Title
MARS: Matching-Driven Analog Sizing
Abstract
This paper presents a new approach for automatic computation of matching constraints for analog sizing. The method automatically analyzes a circuit netlist to generate sizing constraints. These sizing constraints are considered during a subsequent numerical optimization. It is the first method that computes symmetry constraints for sizing, based on the hierarchical structure of the circuit and a qualitative signal flow. As a unique feature, the method is capable of handling multiple signal and feedback paths, leading to multiple symmetry axes. It can be applied to linear and nonlinear circuits and any available sizing algorithm. Experimental results show that the runtime of analog sizing, as well as of result quality, are greatly improved.
Year
DOI
Venue
2012
10.1109/TCAD.2012.2190069
Computer-Aided Design of Integrated Circuits and Systems, IEEE Transactions
Keywords
DocType
Volume
analogue integrated circuits,feedback,optimisation,MARS,analog integrated circuit,feedback path,linear circuit,matching constraint automatic computation,matching-driven analog sizing,multiple signal handling,multiple symmetry axes,nonlinear circuit,numerical optimization,qualitative signal flow,symmetry sizing constraint generation,Analog integrated circuits,constraint generation,symmetry detection,transistor sizing
Journal
31
Issue
ISSN
Citations 
8
0278-0070
9
PageRank 
References 
Authors
0.74
11
2
Name
Order
Citations
PageRank
Michael Eick190.74
Helmut E. Graeb226936.22