774 IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES, VOL. 56, NO. 4, APRIL 2008

Single-Layer High-Order Miniaturized-ElementFrequency-Selective Surfaces

Farhad Bayatpur, Student Member, IEEE, and Kamal Sarabandi, Fellow, IEEE

Abstract—A new miniaturized-element frequency-selective sur-face is presented in this paper. This frequency-selective surface ismade up of a 2-D periodic array of metallic loop and a wire gridof the same period printed on either side of a very thin substrate.Unique features of the new design include localized frequency-se-lective properties, high-order frequency response achieved by asingle substrate, lack of passband harmonics in the frequency re-sponse, and very low frequency response sensitivity to the incidenceangle. High-order frequency response is accomplished through theapplication of a thin substrate that allows considerable couplingsbetween the elements on the two sides of the substrate. The layers’couplings in conjunction with each layer characteristics are de-signed to produce a high- bandpass frequency response, in ad-dition to a transmission zero. It is shown that by inserting vari-able capacitors in the gap between the metallic loops, the centerfrequency of the passband can be tuned over nearly an octave. Inaddition, using a cluster of loops as the unit cell and modifying theparameters of the loops within the cluster, a dual-band character-istic from a single-layer miniaturized-element frequency-selectivesurface can be achieved. A prototype sample of the miniaturized-el-ement frequency-selective surface, whose unit cell can be as smallas � ��, is fabricated to verify the design performance through astandard free-space measurement setup. The transmission charac-teristic of the structure is measured and compared with numericalsimulation results.

Index Terms—Bandpass, bandstop, dual band, miniaturized-el-ement frequency-selective surface, single band.

I. INTRODUCTION

TRADITIONAL frequency-selective surface structures,with resonant unit cells, have been investigated over the

years for a variety of applications. These include bandpass andbandstop spatial filters, absorbers, and artificial electromag-netic bandgap materials. A typical frequency-selective surfaceis a 2-D planar structure consisting of one or more metallicpatterns, each backed by a dielectric substrate. These structuresare usually arranged in a periodic fashion; therefore, theirfrequency response is entirely determined by the geometry ofthe structure in one period called a unit cell. As a result ofresearch on the applications mentioned above, the behavior offrequency-selective surfaces is well understood [1]–[3]. Thefocus of the past studies, however, has been mostly on the band-stop characteristics produced by these surfaces, and structureswith bandpass characteristics have been rarely studied.

Manuscript received August 7, 2007; revised November 28, 2007.The authors are with the Radiation Laboratory, Department of Electrical

Engineering and Computer Science, The University of Michigan at AnnArbor, Ann Arbor, MI, 48109-2122 USA (e-mail: farhadbp@eecs.umich.edu;saraband@eecs.umich.edu).

Digital Object Identifier 10.1109/TMTT.2008.919654

Recently, there has been an interest in design of frequency-se-lective surfaces with unit cell dimensions much smaller than awavelength. In traditional designs, the frequency-selective prop-erties result from mutual interactions of the unit cells. There-fore, to observe a desired frequency-selective behavior, a largenumber of unit cells must be present. Consequently, the overallsize of the surface is electrically large. On the other hand, forsome applications where a low sensitivity with respect to the in-cidence angle of the exciting wave is required or in cases where auniform phase front is difficult to establish, the screen size needsto be small. To address this problem, a new class of frequency-selective surfaces called miniaturized-element frequency-selec-tive surfaces was developed [4]. The new class takes an ap-proach that is different from those of the past designs. In thisapproach, instead of using a resonant structure as the buildingblock of the frequency-selective surface, special unit cells ofsmall dimensions are used. These unit cells act as lumped in-ductive and capacitive elements and are properly arranged sothey couple to the magnetic and electric fields of an incidentwave, respectively. Periodic capacitive patches and an induc-tive wire grid with very small unit cell dimensions ,each printed on a side of a substrate, constitute a parallel L–Ccircuit that behaves as a single-pole bandpass filter [4]. Due to ahigh external coupling coefficient, it has been shown that sucha single-pole design has a poor selectivity and a high insertionloss. To remedy this shortcoming, two such surfaces are cou-pled using an impedance inverter to obtain a high- two-polefilter response with a very low insertion loss. Although the per-formance of this design is high, its fabrication is cumbersome,as it requires two substrates (four printed faces) and a spacer(quarter-wave impedance converter).

Another practical feature of frequency-selective surfacestructures design is the ability to electronically tune the fre-quency response of the structures. In [5]–[13], frequencytunability has been accomplished by altering the substrateconstitutive parameters. Other methods include either changingthe structure geometry using RF microelectromechanicalsystems (MEMS) technology [14]–[17] or manipulating thefrequency-selective surface layers’ reactive characteristics byincorporating tuning elements into the layers’ design [18]–[22].The purpose of this paper is to present a new architecture withcell size dimensions as small as . The new miniatur-ized-element frequency-selective surface exhibits a high-high-order bandpass characteristic, which can be potentiallytuned using varactors mounted on one layer of the structure. Adetailed design procedure for the proposed design is presentedusing an equivalent-circuit model whose parameters are ex-tracted from full-wave analysis. This model is used to design

0018-9480/$25.00 © 2008 IEEE

BAYATPUR AND SARABANDI: SINGLE-LAYER HIGH-ORDER MINIATURIZED-ELEMENT FREQUENCY-SELECTIVE SURFACES 775

Fig. 1. Miniaturized-element frequency-selective surface screen consisting ofthe loop array on one side and wire grid on the other side. Capacitors inter-connect the loops in both �- and �-directions to maintain the symmetry of thestructure.

Fig. 2. Miniaturized-element frequency-selective surface unit cell geometry in-cluding the physical parameters of the loop and wire grid structure that affectthe frequency response.

a prototype at -band, with unit cell dimensions of (4 mm4 mm), which is fabricated and tested using a free-space

measurement setup to show the validity of the design procedureand the performance of the new design.

II. MINIATURIZED-ELEMENT FREQUENCY-SELECTIVE

SURFACE DESIGN SPECIFICATIONS

The proposed miniaturized-element frequency-selective sur-face has two printed layers separated by a very thin dielectricsubstrate. On the top surface, there is a 2-D periodic array ofmetallic square loops, and on the bottom, there is a wire grid.A square portion of the surface containing a few unit cells isshown in Fig. 1. The structure is designed symmetrically withrespect to both - and -axes so that its response is polarizationinsensitive. In Fig. 2, the unit cell static design parameters arealso indicated, including the loop trace width , the spacing be-tween the loops , the grid strip width , the substrate thickness, and the unit cell dimensions and . In the following, a

brief description of the behavior of the layers and their inter-action is presented to give insight into the structure operationmechanism.

The layers’ study starts with the layer containing the wiregrid. It is well known that thin metallic strips supporting axialelectric current excited by an incident wave generate an induc-tive response [3]. The total inductance produced depends on the

strips’ width and length, as well as the field polarization with re-spect to the strips. The wire grid is symmetrical and, therefore,polarization insensitive.

The loop layer has a combination of both inductive and ca-pacitive responses [23]. For an -polarized (see Fig. 1) plane-wave normally incident upon the structure, positive and negativecharge densities are established along the adjacent edges of thesuccessive square loops, storing electric energy in the capacitors(gaps) between the loops. A close examination of the loop struc-ture reveals that the loops can also store magnetic energy, givingthem an inductive characteristic. This is from the metallic stripsof the loops that are parallel with the electric field and supportan electric current, thus giving rise to the inductive behavior.Hence, while the two -directed sides of the loops, which areperpendicular to the electric field, are acting as the plates of acapacitor, the two -directed sides act as inductors. Two neigh-boring square loops, as a result, constitute a series combinationof a capacitor and an inductor, generating the loop layer band-stop characteristic.

As mentioned above, the miniaturized-element frequency-se-lective surface behavior also depends on the interaction of thelayers. A very thin substrate is used in this design to increase thisinteraction that is explained as follows. The magnetic field in-duced by the current flowing in the wire grid encircles the stripsof the grid itself. A portion of that field couples through thesquare loops on the other layer, inducing some electric currenton the loops’ traces. Conversely, the current on the loops pro-duces a magnetic field that couples to the wire grid. This induc-tive mutual coupling becomes stronger as the substrate thicknessdecreases.

In addition to the effect mentioned above, another kind of in-teraction between the layers is also observed. Based on initialfull-wave simulations of the structure, an -polarized incidentelectric field produces an electric field normal ( -directed) tothe surface between the two layers (Fig. 3). This phenomenoncan be explained considering the need for displacement currentsbetween the layers at resonance, where significant electric cur-rents are flowing on the loops and the wire grid. This interestinginteraction, which is observed only for cases where the substrateis very thin, is modeled as a capacitive junction at the locationswhere the strips of the wire grid and the square loops overlap.

In summary, the proposed structure is a parallel combinationof highly coupled bandstop and inductive surfaces. By choosingproper dimensions and aligning the surfaces, desired bandpassor bandstop characteristics are obtained. Finally, the tunabilityof the response can be achieved by altering the loops’ gap capac-itance, e.g., by interconnecting the loops via lumped capacitors,as shown in Fig. 1. The results of a parametric study are pre-sented below for qualitative and quantitative determination ofthe effect of each surface parameter.

III. PARAMETRIC STUDY AND CIRCUIT MODEL DEVELOPMENT

The miniaturized-element frequency-selective surface para-metric study begins with development of a circuit model to de-scribe the frequency behavior of the structure qualitatively. Acircuit model is highly desirable at the design stage to quicklypredict the response of the structure with some level of accu-racy. Finding a circuit model for frequency-selective surfaces is

776 IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES, VOL. 56, NO. 4, APRIL 2008

Fig. 3. Vector electric field plot shows the existence of significant normal fieldcomponents between the two layers justifying the placement of the series ca-pacitance in the equivalent-circuit model.

Fig. 4. Equivalent-circuit model for the coupled wire grid and bandstop sur-faces of the proposed miniaturized-element frequency-selective surface.

not always possible. This is especially true for frequency-selec-tive surfaces composed of resonant geometries for which sig-nificant mode (Bragg) couplings take place. That is, the energyfrom the fundamental TEM mode of the incident wave is dis-tributed among all nonpropagating Bragg modes that eventuallyconstitute the surface current. In these cases, equivalent-circuitmodels do not exist and accurate analysis can only be accom-plished using full-wave numerical simulations.

The proposed design operates entirely in TEM mode andmakes use of very small-size elements. As a result, no significanthigher order mode couplings are expected. Hence, a simple cir-cuit model is sufficient to characterize the behavior of the struc-ture.

To arrive at an accurate model, sensitivity analyses using afull-wave simulator are carried out. The full-wave approach isalso necessary to establish the relationship between the phys-ical parameters of the miniaturized-element frequency-selec-tive surface unit cell and the lumped elements of the circuitmodel. Having an accurate circuit model, one can synthesizea desired frequency response (center frequency, bandwidth, in-sertion loss, and tuning range) by an optimization method in areasonably short time using a circuit simulator. A circuit modelbased on the qualitative description of the structure provided inSection II is shown in Fig. 4. This model ignores all metallicand dielectric losses. The model includes two parallel branches,representing the two layers.

The right branch models the wire grid, which is purely induc-tive. For this branch, sensitivity analysis shows that increasingthe strip width decreases the inductance shown in the equiva-lent circuit. As will be shown below, the structure has an overallbandpass characteristic with a center frequency denoted by .

The reduction in strip width obviously decreases , but simula-tions also show that the bandpass response bandwidth becomeswider.

The left branch in the model, on the other hand, representsthe array of loops that acts like a bandstop circuit (series L–C).The total capacitance used in the circuit model is composed ofthe loops’ gap capacitance and the lumped capacitorthat is mounted in parallel with in the gap between twosuccessive loops. The gap capacitance is a function of thegap spacing , and, to a lesser extent, the loop strip width(see Fig. 2). Characterization of the gap capacitance has beendiscussed in [24]. As expected, increasing decreases the ca-pacitance, which, in turn, increases the notch frequencyat which there is a transmission zero. The simulations indicatethat increasing also increases . It is also observed that in-creasing results in an increase in . This is mostly due to thelower inductance of the wider strip. A more detailed study, how-ever, reveals that changing the loop trace width also changes thegap capacitance, as mentioned above. Although increasing in-creases the gap capacitance, the inductance decreases in sucha manner that both and increase. This is due to the verysmall value of the gap capacitance compared to that of thelumped capacitor mounted within the gap. As a result, theloop trace width primarily affects the inductance of the loop sur-face without changing the gap capacitance significantly.

In addition to the elements discussed above, the circuit modelhas other elements that represent the layers’ interactions. Themagnetic interaction discussed in Section II is modeled by amutual inductance in the circuit model; this coupling be-comes larger to represent increased coupling as the substratethickness decreases. The displacement current flowing betweenthe layers forms a capacitive junction, which is modeled as aseries capacitor . To model the substrate, a very short pieceof transmission line is placed between the shunt branches. Sen-sitivity analysis shows that increasing the substrate thicknessbeyond 200 m increases the insertion loss drastically; hence,degrading the bandpass characteristics of the response. More-over, slightly decreases as increases.

Given the equivalent-circuit model for the miniaturized-el-ement frequency-selective surface, a set of typical values arechosen for the circuit elements to show the expected frequencyresponse of the surface (Fig. 5). A bandpass behavior (ap-proximately 9 GHz) with a transmission zero (approximately13 GHz) is clearly demonstrated. The value of each circuit ele-ment is nH, nH, nH, ,

pF, and pF.

IV. MODEL VERIFICATION AND SIMULATION RESULTS

In Section III, the synthesis of a highly selective bandpassresponse by the miniaturized-element frequency-selective sur-face was shown to be possible, assuming accuracy of the cir-cuit model. Next, employing the trial-and-error approach, a full-wave simulator is used to verify the behavior of the structure.

In this section, three different miniaturized-element fre-quency-selective surface designs whose parameters’ valuesare provided in Table I are presented. Full-wave analyses arebased on the finite element method (FEM); both dielectricand metallic losses are included. The accuracy of the circuit

BAYATPUR AND SARABANDI: SINGLE-LAYER HIGH-ORDER MINIATURIZED-ELEMENT FREQUENCY-SELECTIVE SURFACES 777

Fig. 5. Circuit model simulation shows a bandpass characteristic with high se-lectivity and low insertion loss, including a transmission zero.

TABLE IBANDPASS MINIATURIZED-ELEMENT FREQUENCY-SELECTIVE

SURFACES’ STATIC DESIGN PARAMETERS AT �-BAND

model is also verified using sensitivity analysis by which thecorresponding values for the elements of the circuit model areextracted to best fit the full-wave results.

In the following, the frequency tuning capability is studiednumerically by altering the lumped capacitance mounted in theloops’ gap. The simulation results for the first design (Table I)are shown in Fig. 6. As can be seen, a frequency range from 6.7to 9.4 GHz is swept by altering the lumped capacitance from0.7 to 0.2 pF. The simulations predict a very wide tuning rangewith very little degradation in the structure performance. Fig. 7shows the simulation results for the second design with unit celldimensions of 2.5 mm 2.5 mm. Among the designs consid-ered here, the second design achieves the smallest cell size. Likethe first design, the second design has a very wide frequencytuning range. In this case, the center frequency changes from6.4 to 9.7 GHz by changing the capacitance from 0.3 to 0.1 pF.This capacitance range is rather small and, thus, not very desir-able in practice. Finally, Fig. 8 shows the simulation results forthe third design. This design also has a wide tuning range forthe capacitance range of 0.7 to 0.2 pF.

The third design, however, has a lower insertion loss com-pared to the first design. Another advantage of the third design

Fig. 6. Full-wave simulations of frequency response (return loss and transmis-sion) of the first miniaturized-element frequency-selective surface design in-cluding metallic and dielectric losses. The results show that a wide tuning rangecan be obtained.

Fig. 7. Full-wave simulations of frequency response (return loss and transmis-sion) of the second miniaturized-element frequency-selective surface design in-cluding metallic and dielectric losses. The results show that a wide tuning rangecan be obtained.

is in that its substrate is considerably thicker than that of thefirst and, therefore, is much easier to fabricate and handle ex-perimentally.

Comparison of the circuit model response with full-wave sim-ulations for the second design is shown in Fig. 9. Extracted cir-cuit elements’ values are given in Table II. Two different values

778 IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES, VOL. 56, NO. 4, APRIL 2008

Fig. 8. Full-wave simulations of frequency response (return loss and transmis-sion) of the third miniaturized-element frequency-selective surface design in-cluding metallic and dielectric losses. The results show that a wide tuning rangecan be obtained. This design provides lower insertion loss and uses a thickersubstrate compared to the first design.

Fig. 9. Circuit model simulations are compared against FEM model for thesecond miniaturized-element frequency-selective surface design and for dif-ferent values of the lumped capacitor: � � ���� and ��� pF.

of the lumped capacitance, i.e., and pF, are usedto show the validity of the circuit model. As shown, the modelresponse not only predicts the FEM result, but it also tracks theFEM results as the lumped capacitance is tuned. The circuitmodel, however, ignores the losses associated with the metaland the substrate.

TABLE IICIRCUIT MODEL ELEMENTS’ VALUES—SECOND DESIGN

Fig. 10. Unit cell of a dual-bandpass miniaturized-element frequency-selectivesurface composed of a cluster of four identical loops, but with different lumpedcapacitors.

Fig. 11. Frequency response of a dual-band miniaturized-element frequency-selective surface obtained by using a cluster of four loops as the unit cell andchoosing different values for the lumped capacitor (dashed line) capacitors arethe same �� � � � ��� pF� and (solid line) � � ���� pF and � ���� pF.

V. DUAL BANDPASS FREQUENCY RESPONSE

In many communication and radar applications, frequency-selective surfaces with multiband characteristics are highly de-sirable. The new miniaturized-element frequency-selective sur-face can be designed to provide such characteristics from just asingle substrate layer. This is accomplished in a rather straight-forward manner because of the very small unit cell dimensionsof the new design and the localized nature of the frequency re-sponse. To achieve a multiple bandpass behavior, one needs toincrease the number of poles and zeros. For past designs, anincrease in the number of poles and zeros translates to havinga multilayer frequency-selective surface. Due to the small sizeof the unit cells of the new design, the number of poles and

BAYATPUR AND SARABANDI: SINGLE-LAYER HIGH-ORDER MINIATURIZED-ELEMENT FREQUENCY-SELECTIVE SURFACES 779

Fig. 12. Free-space measurement setup consisting of a receiver, a transmitter,and the miniaturized-element frequency-selective surface in between.

Fig. 13. Fabricated surface using fixed lumped capacitors. This prototype usesapproximately 3000 capacitors with capacitance value of � � ��� pF. Thisfigure shows the loop layer.

zeros can be increased by using a cluster of loops with dif-ferent parameters as the unit cell. In this section, a dual-bandresponse is demonstrated by considering a four-loop cluster asthe unit cell and simply placing different lumped capacitors inthe gaps between the loops, as shown in Fig. 10. The equivalentcircuit of this composite loop-wire grid surface is two parallelcircuits, each similar to that shown in Fig. 4. It can be shownthat by proper choice of the lumped capacitors and , adual bandpass response can be obtained. Using the parametersof the second design given in Table I, changing the loop tracewidth from 0.15 to 0.25 mm, and choosing pF,a single bandpass frequency response, shown in Fig. 11, is pro-duced. By changing the capacitance values to pFand pF, a dual-band response, also shown in Fig. 11,is obtained.

VI. EXPERIMENTAL VERIFICATION

To demonstrate the validity of the simulations, a prototypesample of the miniaturized-element frequency-selective surface

Fig. 14. Miniaturized-element frequency-selective surface measurementresults versus numerical simulation. Angle � measures the orientation onthe surface. In both cases ��� � � and ��� � �� , the incident wavepolarization is parallel to the surface. This measurement shows that the designis polarization insensitive.

Fig. 15. Miniaturized-element frequency-selective surface measurementresults versus numerical simulation. This measurement shows the effect ofoff-normal excitation of the surface. Angle theta represents the angle betweenthe propagation vector of the incident wave and the normal to the surface.

at -band is fabricated and tested using a free-space measure-ment setup. The setup is composed of a lens-corrected horn an-tenna for creation of a uniform phase front over a finite apertureand a high-gain antenna in the far field. The surface is placedadjacent to the horn, as shown in Fig. 12, and the receiving an-tenna is in the far-field region of the horn and surface.

The surface is fabricated using a low-loss Duroid substratethrough a printed circuit board etching process. 0.1-pF thin-filmchip capacitors with a very small package size (0201 standardsize) are then mounted in the loop gaps. The fabricated sample isbased on the third design whose parameters’ values are listed in

780 IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES, VOL. 56, NO. 4, APRIL 2008

Table I. The design screen (Fig. 13) is 6 in 6 in, which includes38 38 unit cells.

An 8720D vector network analyzer is employed to measurerthe transmissivity through the sample. A thru calibrationis performed to calibrate the network analyzer within a fre-quency band of 8.4–11.15 GHz (the lens-horn band of opera-tion) in the absence of the surface. Fig. 14 compares the sim-ulation and experimental results at normal incidence. Fig. 14also includes the case in which the polarization of the incidentwave is rotated by an angle 45 about the surface normal. Aspredicted by the simulations, the structure is polarization insen-sitive. The measurement results for oblique incidence up to 45 ,are shown in Fig. 15. As expected, the new design shows onlyminor frequency response dependence to the incidence angle. Agood agreement between the measured and simulated results isobserved.

VII. CONCLUSION

In this paper, a new miniaturized-element frequency-selectivesurface has been presented. A single substrate has been used togenerate a high-order bandpass characteristics with tuning capa-bilities. The unit cell size of this design can be as small as .The salient feature of the new structures is that their property islocalized and thus suitable for moderate size antennas’ appli-cations at low frequencies. High-order frequency response hasbeen achieved by establishing proper coupling between the in-ductive and bandstop surfaces on either side of the substrate. Anaccurate circuit model for the proposed design has been devel-oped that allows for rapid optimization of a desired frequencyresponse. The accuracy of the circuit model has been verifiedusing a full-wave approach. The frequency tuning capabilityof the response has also been demonstrated through numericalsimulations. A wide tuning range with a negligible loss of per-formance from approximately 6 to 10 GHz has been shown byaltering the lumped capacitance in the loops’ gap from 0.7 to0.2 pF. The validity of the simulation results has been shownthrough careful experimentations. A prototype of the design hasbeen fabricated and tested in a free-space environment. Goodagreement between the simulations and experimental results hasbeen shown.

REFERENCES

[1] J. C. Vardaxoglou, Frequency-Selective Surfaces: Analysis and De-sign. Taunton, U.K.: Res. Studies Press, 1997.

[2] T. K. Wu, Frequency-Selective Surface and Grid Array. New York:Wiley, 1995.

[3] B. A. Munk, Frequency-Selective Surfaces: Theory and Design. NewYork: Wiley, 2000, pp. 5–6.

[4] K. Sarabandi and N. Behdad, “A frequency selective surface withminiaturized elements,” IEEE Trans. Antennas Propag., vol. 55, no. 5,pp. 1239–1245, May 2007.

[5] T. K. Chang, R. J. Langley, and E. A. Parker, “Frequency selectivesurfaces on biased ferrite substrates,” Electron. Lett., vol. 30, no. 15,pp. 1193–1194, Jul. 1994.

[6] G. Y. Li, Y. C. Chan, T. S. Mok, and J. C. Vardaxoglou, “Analysis offrequency selective surfaces on biased ferrite substrate,” in IEEE AP-SDig., Jun. 1995, vol. 3, pp. 1636–1639.

[7] Y. Liu, C. G. Christodoulou, P. F. Wahid, and N. E. Buris, “Analysisof frequency selective surfaces with ferrite substrates,” in IEEE AP-SDig., Jun. 1995, vol. 3, pp. 18–23.

[8] Y. C. Chan, G. Y. Li, T. S. Mok, and J. C. Vardaxoglou, “Analysisof a tunable frequency-selective surface on an in-plane biased ferritesubstrate,” Microw. Opt. Technol. Lett., vol. 13, no. 2, pp. 59–63, Oct.1996.

[9] Y. Liu, C. G. Christodoulou, and N. E. Buris, “Fullwave analysismethod for frequency selective surfaces on ferrite substrates,” J.Electromagn. Waves Applicat., vol. 11, no. 5, pp. 593–607, 1997.

[10] E. A. Parker and S. B. Savia, “Active frequency selective surfaces withferroelectric substrates,” Proc. Inst. Elect. Eng.—Microw., Antennas,Propag., vol. 148, pp. 103–108, Apr. 2001.

[11] A. C. d. C. Lima, E. A. Parker, and R. J. Langley, “Tunable frequencyselective surface using liquid substrates,” Electron. Lett., vol. 30, no. 4,pp. 281–282, Feb. 1994.

[12] J. C. Vardaxoglou, “Optical switching of frequency selective surfacebandpass response,” Electron. Lett., vol. 32, no. 25, pp. 2345–2346,Dec. 1996.

[13] T. Anderson, I. Alexeff, and J. Raynolds, “Plasma frequency selectivesurfaces,” in Proc. IEEE Int. Plasma Sci. Conf., Jun. 2003, pp. 237–237.

[14] B. Schoenlinner, A. Abbaspour-Tamijani, L. C. Kempel, and G. M.Rebeiz, “Switchable low-loss RF MEMS ��-band frequency-selec-tive surface,” IEEE Trans. Microw. Theory Tech., vol. 52, no. 11, pp.2474–2481, Nov. 2004.

[15] D. Lockyer, R. Seager, and J. C. Vardaxoglou, “Reconfigurable FSSusing multilayer conducting and slotted array structures,” in IEE Adv.Electromagn. Screens, Radomes, Mate. Colloq., Oct. 1996, pp. 6/1–6/4,Dig. 270.

[16] J. P. Gianvittorio, J. Zendejas, Y. Rahmat-Samii, and J. Judy, “Recon-figurable MEMS-enabled frequency selective surfaces,” Electron. Lett.,vol. 38, no. 25, pp. 1627–1628, Dec. 2002.

[17] D. S. Lockyer and J. C. Vardaxoglou, “Reconfigurable FSS responsefrom two layers of slotted dipole arrays,” Electron. Lett., vol. 32, no. 6,pp. 512–513, Mar. 1996.

[18] C. Mias, “Frequency selective surfaces loaded with surface-mount re-active components,” Electron. Lett., vol. 39, no. 9, pp. 724–726, May2003.

[19] T. K. Chang, R. J. Langley, and E. Parker, “An active square loop fre-quency selective surface,” IEEE Microw. Guided Wave Lett., vol. 3, pp.387–388, Oct. 1993.

[20] T. K. Chang, R. J. Langley, and E. A. Parker, “Active frequency-se-lective surfaces,” Proc. Inst. Elect. Eng.—Microw., Antennas, Propag.,vol. 143, pp. 62–66, Feb. 1996.

[21] B. Philips, E. A. Parker, and R. J. Langley, “Active FSS in an exper-imental horn antenna switchable between two beamwidths,” Electron.Lett., vol. 31, no. 1, pp. 1–2, Jan. 1995.

[22] B. M. Cahill and E. A. Parker, “Field switching in an enclosure withactive FSS screen,” Electron. Lett., vol. 37, no. 4, pp. 244–245, Feb.2001.

[23] B. Monacelli, J. B. Pryor, B. A. Munk, D. Kotter, and G. D. Boreman,“Infrared frequency selective surface based on circuit-analog squareloop design,” IEEE Trans. Antennas Propag., vol. 53, no. 2, pp.745–752, Feb. 2005.

[24] W. R. Smythe, Static and Dynamic Electricity. New York: McGraw-Hill, 1968, pp. 63–120.

Farhad Bayatpur (S’06) received the B.Sc. degree in electrical engineeringfrom the Sharif University of Technology, Tehran, Iran, in 2005, the M.S. de-gree in electrical engineering from The University of Michigan at Ann Arbor,in 2007, and is currently working toward the Ph.D. degree at The University ofMichigan at Ann Arbor.

Kamal Sarabandi (S’87–M’90–SM’92–F’00) re-ceived the B.S. degree in electrical engineering fromthe Sharif University of Technology, Tehran, Iran, in1980, and the M.S.E. and Ph.D. degrees from TheUniversity of Michigan at Ann Arbor, in 1986 and1989, respectively, both in electrical engineering.

He is currently Director of the Radiation Labora-tory and a Professor with the Department of ElectricalEngineering and Computer Science, The Universityof Michigan at Ann Arbor. He possesses 22 years ofexperience with wave propagation in random media,

communication channel modeling, microwave sensors, and radar systems and

BAYATPUR AND SARABANDI: SINGLE-LAYER HIGH-ORDER MINIATURIZED-ELEMENT FREQUENCY-SELECTIVE SURFACES 781

leads a large research group including two research scientists, 12 Ph.D. stu-dents, and two M.S. students. He has graduated 28 Ph.D. and supervised nu-merous post-doctoral students. He has served as the Principal Investigator onmany projects sponsored by the National Aeronautics and Space Administration(NASA), Jet Propulsion Laboratory (JPL), Army Research Office (ARO), Officeof Naval Research (ONR), Army Research Laboratory (ARL), National ScienceFoundation (NSF), Defense Advanced Research Projects Agency (DARPA),and numerous industries. He has authored or coauthored many book chaptersand over 150 papers in refereed journals on miniaturized and on-chip antennas,metamaterials, electromagnetic scattering, wireless channel modeling, randommedia modeling, microwave measurement techniques, radar calibration, inversescattering problems, and microwave sensors. He has also had over 380 papersand invited presentations in many national and international conferences andsymposia on similar subjects. He is listed in American Men and Women of Sci-ence, Who’s Who in America, and Who’s Who in Science and Engineering. Hisresearch interests include microwave and millimeter-wave radar remote sensing,metamaterials, electromagnetic wave propagation, and antenna miniaturization.

Dr. Sarabandi is a member of NASA Advisory Council appointed bythe NASA Administrator. He also served as a vice president of the IEEEGeoscience and Remote Sensing Society (GRSS) and a member of the IEEETechnical Activities Board Awards Committee. He serves on the EditorialBoard of PROCEEDINGS OF THE IEEE. He was an associate editor of the IEEE

TRANSACTIONS ON ANTENNAS AND PROPAGATION and the IEEE SENSORS

JOURNAL. He is a member of Commissions F and D of URSI. He was therecipient of the Henry Russel Award presented by the Regent of The Uni-versity of Michigan at Ann Arbor, the 1999 GAAC Distinguished LecturerAward presented by the German Federal Ministry for Education, Science,and Technology, which is given to approximately ten individuals worldwidein all areas of engineering, science, medicine, and law. He was a recipient ofthe 1996 Electrical Engineering and Computer Science (EECS) DepartmentTeaching Excellence Award and a 2004 College of Engineering ResearchExcellence Award. He was the recipient of the 2005 IEEE GRSS DistinguishedAchievement Award and The University of Michigan at Ann Arbor’s FacultyRecognition Award. He was the recipient of the 2006 Best Paper Awardpresented at the Army Science Conference. He was also the recipient of the2008 Humboldt Research Award for Senior U.S. Scientist presented by TheAlexander von Humboldt Foundation of Germany, which is granted to scientistsand scholars in all disciplines with internationally recognized academic quali-fications. Over the past several years, joint papers presented by his students ata number of international symposia (IEEE APS’95,’97,’00,’01,’03,’05,’06,’07;IEEE IGARSS’99,’02,’07; IEEE IMS’01, USNC URSI’04,’05,’06, AMTA ’06)have received Student Paper Awards.