Abstract

Analog and mixed signal (AMS) circuits play an important role in system on chip designs. They pose, however, many challenges in the verification of the overall system due to their complex behaviors and expensive consumption of simulation resources. Besides functionality, AMS systems also suffer from stochastic processes such as random noise which exhibits statistical properties. Among many developed verification techniques, runtime verification has been shown to be effective by experimenting finite executions instead of going through the whole state space. In this thesis, we propose a methodology for the verification of AMS designs using functional and statistical runtime verification. Functional runtime verification is used to check the functional behavior of the AMS design. A system of recurrence equation (SRE) is used to model the AMS design and construct a functional property monitor. This functional runtime verification is carried out in an online fashion. Statistical runtime verification is used to verify the statistical properties of the AMS design. Hypothesis test, which is a method to make statistical decisions about rejecting or accepting some statement about the information of a sample, is used to verify the statistical properties. We use Monte Carlo simulation for the hypothesis test and for evaluating its performance. The proposed methodology is applied to a phase lock loop based frequency synthesizer where several functional properties and stochastic noise properties are verified.