fluctuations. Characterization of these fluctuations in behavioral modeling is extremely important if the high yield of the manufactured circuits is to be achieved. The methodology presented in the paper allows one to build behavioral statistical models which give designers insight into the influence of process and mismatch variations on the operation of the complex IC. Successful statistical simulation of an analog phase-locked loop is presented.