Enabling Efficient Analog Synthesis by Coupling Sparse Regression and Polynomial Optimization

摘要: The challenge of equation-based analog synthesis comes from its dual nature: functions producing good least-square fits to SPICE-generated data are non-convex, hence not amenable efficient optimization. In this paper, we leverage recent progress on Semidefinite Programming (SDP) relaxations polynomial (non-convex) Using a general allows for much more accurate fitting SPICE compared the restricted functional forms. Recent SDP techniques convex optimizations powerful but alone still insufficient: even small problems, resulting prohibitively high dimensional.We harness these new tools and realize their promise by introducing novel regression technique that non-convex polynomials with special sparsity structure. We show coupled sparse optimization (CSFO) flow introduce us find high-order while keeping tractable.Using established circuits experiments, demonstrate handling higher-order reduce error 3.6% 10%, average. This translates into dramatic increase in rate constraint satisfaction: 1% violation threshold, success is increased 0% 78%.