Characterization of a next generation step-and-scan system
Timothy J. Wiltshire, Joseph P. Kirk, et al.
SPIE Advanced Lithography 1998
It has been known for some time that the trapezoidal rule Tnf = 1 2f(0) + f(1) + ... + f(n - 1) + 1 2f(n) is the best quadrature formula in the sense of Sard for the space W1,p, all functions such that f′ ε{lunate} Lp. In other words, the norm of the error functional Ef = ∝0n f(x) dx - ∑k = 0n λkf(k) in W1,p is uniquely minimized by the trapezoidal sum. This paper deals with quadrature formulas of the form ∑k = 0n ∑lε{lunate}J cklf(l)(k) where J is some subset of {0, 1,..., m - 1}. For certain index sets J we identify the best quadrature formula for the space Wm,p, all functions such that f(m) ε{lunate} Lp. As a result, we show that the Euler-Maclaurin quadrature formula Tnf + ∑ o<2v≤m B2v (2v)! (f (2v-1)(0) - f (2v-1) (n)) is the best quadrature formula of the above form with J = {0, 1, 3,..., ≤m - 1} for the space Wm,p, providing m is an odd integer. © 1974.
Timothy J. Wiltshire, Joseph P. Kirk, et al.
SPIE Advanced Lithography 1998
David W. Jacobs, Daphna Weinshall, et al.
IEEE Transactions on Pattern Analysis and Machine Intelligence
Paul J. Steinhardt, P. Chaudhari
Journal of Computational Physics
Simeon Furrer, Dirk Dahlhaus
ISIT 2005