Professional Documents
Culture Documents
14, 2015
Abstract—In this letter, a structure of a small ultra-wideband performance evaluation [4]. As conventional optimization
(UWB) monopole antenna, its design optimization procedure as methods (such as gradient-based algorithms [5]) require large
well as experimental validation are presented. According to our ap- number of EM analyses to converge (which is extremely
proach, antenna compactness is achieved by means of a meander
expensive in computational terms), simulation-driven design
line for current path enlargement as well as the two parameterized
slits providing additional degrees of freedom that help to ensure is typically realized through parameter sweeps. This is still
good impedance matching. For the sake of reliability, the antenna laborious and does not lead to truly optimum results because
design process (simultaneous adjustment of multiple geometry pa- only a few parameters can be handled this way.
rameters) is carried out using high-fidelity EM analyses. Surro- In this letter, we propose a structure of a compact UWB
gate-based optimization involving an auxiliary coarse-discretiza- monopole with meander line for current path enlargement
tion EM model it utilized to accomplish the design in practical time- as well as two parameterized slits that introduce additional
frame. Penalty function approach allows us to reduce the antenna
footprint (to only mm ) while maintaining acceptable degrees of freedom when controlling the current flow in the
reflection in the UWB frequency range. Experimental validation of ground plane. In order to achieve the smallest possible size,
the design is also provided. we carry out explicit minimization of the antenna footprint
Index Terms—Computer-aided design, simulation-driven
with penalty functions utilized to enforce sufficient matching
design, small antennas, space mapping, UWB monopole. in the UWB band. The novelty of the method lies in its ability
to reduce the antenna footprint while maintaining acceptable
reflection response. Computational efficiency of the design
I. INTRODUCTION process is obtained by means of surrogate-based optimization
(SBO) [6] exploiting space mapping [7] and underlying low-fi-
1536-1225 © 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
BEKASIEWICZ AND KOZIEL: STRUCTURE AND COMPUTATIONALLY EFFICIENT SIMULATION-DRIVEN DESIGN OF COMPACT UWB MONOPOLE ANTENNA 1283
(3)
Fig. 1. Compact UWB monopole antenna—geometry. where is a surrogate model obtained by a suitable cor-
rection of the low-fidelity model (here, obtained through
coarse-mesh EM antenna simulations, cf. Section II), and
The structure is described by 14 independent design param- , , is a sequence of approximated solu-
eters: , whereas tions to the original problem. In this letter, we employ
, , , remain fixed (all dimensions in output and frequency space mapping [7] as the low-fi-
mm). The antenna is designated to operate on TLP-5 dielectric delity model correction techniques, which is well suited
substrate ( , , mm). The sim- for wideband antenna response handling. More specifically,
ulation model of the structure (referred to as the high-fidelity let , where
model ) is implemented in CST Microwave Studio [10] is evaluation of the low-fidelity model at frequency
( mesh cells, average evaluation time 42 min). In (here, ). The frequency-scaled model is de-
the design process (cf. Section III), we also utilize an auxiliary fined as
model ( cells, simulation time 28 s), which is a
coarsely discretized counterpart of . The design objective is (4)
to achieve minimum footprint while ensuring dB
in 3.1 to 10.6 GHz frequency band of interest. where and are scaling parameters obtained to minimize
the misalignment between and at as
III. DESIGN OPTIMIZATION USING SPACE MAPPING (5)
In this section, we discuss the simulation-driven design opti-
mization approach utilized to adjust the geometry parameters of The CPU cost of (5) is negligible as the responses of
the antenna structure proposed in the paper. The aim is to main- , is obtained by inter-
tain reliability of the design process (by utilizing high-fidelity polating the known values of . The response
EM model) and low computational cost (by exploiting SBO). correction is realized with output SM [7], i.e., we have
. This ensures
A. Problem Formulation zero-order consistency [11] between the corrected model and
Our objective is to optimize the high-fidelity antenna at , i.e., .
model , specifically, to minimize the antenna size while Good correlation between the low- and high-fidelity models
keeping its reflection response at the acceptable level (i.e., (both are evaluation using the same EM solver) normally leads
dB in the UWB range). Here, is a vector to a convergence of the process (3) after a few iterations (cf. [7]),
of geometry parameters of the structure. The design task is each requiring just one evaluation of the high-fidelity model.
formulated as This, in conjunction with the fact that is much faster than
, leads to the low overall cost of the design process.
(1)
IV. NUMERICAL RESULTS
where is an objective function that encodes given perfor-
In this section, we summarize the optimization results. We
mance specifications; is the optimum design to be found.
consider two cases: without and with a connector included in the
Here, is defined as
EM antenna model. The latter is necessary to perform adequate
experimental validation reported in Section V.
(2)
A. Optimization Results
where is the antenna footprint, is a penalty factor
(here, ), whereas is a penalty function defined as The initial antenna dimensions are
(i.e., . A
if dB in the UWB band, or corresponds to set of optimized antenna parameters
relative violation of the dB threshold otherwise). Formu-
lation (2) penalizes designs that violate reflection requirements is obtained using methodology of Section III
so that the problem can be solved as an unconstrained one. after only four iterations of (3). The overall area of optimized
1284 IEEE ANTENNAS AND WIRELESS PROPAGATION LETTERS, VOL. 14, 2015
TABLE I
UWB MONOPOLE: INITIAL AND REDUCED VARIABLE BOUNDS
VI. CONCLUSION
Fig. 5. Fabricated prototype of the compact UWB monopole: front view (left) In this letter, a geometrically complex structure of a compact
and back view (right). UWB antenna has been successfully designed and optimized
using SBO scheme. A comparison of the optimization time of
SBO against direct optimization exhibits its superiority in terms
of computational cost. The introduced ground plane modifica-
tion and simultaneous adjustment of all geometry parameters
through computationally efficient simulation-driven optimiza-
tion led to a very compact size of mm . Experimental
validation confirms the correctness of the design approach.
REFERENCES
[1] M. Bod, H. R. Hassani, and M. M. S. Taheri, “Compact UWB printed
Fig. 6. Reflection responses of the antenna with SMA connector: simulation
slot antenna with extra Bluetooth, GSM, and GPS bands,” IEEE An-
(—) and measurement (- - -) results.
tennas Wireless Propag. Lett., vol. 11, pp. 531–534, 2012.
[2] P. Thomas, D. D. Krishna, M. Gopikrishna, U. G. Kalappura, and C.
K. Aanandan, “Compact planar ultra-wideband bevelled monopole for
portable UWB systems,” Electron. Lett., vol. 47, no. 20, pp. 1112–1114,
2011.
[3] J.-F. Li, Q.-X. Chu, Z.-H. Li, and X.-X. Xia, “Compact dual band-
notched UWB MIMO antenna with high isolation,” IEEE Trans. An-
tennas Propag., vol. 61, no. 9, pp. 4759–4766, Sep. 2013.
[4] N. Chahat, M. Zhadobov, R. Sauleau, and K. Ito, “A compact UWB
antenna for on-body applications,” IEEE Trans. Antennas Propag., vol.
59, no. 4, pp. 1123–1131, Apr. 2011.
[5] J. Nocedal and S. J. Wright, Numerical Optimization. New York, NY,
USA: Springer, 2006.
[6] M. B. Yelten, T. Zhu, S. Koziel, P. D. Franzon, and M. B. Steer, “De-
mystifying surrogate modeling for circuits and systems,” IEEE Circuits
Syst. Mag., vol. 12, no. 1, pp. 45–63, 1st Quart., 2012.
[7] S. Koziel, J. W. Bandler, and K. Madsen, “Towards a rigorous formu-
lation of the space mapping technique for engineering design,” in Proc.
Int. Symp. Circuits Syst., 2005, vol. 1, pp. 5605–5608.
[8] J. Liang, C. C. Chiau, X. Chen, and C. G. Parini, “Printed circular disc
monopole antenna for ultra-wideband applications,” Electron. Lett.,
vol. 40, no. 20, pp. 1246–1247, 2004.
[9] T. Li, H. Zhai, G. Li, L. Li, and C. Liang, “Compact UWB band-
notched antenna design using interdigital capacitance loading loop res-
onator,” IEEE Antennas Wireless Propag. Lett., vol. 11, pp. 724–727,
2012.
[10] CST Microwave Studio. ver. 2013, CST AG, Darmstadt, Germany,
2013.
[11] N. M. Alexandrov and R. M. Lewis, “An overview of first-order model
management for engineering optimization,” Optim. Eng., vol. 2, no. 4,
Fig. 7. Comparison of simulated ( ) and measured (––) radiation patterns of
a compact UWB monopole: (a) 5 GHz; (b) 7 GHz; (c) 9 GHz; (d) 10 GHz. pp. 413–430, Dec. 2001.
[12] S. Koziel and A. Bekasiewicz, “Small antenna design using surrogate-
based optimization,” in Proc. IEEE Int. Symp. Antennas Propag., 2014,
pp. 585–586.
antennas with complex topologies. A detailed comparison of [13] T. G. Kolda, R. M. Lewis, and V. Torczon, “Optimization by direct
optimization cost for discussed antenna realizations has been search: New perspectives on some classical and modern methods,”
collected in Table II. SIAM Rev., vol. 45, no. 3, pp. 385–482, 2003.