An Adjoint Network Approach for RLCG Interconnect Model Order Reductions

Chia-Chi CHU; Herng-Jer LEE; Ming-Hong LAI; Wu-Shiung FENG

doi:10.1093/ietfec/e89-a.2.439

An Adjoint Network Approach for RLCG Interconnect Model Order Reductions

Chia-Chi CHU, Herng-Jer LEE, Ming-Hong LAI, Wu-Shiung FENG

Full Text Views

0

Cite this

Summary :

This work proposes a new method for RLCG interconnect model-order reductions in consideration with the adjoint network. Relationships between an original MNA network and its corresponding adjoint MNA network will be explored first. It will be shown that the congruence transformation matrix used in the one-sided projection can be constructed by using the bi-orthogonal bases developed from the Lanczos-type algorithms. In particular, if the multi-port driving-point impedance of RLCG interconnect circuits is the main concern, the transfer functions and system moments of the adjoint network can be directly calculated from those of the original RLCG interconnect network by exploring symmetric properties of the MNA formulation. Therefore, the cost of constructing the congruence transformation matrix can be simplified by up to 50% of the previous methods. Comparative studies among various standard methods and the proposed methods are also investigated. Experimental results on large-scale RLCG interconnect circuits will demonstrate the accuracy and the efficiency of the proposed method.

Publication: IEICE TRANSACTIONS on Fundamentals Vol.E89-A No.2 pp.439-447

Publication Date: 2006/02/01

Publicized

Online ISSN: 1745-1337

DOI: 10.1093/ietfec/e89-a.2.439

Type of Manuscript: Special Section PAPER (Special Section on Analog Circuit Techniques and Related Topics)

Category

Cite this

Copy

Chia-Chi CHU, Herng-Jer LEE, Ming-Hong LAI, Wu-Shiung FENG, "An Adjoint Network Approach for RLCG Interconnect Model Order Reductions" in IEICE TRANSACTIONS on Fundamentals, vol. E89-A, no. 2, pp. 439-447, February 2006, doi: 10.1093/ietfec/e89-a.2.439.
Abstract: This work proposes a new method for RLCG interconnect model-order reductions in consideration with the adjoint network. Relationships between an original MNA network and its corresponding adjoint MNA network will be explored first. It will be shown that the congruence transformation matrix used in the one-sided projection can be constructed by using the bi-orthogonal bases developed from the Lanczos-type algorithms. In particular, if the multi-port driving-point impedance of RLCG interconnect circuits is the main concern, the transfer functions and system moments of the adjoint network can be directly calculated from those of the original RLCG interconnect network by exploring symmetric properties of the MNA formulation. Therefore, the cost of constructing the congruence transformation matrix can be simplified by up to 50% of the previous methods. Comparative studies among various standard methods and the proposed methods are also investigated. Experimental results on large-scale RLCG interconnect circuits will demonstrate the accuracy and the efficiency of the proposed method.
URL: https://global.ieice.org/en_transactions/fundamentals/10.1093/ietfec/e89-a.2.439/_p

Copy

@ARTICLE{e89-a_2_439,
author={Chia-Chi CHU, Herng-Jer LEE, Ming-Hong LAI, Wu-Shiung FENG, },
journal={IEICE TRANSACTIONS on Fundamentals},
title={An Adjoint Network Approach for RLCG Interconnect Model Order Reductions},
year={2006},
volume={E89-A},
number={2},
pages={439-447},
abstract={This work proposes a new method for RLCG interconnect model-order reductions in consideration with the adjoint network. Relationships between an original MNA network and its corresponding adjoint MNA network will be explored first. It will be shown that the congruence transformation matrix used in the one-sided projection can be constructed by using the bi-orthogonal bases developed from the Lanczos-type algorithms. In particular, if the multi-port driving-point impedance of RLCG interconnect circuits is the main concern, the transfer functions and system moments of the adjoint network can be directly calculated from those of the original RLCG interconnect network by exploring symmetric properties of the MNA formulation. Therefore, the cost of constructing the congruence transformation matrix can be simplified by up to 50% of the previous methods. Comparative studies among various standard methods and the proposed methods are also investigated. Experimental results on large-scale RLCG interconnect circuits will demonstrate the accuracy and the efficiency of the proposed method.},
keywords={},
doi={10.1093/ietfec/e89-a.2.439},
ISSN={1745-1337},
month={February},}

Copy

TY - JOUR
TI - An Adjoint Network Approach for RLCG Interconnect Model Order Reductions
T2 - IEICE TRANSACTIONS on Fundamentals
SP - 439
EP - 447
AU - Chia-Chi CHU
AU - Herng-Jer LEE
AU - Ming-Hong LAI
AU - Wu-Shiung FENG
PY - 2006
DO - 10.1093/ietfec/e89-a.2.439
JO - IEICE TRANSACTIONS on Fundamentals
SN - 1745-1337
VL - E89-A
IS - 2
JA - IEICE TRANSACTIONS on Fundamentals
Y1 - February 2006
AB - This work proposes a new method for RLCG interconnect model-order reductions in consideration with the adjoint network. Relationships between an original MNA network and its corresponding adjoint MNA network will be explored first. It will be shown that the congruence transformation matrix used in the one-sided projection can be constructed by using the bi-orthogonal bases developed from the Lanczos-type algorithms. In particular, if the multi-port driving-point impedance of RLCG interconnect circuits is the main concern, the transfer functions and system moments of the adjoint network can be directly calculated from those of the original RLCG interconnect network by exploring symmetric properties of the MNA formulation. Therefore, the cost of constructing the congruence transformation matrix can be simplified by up to 50% of the previous methods. Comparative studies among various standard methods and the proposed methods are also investigated. Experimental results on large-scale RLCG interconnect circuits will demonstrate the accuracy and the efficiency of the proposed method.
ER -