Analytical and numerical procedures for fast periodic steady-state and transient analyses of nonlinear circuits

Abstract

Effective and efficient simulation and verification techniques are always highly demanded in the electronic design automation (EDA) community. However, existing modeling and simulation approaches could no longer fulfill this growing demand due to the presence of large-scale distributed components and strongly nonlinear devices. In this thesis, two novel techniques are presented for fast simulation and verification of analog and radio-frequency (RF) circuits. Emphasis is placed on developing an analytical approach for periodic steady-state (PSS) analysis and an effective model order reduction (MOR) technique for nonlinear, especially highly nonlinear, systems arising in analog/RF circuit applications.

The first approach, named autonomous Volterra (AV), achieves efficient PSS analysis of nonlinear circuits. With elegant analytic forms and availability of efficient solvers, AV constitutes a competitive steady-state algorithm besides the two mainstream PSS algorithms, namely, shooting Newton (SN) and harmonic balance (HB). Nonlinear systems are first captured in nonlinear differential algebraic equations (DAEs), followed by expansion into linear Volterra subsystems. A key step of steady-state analysis lies in modeling each Volterra subsystem with autonomous nonlinear inputs. The PSS solution of these subsystems then proceeds with a series of Sylvester equation solves, completely avoiding the guesses of initial condition and time stepping as in SN, as well as the uncertain length of Fourier series as in HB. Error control in AV is also straightforward by monitoring the norms of the Sylvester equation solutions. It is further demonstrated that AV is readily parallelizable with superior scalability towards large-scale problems.

Besides PSS analysis, another important step in analog/RF design is transient as well as general time simulation. To this end, the second part of this thesis features a tensorbased nonlinear model order reduction (TNMOR) algorithm that allows efficient simulation of nonlinear circuits via the emerging techniques utilizing tensors, namely, a multidimensional generalization of matrices. Unlike existing nonlinear model order reduction (NMOR) methods, high-order nonlinearities are captured using tensors in TNMOR, followed by decomposition and reduction to a compact tensor-based reduced-order model (ROM). Consequently, TNMOR completely avoids the dense reduced-order system matrices, which in turn permits faster simulation and less memory requirement. Numerical experiments on transient and PSS analyses confirm the superior accuracy and efficiency of TNMOR, particularly in highly nonlinear scenarios.