Browsing by Author "Wang, CS"

Now showing 1 - 12 of 12

A genetic algorithm approach to InGaP/GaAs HBT parameter extraction and RF characterization
(2003-04-01)
Li, YM
;
Cho, YY
;
Wang, CS
;
Huang, KY
;
電子工程學系及電子研究所
In this paper, a computational intelligence technique is applied to extract and simulate the stationary and high-frequency properties of heterojunction bipolar transistors (HBTs). A set of HBT circuit equations formulated with the Gummel-Poon model in time domain is solved with (1) the waveform relaxation (WR), (2) monotone iterative (MI) method, and (3) genetic algorithm (GA) with floating-point operators. The coupled nonlinear equations are decoupled and solved with the WR and MI methods in time domain, and the results obtained are used for the optimization of the characteristics with the GA method. The iteration can be terminated when the final convergent global solution is obtained. The time domain result is used in analyzing the property of the output third-order intercept point (OIP3) with the fast Fourier transform (FFT). Compared with the SPICE result, our simulation results demonstrate that this method is accurate and stable in high frequency simulation. This approach has practical applications in HBT characterization and radio frequency (RF) circuit optimal design.
A structure-preserving Lanczos-type algorithm with application to control problems
(1997)
Ferng, WR
;
Lin, WW
;
Wang, CS
;
應用數學系
;
Department of Applied Mathematics
A Hamiltonian structure-preserving Lanczos-type method, named the J-Lanczos algorithm, is introduced for solving large sparse Hamiltonian eigenvalue problem which arises in both continuous-time and discrete-time optimal control applications. Shift and invert techniques are incorporated to approximate all stable eigenvalues and the associated invariant subspace. Numerical results for solving high order continuous-time Riccati equation arising from position and velocity control for a string of high speed vechicles are presented.
New triple self-aligned (SA3) split-gate flash cell with T-shaped source coupling
(2005-10-01)
Sung, HC
;
Lei, TF
;
Huang, CM
;
Kao, YC
;
Lin, YT
A new triple self-aligned (SA3) split-gate flash cell with a T-shaped source coupling approach is described in this paper. This novel structure can significantly enhance coupling capacitance between the Source and floating gate without increasing cell size. The enhancement can be simply modulated by an oxide-etching step. This new structure can be applied to program voltage reduction and cell size scaling. For program voltage reduction, the maximum program voltage of the new cell can be reduced from 7.4 to 6AV, which is characterized by a newly developed methodology for program vs disturb window characterization. For cell size scaling, comparable sort-1 and sort-2 yields are demonstrated using the new cell with a shorter floating length and a shallower source junction. To understand the relationship between source coupling ratio (SCR) and the program/erase mechanism, an insightful discussion on the program and erase mechanisms for our split-gate flash cell is also shown in this paper.
Nonequivalence transformation of lambda-matrix eigenproblems and model embedding approach to model tuning
(2001-01-01)
Ferng, WR
;
Lin, WW
;
Pierce, DJ
;
Wang, CS
;
應用數學系
In this paper we present some theory for a non-equivalence transformation of matrix eigenvalues for lambda -matrix polynomials. Application of this transformation to eigenvalue embedding for model tuning on a realistic industry problem is illustrated. The new approach has several advantages such as flexibility, efficiency, and structure-preservation. A numerical experiment on a pseudosimulation data set from The Boeing Company is retorted. Copyright (C) 2001 John Wiley & Sons, Ltd.
Novel program versus disturb window characterization for split-gate flash cell
(2005-03-01)
Sung, HC
;
Lei, TF
;
Hsu, TH
;
Kao, YC
;
Lin, YT
In this letter, a new methodology for program versus disturb window characterization on split gate Flash cell is presented for the first time. The window can be graphically illustrated in V-wI (word-line)-V-ss (source) domain under a given program current. This method can help us understand quantitatively how the window shifts versus bias conditions and find the optimal program condition. The condition obtained by this method can have the largest tolerance for program, bias variations. This methodology was successfully implemented in 0.18-mum triple self-aligned (SA3) split-gate cell characterization to provide program condition for 32 M products.
Novel single-poly EEPROM with damascene control-gate structure
(2005-10-01)
Sung, HC
;
Lei, TF
;
Hsu, TH
;
Wang, SW
;
Kao, YC
A novel single-poly EEPROM using damascene control gate (CG) structure is presented in this letter. The CG is tungsten (W) line made by a damascene process, and intergate dielectric is Al2O3 grown by atomic layer deposition (ALD). The program and erase mechanism is the same as the one for traditional stacked-gate cell, which uses the channel hot electron injection for programming and Fowler-Nordheim tunneling for channel erasing. With the high dielectric constant (K) property of Al2O3, we can perform the program and erase function with a voltage less than 6.5 V, which can be handled by 3.3 V devices instead of traditional high voltage devices. In the process compatibility aspect, this new cell needs only two extra masking steps over the standard CMOS process, and the high-kappa material is deposited in the back-end metallization steps without the contamination concerns on the front-end process. Therefore, this new technology is suitable for embedded application. In this letter, the good cell performance is demonstrated; such as, fast programming/erasing, good endurance and data retention.
Numerical algorithms for undamped gyroscopic systems
(1999-01-01)
Ferng, WR
;
Lin, WW
;
Wang, CS
;
應用數學系
;
Department of Applied Mathematics
The solutions of a gyroscopic vibrating system oscillating about an equilibrium position, with no external applied forces and no damping forces, are completely determined by the quadratic eigenvalue problem (-lambda(i)(2)M + lambda(i)G + K)x(i) = 0, for i = 1, ..., 2n, where M, G, and K are real n x n matrices, and M is symmetric positive definite (denoted by M > 0), G is skew symmetric, and either K > 0 or -K > 0. Gyroscopic systemin motion about a stable equilibrium position (with -K > 0) are well understood. Two Lanczos-type algorithms, the pseudo skew symmetric Lanczos algorithm and the J-Lanczos algorithm, are studied for computing some extreme eigenpairs for solving gyroscopic systems in motion about an unstable equilibrium position (with K > 0). Shift and invert strategies, error bounds, implementation issues, and numerical results for both algorithms are presented in details. (C) 1998 Elsevier Science Ltd. All rights reserved.
On computing stable Lagrangian subspaces of Hamiltonian matrices and symplectic pencils
(1997-07-01)
Lin, WW
;
Wang, CS
;
應用數學系
;
Department of Applied Mathematics
This paper presents algorithms far computing stable Lagrangian invariant subspaces of a Hamiltonian matrix and a symplectic pencil, respectively, having purely imaginary and unimodular eigenvalues. The problems often arise in solving continuous- or discrete-time H-infinity-optimal control, linear-quadratic control and filtering theory, etc. The main approach of our algorithms is to determine an isotropic Jordan subbasis corresponding to purely imaginary (unimodular) eigenvalues by using the associated Jordan basis of the square of the Hamiltonian matrix (the S + S-1-transformation of. the symplectic pencil). The algorithms preserve structures and are numerically efficient and reliable in that they employ only orthogonal transformations in the continuous case.
On computing stable Lagrangian subspaces of Hamiltonian matrices and symplectic pencils
(1997-07-01)
Lin, WW
;
Wang, CS
;
應用數學系
;
Department of Applied Mathematics
This paper presents algorithms far computing stable Lagrangian invariant subspaces of a Hamiltonian matrix and a symplectic pencil, respectively, having purely imaginary and unimodular eigenvalues. The problems often arise in solving continuous- or discrete-time H-infinity-optimal control, linear-quadratic control and filtering theory, etc. The main approach of our algorithms is to determine an isotropic Jordan subbasis corresponding to purely imaginary (unimodular) eigenvalues by using the associated Jordan basis of the square of the Hamiltonian matrix (the S + S-1-transformation of. the symplectic pencil). The algorithms preserve structures and are numerically efficient and reliable in that they employ only orthogonal transformations in the continuous case.
Plausible origin of electromigration lifetime extrapolation difference between wafer level isothermal test and package level constant current test
(2004)
Wang, CS
;
Chen, MJ
;
Chang, WC
;
Ke, WS
;
Lee, CF
Electromigration lifetime extrapolated from the wafer level isothermal test on a metal line does not coincide with that from the package level constant current test. The discontinuity of mass flux density in wafer level isothermal test as origin of such difference is identified based on a temperature profile along the metal line. Once this issue is clarified in advance, improved correlation between both tests can be achieved and as a result, wafer level isothermal test is expected to replace the conventional time-consuming package level constant current test.
The shift-inverted J-Lanczos algorithm for the numerical solutions of large sparse algebraic Riccati equations
(1997-05-01)
Ferng, WR
;
Lin, WW
;
Wang, CS
;
應用數學系
;
Department of Applied Mathematics
The goal of solving an algebraic Riccati equation is to find the stable invariant subspace corresponding to all the eigenvalues lying in the open left-half plane. The purpose of this paper is to propose a structure-preserving Lanczos-type algorithm incorporated with shift and invert techniques, named shift-inverted J-Lanczos algorithm, for computing the stable invariant subspace for large sparse Hamiltonian matrices. The algorithm is based on the J-tridiagonalization procedure of a Hamiltonian matrix using symplectic similarity transformations. We give a detailed analysis on the convergence behavior of the J-Lanczos algorithm and present error bound analysis and Paige-type theorem. Numerical results for the proposed algorithm applied to a practical example arising from the position and velocity control for a string of high-speed vehicles are reported.
The shift-inverted J-Lanczos algorithm for the numerical solutions of large sparse algebraic Riccati equations (vol 33, pg 23, 1997)
(1999-08-01)
Ferng, WR
;
Lin, WW
;
Wang, CS
;
應用數學系
;
Department of Applied Mathematics