We propose a hierarchical mixed signal design methodology based on the principles of Platform-Based Design (PBD). The methodology is a meet-in-the-middle approach where design components are modeled bottom-up at various abstraction levels and performance constraints are mapped top-down to select among the available components the ones that best meet the constraints. The design methodology can seamlessly operate on both analog and digital designs, thus dealing with mixed signal designs in a consistent way. We demonstrate the effectiveness of the approach optimizing an 80 MS/s 14 bit pipelined Analog-to-Digital Converter (ADC) including digital calibration, yielding 64 % power reduction compared to the original hand optimized design.

Behavioral-level performance modeling of analog and mixed-signal systems using support vector machines This paper presents a novel behavioral-level analog and mixedsignal (AMS) system performance modeling methodology using support vector machines (SVM). The method relies on linearly graded sub-spaces to model complex multi-dimensional performance spaces. A detailed evaluation of the method has been carried out for the purpose of potential use for AMS synthesis. The method has been applied to a complex nonideal 2 nd order Sigma-Delta modulator (SDM) and results show good accuracy of performance modeling and numerical efficiency. 1.

analog circuit synthesis. It usually consist of two steps: feasibility design space identification and performance macromodels generation. A feasibility design space is defined as a multidimensional space in which every design satisfies all the design constraints. The minimum set of constraints is the one that ensures the correct functionality of the given circuit topology. Performance macromodels are only constructed and thereby valid in the functionally correct design space. Support vector machines (SVMs) are used as classifier to identify the feasible design space of analog circuits. A kernel is an integral part of the SVM and contributes in obtaining an optimized and accurate classifier. The most commonly used kernels are Radial Basis Function (RBF), polynomial, spline, multilayer perceptron. In this paper, some new kernels and some other kernels composed through modifications on the some of the standard kernels, are explored. The classifiers using these kernel functions have been tested on different analog circuits in order to identify the feasible design space. HSPICE has been used for generation of learning data. Least Square SVM toolbox interfaced with MATLAB was used for classification. We found that use of modified kernels improves classification accuracy as well as shortens classifier generation time.

Design tool solutions for mixed-signal/RF circuit design