feed-gap height (g) suggests that the feed-gap height of 1 mm can
yield the widest BW. It is shown that the proposed antenna can
operate in lower band from 1525 to 2440 MHz continuously,
which may find applications in many wireless communication
systems such as GPS, DCS, PCS, UMTS, and IMT-2000.
ACKNOWLEDGMENTS
This work was supported by the National Science Council of the
Republic of China under grant NSC 94 –2213-E-006 – 045.
REFERENCES
1. S. Dey, K.A. Jose, C.K. Aanandan, P. Mohanan, and K.G. Nair,
Wideband printed dipole antenna, Microwave Opt Technol Lett 4
(1991), 417– 419.
2. G. Kumar and K.P. Ray, Broadband microstrip antennas, Artech
House, Norwood, 2003.
3. K.L. Wong, Compact and broadband microstrip antennas, Wiley, New
York, 2002.
4. H.M. Chen, Y.F. Lin, C.C. Kuo, and K.C. Huang, A compact dual-
band microstrip-fed monopole antenna, IEEE Antennas Propag Soc Int
Symp 2 (2001), 124 –127.
5. Y.F. Lin, H.D. Chen, and H.M. Chen, A dual-band printed L-shaped
monopole for WLAN applications, Microwave Opt Technol Lett 37
(2003), 214 –216.
6. C.Y. Pan, C.H. Huang, and T.S. Horng, A novel printed monopole
antenna with a square conductor-backed parasitic plane for dual-band
WLAN applications, IEEE Antennas Propag Soc Int Symp 1 (2004),
261–264.
7. W.C. Liu, Compact microstrip-line-fed ring monopole antenna with
tuning strip for 5 GHz WLAN operation, Electron Lett 41 (2005),
831– 832.
8. H.D. Chen, J.S. Chen, and Y.T. Cheng, Modified inverted-L monopole
antenna for 2.4/5 GHz dual-band operations, Electron Lett 39 (2003),
1567–1568.
9. J.Y. Jan and T.H. Wu, Planar dual-band monopole antenna designs
embedded with an open square slot, IEEE Antennas Propag Soc Int
Symp 4 (2003), 142–145.
10. Ansoft High-Frequency Structure Simulator (HFSS), Ansoft Corpora-
tion.
11. J. Liang, C.C. Chiau, X. Chen, and C.G. Parini, Analysis and design of
UWB disc monopole antennas, IEE Seminar Ultra Wideband Commun
Technol Syst Des (2004), 103–106.
© 2007 Wiley Periodicals, Inc.
SUPER-HYPER SINGULARITY
TREATMENT FOR SOLVING 3D
ELECTRIC FIELD INTEGRAL
EQUATIONS
Mei Song Tong and Weng Cho Chew
Center for Computational Electromagnetics and Electromagnetics,
Laboratory, Department of Electrical and Computer Engineering,
University of Illinois at Urbana-Champaign, Urbana, IL 61801
Received 20 October 2006
ABSTRACT: The super-hyper singularity treatment is developed for
solving three-dimensional (3D) electric field integral equations (EFIE).
EFIE usually takes two forms: one of which includes a super-hyper sin-
gular kernel generated by the double gradient of the Green’s function.
So far, there is no way to evaluate these super-hyper singular integrals
needed for constructing the pertinent matrix equation. We apply the se-
ries expansion of the Green’s function to the super-hyper singular ker-
nel and derive closed-form expressions for their evaluations in the
Cauchy principal-value sense. The derivation is based on the constant
current approximation over a flat triangle patch, but it can be extended
easily to higher-order approximations of the current. The scheme can be
used to accurately calculate the near and self interaction terms in the
impedance matrix for the method of moments (MoM), Nystro ¨m method
or boundary element method (BEM). © 2007 Wiley Periodicals, Inc.
Microwave Opt Technol Lett 49: 1383–1388, 2007; Published online in
Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/mop.
22443
Key words: singularity treatment; electric field integral equation;
Cauchy principal value
1. INTRODUCTION
The widely-used electric field integral equation (EFIE) in electro-
magnetic (EM) scattering and radiation problems assumes two
forms with different kernels: one includes the gradient of the
Green’s function times the divergence of current (first form); the
other includes the double gradient of the Green’s function times
the current (second form). These two forms are equivalent to each
other, demonstrable using integration by parts.
The first form has a 1/R
2
singularity in the kernel for self-terms,
where R is the distance between the source point and the obser-
vation point. This singularity, together with 1/R singularity in the
Green’s function in the kernel, can be efficiently handled by
singularity subtraction technique and Duffy’s method [1–7].
Therefore, the first form is more widely used, and because of the
divergence of current in the kernel, a linear approximation for the
current expansion is often used.
The second form is free of the divergence of current. This form
allows constant current approximation over a discretized patch
with great simplicity. It also provides a simple interface with fast
algorithms, such as the fast multipole algorithm (FMA) [8], be-
cause it associates an unknown with a point in space, where this
locality improves FMAs efficiency. However, this form includes a
1/R
3
super-hyper singularity in the kernel generated by the double
gradient of the Green’s function. Due to the higher degree of the
singularity, there is no direct method to handle such a kernel. One
usually changes the kernel back to the first form to evaluate the
self-terms and uses the second form just for far-interaction terms in
the Nystro ¨ m method [9 –11]. This treatment is inefficient and is not
convenient when working with FMA.
Under some circumstances, it is obviously very desirable to
apply the second form without changing the kernel, e.g., in de-
signing a localized Nystro ¨ m method. The problem, however, is the
Figure 8 Measured peak gain of the planar monopole antenna: (a)
1.5–2.7 GHz; (b) 4 –5.2 GHz; W = 23 mm, D = 5 mm, S = 0 mm, g =
1 mm
DOI 10.1002/mop MICROWAVE AND OPTICAL TECHNOLOGY LETTERS / Vol. 49, No. 6, June 2007 1383