Switchable Low-Loss RF MEMS Ka-Band Frequency ... - IEEE Xplore

2474

IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES, VOL. 52, NO. 11, NOVEMBER 2004

Ka

Switchable Low-Loss RF MEMS -Band Frequency-Selective Surface Bernhard Schoenlinner, Student Member, IEEE, Abbas Abbaspour-Tamijani, Member, IEEE, Leo C. Kempel, Senior Member, IEEE, and Gabriel M. Rebeiz, Fellow, IEEE

Abstract—A switchable frequency-selective surface (FSS) was developed at 30 GHz using RF microelectromechanical systems (MEMS) switches on a 500- m-thick glass substrate. The 3-indiameter FSS is composed of 909 unit cells and 3636 MEMS bridges with a yield of 99.5%. The single-pole FSS shows a transmission loss of 2.0 dB and a 3-dB bandwidth of 3.2 GHz at a resonant frequency of 30.2 GHz with the MEMS bridges in the up-state position. The 1-dB bandwidth is 1.6 GHz. When the MEMS bridges are actuated to the down-state position, an insertion loss of 27.5 dB is measured. Theory and experiment agree quite well. The power handling is limited to approximately 25 W with passive air cooling and 150 W with active air cooling due to the increased temperature of the overall circuit resulting from the transmission loss (for continuous-wave operation with the assumed maximum allowable temperature of 80 C), or 370 W–3.5 kW due to self-actuation of the RF MEMS bridges (for pulsed incident power). Experimental results validate that 20 W of continuous-wave power can be transferred by the RF MEMS FSS with no change in the frequency response. This is -band the first demonstration of a switched low-loss FSS at frequencies. Index Terms—Frequency selective surface (FSS), microelectromechanical devices, quasi-optical, RF microelectromechanical systems (MEMS), tunable filters.

I. INTRODUCTION

F

REQUENCY-SELECTIVE surfaces (FSSs) have found applications in multiband reflector antennas [1], [2] and radomes [3]–[5], especially in communications and defense applications. FSSs are usually designed for a fixed frequency response. However, for certain applications, it is desirable to be able to change the frequency behavior over time. This is done using a ferrite substrate when the relative permeability is tuned with a bias magnetic field [6], [7]. The properties of the resonating elements themselves can also be changed using varactor diodes [8]–[10] or microelectromechanical systems (MEMS) such as rotating dipoles [11]. However, these methods have many disadvantages such as high losses (ferrite substrate), high bias currents (varactor or p-i-n diodes), or high cost (ferrite substrate, varactor diodes). Manuscript received February 24, 2004; revised June 13, 2004. This work was supported by the Automotive Systems Laboratory. This work was supported in part by the Defense Advanced Research Projects Agency Intelligent Reconfigurable Front-End Program under a subcontract from Rockwell Scientific. B. Schoenlinner, A. Abbaspour-Tamijani, and G. M. Rebeiz are with the Radiation Laboratory, Department of Electrical Engineering and Computer Science, The University of Michigan at Ann Arbor, Ann Arbor, MI 49109-2122 USA (e-mail: [email protected]; [email protected]). L. C. Kempel is with the Department of Electrical and Computer Engineering, Michigan State University, East Lansing, MI 48824-1326 USA (e-mail: [email protected]). Digital Object Identifier 10.1109/TMTT.2004.837148

Alternatively, a frequency shift in an FSS can be achieved using RF MEMS, and several advantages make the use of this technology as tuning elements very attractive [12]. The switching speed is reasonably fast (in the order of 10 s), the dc power consumption is extremely low (in the order of 100 nW per bridge due to a 10-nA leakage current in the down-state position and 10-V hold-down voltage), the tuning element loss at 30 GHz as compared to GaAs varactor is very low ( diodes with at 30 GHz), and the cost of fabrication does not increase with the number of elements since it is done using standard lithography. II. DESIGN OF THE SWITCHABLE FSS For a bandpass-type filter characteristic, an element of choice is the four-legged loaded element. It is polarization independent due to its 90 -rotational symmetry. It is also electrically small , which ensures that the grating lobes with a typical size of are of no concern in the frequency range of interest. The element design, together with the choice of the MEMS loading bridges and bias bridges, is well detailed in [12] (Fig. 1). , at The substrate is a glass wafer ( 30 GHz) with a thickness of 0.5 mm. The design frequency of the FSS is 32 GHz. The average bridge height on the fabricated circuit is measured to be 1.7 m using optical interferometry instead of originally assumed 2.0 m. Therefore, the presented simulations are retrofitted for the new bridge height. For a bridge height of 1.7 m, the up-state capacitance of the loading bridges is 50 fF (obtained from full-wave analysis using Sonnet1). Once the geometry of the bridge is set, the only remaining parameter, the length of the “legs,” is determined by a three-dimensional (3-D) full-wave simulation using the commercial finite-element method (FEM) High Frequency Structure Simulator (HFSS).2 A. Equivalent-Circuit Model An equivalent-circuit model is derived using the schematic shown in Fig. 2, which represents the transmission lines and bridge capacitances of a single unit cell. The input and output ports are connected to the center of the element via a transformer, which represents the coupling between the input/output plane waves and the guided modes in the resonator circuit can be determined (Fig. 3). A transformer ratio of using the method introduced in Section V or by fitting the calculated and measured -parameters. The two “legs” that carry the even mode (slot-line mode) are in parallel (same 1Sonnet 2HFSS,

EM Suite, Sonnet Software Inc., Liverpool, NY, 2002. Ansoft Corporation, Pittsburgh, PA, 2002.

0018-9480/04$20.00 © 2004 IEEE

SCHOENLINNER et al.: SWITCHABLE LOW-LOSS RF MEMS

-BAND FSS

Fig. 1. (a) Unit cell of the switchable FSS with bridges for capacitive loading of the slot, and dimensions in micrometers. (b) Unit cell with four bias bridges (loading bridges not shown).

2475

potential) and are represented as a single transmission-line stub . The with half of the impedance of a single stub two “legs” that carry the odd mode (CPW mode) are connected in series because the outer conductors are at opposite potentials with respect to the element center and are represented as a single transmission-line stub with twice the corresponding . The four bridge capacitors form a impedance parallel-series combination (Fig. 2), which can be represented , an as a single MEMS bridge with a resistance of pH, and a capacitance of fF inductance of at 30 GHz). The length in series ( can be determined using the of the legs is 825 m. Advanced Design System (ADS)3 and is 76 . Generally, the characteristic impedance in the even mode is not well defined and it can take any value less than depending on the width of the ground half-planes, which carry differential currents. The differential current between the ground half-planes results from their direct capacitive coupling, as well as their coupling through the center conductor. However, in the present case, the currents due to the direct coupling of the half-planes are short circuited due to the presence of the perfect electric conductor (PEC) walls at the bottom and top of each cell, and the only component of the differential current that contributes to the resonance is through the center conductor. For this component, the effective value of characteristic impedance . The attenuation of the transmission line in the odd is mode was measured in [13] at 21 GHz and is 70 dB/m. For the FSS at 30 GHz, the attenuation for the even and odd modes . The is approximated with 75 dB/m nonpropagating Floquet modes can be modeled using a shunt [14]. The value of this capacitor is found capacitance from matching the circuit model and HFSS simulations. Fig. 4 shows the transmission and reflection coefficients that result from the circuit model (using ADS) and the FEM simulation (using FSS). III. FABRICATION

Fig. 2.

Transmission-line schematic of a single unit cell.

Fig. 3.

Equivalent-circuit model of the switchable FSS.

The FSS is fabricated on a glass wafer with a diameter of mm (3 in) and consists of 909 unit cells with 3636 loading bridges and 1686 bias bridges (Fig. 5 and 6). The yield of movable (MEMS) loading bridges is 99.5%. Three unit cells are tapped on each side of the array to apply dc-bias voltage. The actuation of the loading bridges across the wafer can fail if any unwanted dc-connections are present due to fabrication errors, thus, a post-fabrication process is necessary to remove these faulty connections. A method to find the dc connections that are partly invisible under a microscope is to apply a dc current between the tapped unit cells and ground. As the current spreads throughout the array, the high current density in the short circuits causes heat generation in faulty cells. The hot spots are typically approximately 2 mm in size, and can be detected at the backside of the circuit using an infrared camera. Once located, the short circuits can be removed with a sharp strong metal tip under the microscope.

3ADS

2002, Agilent Technol. Inc., Santa Clara, CA, 2002.

2476


Fig. 7.

Free-space measurement system using hard horns.

Fig. 4. Simulated S -parameters of the switchable FSS with bridges in the up-state position using the equivalent-circuit model in comparison with the HFSS simulation.

Fig. 5. Unit cell of the switchable FSS.

Fig. 8. Measured transmission and reflection coefficients of the FSS versus frequency in comparison with the HFSS simulation: (a) with MEMS bridges in the up-state and (b) down-state positions (V = 17 V).

Fig. 6. Switchable FSS with 909 unit cells, 3636 loading bridges, and 1686 bias bridges.

IV.

-PARAMETER MEASUREMENTS

A. Measurement Results The measurement apparatus is well described in [12] and thru-reflect-line (TRL) calibration is used to bring the reference planes to the device-under-test (DUT) (Fig. 7). Due to the scattering at the edge of an actual circuit, an FSS is only an approximation of an infinite periodic structure. Therefore, the -parameter measurements of the FSS are specific to the size of the FSS and the measurement setup. With the measurement method described above, edge effects are minimized and the scenario of an incident plane wave is reasonably well approximated. However,

for large angles of incidence, the effective area of the FSS is reduced and the results are somewhat qualitative. The dynamic range of the measurement using TRL calibration is better than 40 dB for both co- and cross-polarized components from 24 to 40 GHz. Normal Incidence: The transmission and reflection coefficients of the FSS with bridges in up- and down-state positions for normal incidence are shown in Fig. 8(a) and (b). In the up-state position, the resonant frequency is 30.2 GHz with an insertion loss of 2.0 dB and a reflection coefficient of 11.5 dB. The 3-dB bandwidth is 3.2 GHz. The increase in the 3-dB bandwidth is likely due to the nonuniformity of the MEMS bridges over the whole circuit, which causes a deviation in resonant frequency of the individual resonators. When applying a dc-bias voltage between 0–13 V, the resonant frequency is


-BAND FSS

2477

Fig. 9. Measured transmission coefficient of the FSS versus frequency with different applied dc-bias voltage levels.

shifted to 29.4 GHz due to the increased capacitance of the loading bridges (Fig. 9) and the transmission loss is increased to 2.9 dB while the 3-dB bandwidth stays constant. When the bridges are in the down-state position, which requires a dc-bias voltage of 17 V, the measured transmission coefficient is 27.5 dB at 30.2 GHz and the reflection is close to 0 dB, which agrees well with simulation. In the simulations, the pF, which down-state capacitance is assumed to be . corresponds to a capacitance ratio of As a measure of polarization independence, the transmission coefficient for normal incidence is measured for different angular orientation of the FSS. It is observed that the transmission coefficient changes only by 0.05 dB for a full 360 turn of the FSS. A detailed analysis of the 2.0-dB loss indicates that 0.5 dB [13], of the loss is due to the dielectric loss , and 1.2 dB is [15], 0.3 dB due to the reflection loss . due to the finite of the RF MEMS cross-resonator For a perfectly matched FSS and a low-loss dielectric substrate , the estimated loss is 1.2 dB. Oblique Incidence: The transmission coefficient for different angles of incidence is shown in Fig. 10. For orthogonal polarization of the electric field (TE wave), the resonant frequency is stable versus angle of incidence and the 3-dB bandwidth is reduced only slightly to 2.9 GHz at an angle of incidence of 50 . For parallel polarization (TM wave), the resonant frequency is shifted to 28.6 GHz at an angle of incidence of 50 and the transmission loss is increased to 3.3 dB, while maintaining a 3-dB bandwidth of 3.2 GHz. The reason for the shift in frequency is a transmission minimum at 35 GHz, which becomes more prominent with the increasing angle of incidence. The measured cross-polarization level for normal incidence is below 29 dB and is below 27 dB for all measured angles of incidence from 24 to 40 GHz (Fig. 11). The existence of the transmission zero can neither be explained by substrate modes, nor by grating lobes. The reason can be found when considering the modes in a unit cell of the switchable FSS. For normal incidence, the PEC symmetry walls force virtual grounds at the location of the bias bridges. The only fundamental mode that can exist is the mode with a voltage maximum at the location of the loading bridges and a voltage minimum at the location of the bias bridges [see Fig. 12(a)]. At oblique angles of incidence with parallel (TM) polarization, no

Fig. 10. Measured transmission coefficient of the FSS versus frequency for different angles of incidence. (a) Orthogonal and (b) parallel polarization.

Fig. 11. Measured transmission coefficient for the cross-polarized component of the switchable FSS for normal and for oblique incidence.

PEC walls exist and a mode with a voltage maximum at the location of the bias bridges and a voltage zero at the location of the loading bridges can be excited in the resonators [see Fig. 12(b)]. This mode does not radiate since the electric fields (magnetic currents) in the slots propagate in the odd mode in all “legs” and cancel in the far-field, but it creates a reactive loading and impedance mismatch. Based on the transmission-line schematic of Fig. 13, the equivalent-circuit model can be extended to account for the additional propagating mode (Fig. 14). For the second mode, all four “legs” carry odd modes and are connected in parallel. The transmission lines are connected to ground through the

2478


Fig. 12. Fundamental modes in the switchable FSS. (a) The single excited mode for normal incidence. (b) A second mode, which is excited for oblique angles of incidence and TM polarization.

Fig. 15. Measured and simulated (circuit model) transmission coefficient of the switchable FSS for an angle of incidence of 30 (TM polarization) using the model of Fig. 14.

Fig. 16. One-pole RLC bandpass filter configuration with external loading, where C is either C or C depending on the actuation voltage.

Fig. 13. Transmission-line schematic of a single unit cell for the second propagating mode for oblique angles of incidence (TM polarization).

incidence, and is found to be 0.14 to match the measurement for 30 (Fig. 15). An easy way to prevent the existence of the unwanted mode close to the midband frequency is to avoid the air gap of the bias circuitry and design the bias bridges as metal–insulator–metal (MIM) capacitors. This will have no effect on the fundamental transmitting mode, but will increase the capacitive loading of the unwanted mode and will shift its resonant frequency to lower than 10 GHz. V. HIGH-POWER CONSIDERATIONS

Fig. 14. Equivalent-circuit model of the switchable FSS for oblique angles of incidence (TM polarization).

capacitance of the bias bridges, fF. For symmetry reasons, only half of is effective for each of the four transmission lines (legs). The loading bridge capaciand the compensation capacitance have tances no effect because they are connected across a voltage minimum. The resulting input impedance at the center of the , resonator can then be written as represents the equivalent where . The transelectrical length of terminating capacitors former ratio for the new mode depends on the angle of

As a distributed circuit, the FSS has the potential of handling high RF powers. However, simply multiplying the power-handling capabilities of a single MEMS switch in a transmission line by the number of switches implemented in the FSS will lead to wrong results since this is a resonant structure and not a bridge in a 50- transmission line. There are different possible failure mechanisms due to high incident RF power. The predominant effects are self-actuation and hold-down of the MEMS bridges, and thermal destruction of the entire glass wafer due to transmission loss in the transmit mode. In [16] and [17] it was shown that the current density in the bridges in either the up- or down-state positions and the thermal destruction of the MEMS bridges due to their finite electrical and thermal conductivity are not limiting factors. In the following, each single failure mechanism is treated individually. Self-Actuation: A one-pole filter equivalent circuit with resonator is shown in Fig. 16. The external a parallel and are the characteristic impedance loading impedances , and is the equivalent parallel resisof free space tance of the resonator. The transformer ratio at the input and output ports is assumed equal because the thickness of the wafer


-BAND FSS

2479

is much smaller than a wavelength. The loaded quality factor can be calculated from the measured frequency response (1) and is equal to 9.4. The unloaded quality factor using

is then 44.8 (2)

and the external quality factor

is determined using (3)

and is . Knowing that ricated circuit, the equivalent values for using

fF for the faband can be found

(4) , the resonant freFor the bridges in the up state , quency is 30.2 GHz. For the circuit with no bridges the resonant frequency is determined through HFSS simulations pH and fF. to be 52 GHz. This results in is found using (5)

Fig. 17. Maximum incident power on the switchable FSS versus heat coefficient for an ambient temperature of 20 C.

the up-state position, the measured transmission coefficient is 2.0 dB or 63% of the power is transmitted. With a reflection coefficient of 11.5 dB (7% is reflected), this means that 30% of the incoming power is absorbed by the FSS, and leads directly to an increase of the overall temperature of the FSS. Since the aspect ratio of the circuit is very high (76.4-mm wide, 0.5-mm thick), the FSS can be treated using a one-dimensional thermal model. The temperature gradient across the thickness of the glass wafer is very small because of the thin wafer used and, therefore, it can be accurately assumed that the power is dissipated uniformly over the substrate thickness. The governing equation is (steady-state analysis)

and is 3.13 k . The transformer ratio can be found using (9) (6) with

being defined by (7)

and is (which agrees well with Figs. 3 and 4). With a maximum allowable RF voltage in the slotline of before self-actuation occurs, the input is limited to . Using voltage (8) where V. Hence, the maximum incident power per bridge is 102–962 mW, which results in an overall maximum input power of 370 W–3.5 kW. Hold Down: The hold-down voltage for MEMS bridges is V. considerably lower than the pull-down voltage When the bridges are in the down-state position, no standing wave is induced in the resonator at , and using the circuit model of Fig. 16, the induced RF voltage across the MEMS bridge is decreased by a factor of 125 assuming a down-state capacitance of 1.1 pF. This means that the power handling of the array for the hold-down condition is approximately 100 higher than for the pull-down condition. Thermal Destruction Due to Transmission Loss: The maximum allowable temperature for the FSS is approximately 80 C since, for higher temperatures, the gold membranes start to deform and the performance of the circuit is affected. In

where is the dissipated power per unit area (W/m ), is the convective heat transfer coefficient per unit area [W K m ], is the temperature difference between the circuit and and environment. Values for are empirical and range from 2 to 25 for natural air flow and from 25 to 250 for forced air flow [18]. Fig. 17 shows the relationship between and the maximum incident power on the FSS for a maximum temperature of 80 C C . The temand an ambient temperature of 20 C perature of 80 C is chosen based on the elongation/stress model presented in [19]. For an incident power of 25 W (absorbed mW mm ), the required power in the FSS is 7.5 W, value of is 15 W K m , which is achievable using normal W K m , which requires active air air flow. For cooling, the incident power cannot be as high as 180 W (54-W absorbed power). The validity of the one-dimensional model is confirmed with a two-dimensional simulation using MATLAB4 (not presented here). Calculations done with a heat sink around the edge of the FSS show no improvement in the power handling due to the large thermal resistance from the center of the FSS to the heat sink. Summary: For a continuous-wave operation, the power handling of the fabricated FSS is limited by the increased temperature of the overall circuit due to the insertion loss and is approximately 25 W if no air cooling is available. For short pulses of incident power, the power limit is set by the selfactuation due to RF power and is 370 W–3.5 kW depending on 4MATLAB,

The Mathworks Inc., Natick, MA, 2002.

2480


the pull-down voltage. In these considerations, it is assumed that the incident power is distributed evenly across the FSS. Experiment: The power-handling capabilities were investigated using a traveling-wave tube (TWT) amplifier with a maximum output power of 50 W and the hard horns of Fig. 7. With passive air cooling and no bias on the array, there is no change is for 1–20-W incident power. However, at 40 W, numerous bridges near the center of the array are heated to 100 C and change height, therefore, resulting a 2–3-dB reduction in (the temperature of the FSS could not be monitored during the experiment) [19].

VI. CONCLUSIONS A bandpass-type FSS with switching capabilities has been demonstrated. A four-legged loaded element as a unit cell has been modified to incorporate RF MEMS shunt varactors as tuning elements and to facilitate a simple biasing scheme. A 3-in FSS with 909 unit cells and 3636 RF MEMS varactors with a yield of 99.5% has been fabricated and tested. A circuit model has been developed to explain an unwanted transmission minimum for oblique angles of incidence and parallel polarization and a simple method to avoid this phenomenon has been proposed. The power handling is limited to approximately 25 W (no active cooling) for continuous-wave operation and to 370 W–3.5 kW for pulsed incident power. We believe that this is the first demonstration of a switchable low control-power FSS layer with high performance above 6 GHz.

REFERENCES [1] L. C. Comtesse, R. J. Langley, E. A. Parker, and J. C. Vardaxoglou, “Frequency selective surfaces in dual and triple band offset reflector antennas,” in 17th Eur. Microwave Conf., Rome, Italy, 1987, pp. 208–213. [2] T.-K. Wu and S.-W. Lee, “Multiband frequency selective surface with multiring patch elements,” IEEE Trans. Antennas Propagat., vol. 42, pp. 1084–1092, Sept. 1992. [3] P. Callaghan, E. A. Parker, and R. J. Langley, “Influence of supporting dielectric layers on the transmission properties of frequency selective surfaces,” Proc. Inst. Elect. Eng., pt. H, vol. 138, no. 5, pp. 448–454, Oct. 1991. [4] W. R. Bushelle, L. C. Hoots, and R. M. Van Vliet, “Development of a resonant metal radome,” in Electromagnetic Windows Conf., 1978, pp. 179–185. [5] B. A. Munk, Frequency Selective Surfaces. New York: Wiley, 2000. [6] T. K. Chang, R. J. Langley, and E. A. Parker, “Frequency selective surfaces on biased ferrite substrates,” Electron. Lett., no. 15, pp. 1193–1194, July 1994. [7] D. M. Pozar, “Radiation and scattering characteristics of microstrip substrates on normally biased ferrite substrates,” IEEE Trans. Antennas Propagat., vol. 40, pp. 1084–1092, Sept. 1992. [8] T. K. Chang, R. J. Langley, and E. A. Parker, “Active frequency-selective surfaces,” in Proc. Inst. Elect. Eng., vol. 143, Feb. 1996, pp. 62–66. [9] K. D. Stephan, F. H. Spooner, and P. F. Goldsmith, “Quasi-optical millimeter-wave hybrid and monolithic PIN diode switches,” IEEE Trans. Microwave Theory Tech., vol. 41, pp. 1791–1798, Oct. 1993. [10] C. Mias, “Frequency selective surfaces loaded with surface-mount reactive components,” Electron. Lett., vol. 39, no. 9, pp. 724–726, May 2003.

[11] J. P. Gianvittorio, J. Zendejas, and Y. Rahmat-Samii, “MEMS enabled reconfigurable frequency selective surfaces: Design, simulation, fabrication, and measurement,” in IEEE AP-S Int. Symp., vol. 2, June 2002, pp. 404–407. [12] B. Schoenlinner, L. C. Kempel, and G. M. Rebeiz, “Switchable RF MEMS -band frequency-selective surface,” in IEEE MTT-S Int. Microwave Symp. Dig., June 2004, pp. 1241–1244. [13] A. Abbaspour-Tamijani, L. Dussopt, and G. M. Rebeiz, “Miniature and tunable filters using MEMS capacitors,” IEEE Trans. Microwave Theory Tech., vol. 51, pp. 1878–1885, July 2003. [14] R. S. Elliott, Antenna Theory and Design. Englwood Cliffs, New Jersey: Prentice-Hall, 1981. [15] J. S. Hayden and G. M. Rebeiz, “Very low-loss distributed -band and -band MEMS phase shifters using metal–air–metal capacitors,” IEEE Trans. Microwave Theory Tech., vol. 51, pp. 309–314, Jan. 2003. [16] G. M. Rebeiz, RF MEMS: Theory, Design, and Technology. New York: Wiley, 2003. [17] J. B. Rizk, E. Chaiban, and G. M. Rebeiz, “Steady state thermal analysis and high-power reliability considerations of RF MEMS capacitive switches,” in IEEE MTT-S Int. Microwave Symp. Dig., vol. 1, June 2002, pp. 239–242. [18] Conventorware Manual, Coventor Inc., Cary, NC, 2003. [19] J. R. Reid, L. A. Starman, and R. T. Webster, “RF actuation of capacitive MEMS switches,” in IEEE MTT-S Int. Microwave Symp. Dig., vol. 3, June 2003, pp. 1919–1922.

Ka

Ka

X

Bernhard Schoenlinner (S’00) was born in Trostberg, Germany, in 1973. He received the Dipl.-Ing. degree from the Technische Universität München, Munich, Germany, in 2000, and the Ph.D. degree in electrical engineering from The University of Michigan at Ann Arbor, in 2004. His research interests are in microwave and millimeter-wave circuits and devices. His work focuses on millimeter-wave radar systems and RF MEMS devices. Dr. Schoenlinner received second place in the Student Paper Competition at the European Microwave Conference 2003, Munich, Germany. He also received third place in the Student Paper Competition at the IEEE Microwave Theory and Techniques Society (IEEE MTT-S) International Microwave Symposium (IMS), 2004, Fort Worth, TX.

Abbas Abbaspour-Tamijani (S’00–M’04) received the B.S. and M.S. degrees from the University of Tehran, Tehran, Iran, in 1994 and 1997, respectively, and the Ph.D. degree from The University of Michigan at Ann Arbor, in 2003, all in electrical engineering. From 1997 to 1999, he was an RF and Antenna Engineer in the telecommunication industry, during which time he was involved in the design of antennas and RF circuits and subsystems. From 1999 to 2000, he was with the Antenna Laboratory, University of California at Los Angeles (UCLA), where he was involved with the design of slot arrays and feed systems for space-borne reflector antennas. In Fall 2000, he joined The Radiation Laboratory, The University of Michigan at Ann Arbor. He is a Senior RF Engineer with Motia Inc., Pasadena, CA. In Fall 2004, he will join Arizona State University, Tempe, as an Assistant Professor of electrical engineering. His research area includes RF MEMS, phased arrays, focal plane scanning systems, passive and active quasi-optics, and multifunctional integrated devices for RF front-ends.


-BAND FSS

Leo C. Kempel (S’89–M’94–SM’99) was born in Akron, OH, in October 1965. He earned the B.S.E.E. degree from the University of Cincinnati, Cincinnati, OH, in 1989, and the M.S.E.E. and Ph.D. degrees from The University of Michigan at Ann Arbor, in 1990 and 1994, respectively. In 1994, following a brief post-doctoral appointment with The University of Michigan at Ann Arbor, he joined the Mission Research Corporation as a Senior Research Engineer. He led several projects involving the design of conformal antennas, computational electromagnetics, scattering analysis, and high-power/ultrawide-band microwaves. In 1998, he joined Michigan State University, East Lansing, where he conducts research in computational electromagnetics and electromagnetic materials characterization, teaches undergraduate and graduate courses in electromagnetics, and supervises the research of several M.S. and Ph.D. students. His current research interests include computational electromagnetics, conformal antennas, microwave/millimeter-wave materials, mixed-signal electromagnetic interference techniques, and measurement techniques. He is also affiliated with the National Science Foundation (NSF) Wireless Integrated Microsystems Engineering Research Center. He coauthored The Finite Element Method for Electromagnetics (Piscataway, NJ: IEEE Press, 1998). He is a reviewer for the Journal of Electromagnetic Waves and Applications and Radio Science. Dr. Kempel is a member of Tau Beta Pi, Eta Kappa Nu, and Commission B, International Scientific Radio Union (URSI). He served as technical chairperson for the 2001 Applied Computational Electromagnetics Society (ACES) Conference and technical co-chair for the Finite Element Workshop, Chios, Greece, 2002. He is a member of the Applied Computational Electromagnetic Society (ACES) Board of Directors. He organized several sessions at recent URSI and ACES meetings. He is a reviewer for several IEEE publications. He serves as an Associate Editor for the IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION. He was the recipient of a CAREER award presented by the NSF and the 2002 Teacher–Scholar Award presented by Michigan State University. He was also the recipient of the 2001 Michigan State University College of Engineering’s Withrow Distinguished Scholar (Junior Faculty) Award.

2481

Gabriel M. Rebeiz (S’86–M’88–SM’93–F’97) received the Ph.D. degree in electrical engineering from the California Institute of Technology, Pasadena. He is a Full Professor of electrical engineering and computer science (EECS) at The University of Michigan at Ann Arbor. He authored RF MEMS: Theory, Design and Technology (New York: Wiley, 2003). His research interests include applying MEMS for the development of novel RF and microwave components and subsystems. He is also interested in SiGe RF integrated-circuit (RFIC) design, and in the development of planar antennas and millimeter-wave front-end electronics for communication systems, automotive collision-avoidance sensors, and phased arrays. Prof. Rebeiz was the recipient of the 1991 National Science Foundation (NSF) Presidential Young Investigator Award and the 1993 International Scientific Radio Union (URSI) International Isaac Koga Gold Medal Award. He was selected by his students as the 1997–1998 Eta Kappa Nu EECS Professor of the Year. In October 1998, he was the recipient of the Amoco Foundation Teaching Award, presented annually to one faculty member of The University of Michigan at Ann Arbor for excellence in undergraduate teaching. He was the corecipient of the IEEE 2000 Microwave Prize. In 2003, he was the recipient of the Outstanding Young Engineer Award of the IEEE Microwave Theory and Techniques Society (IEEE MTT-S). He is a Distinguished Lecturer for the IEEE MTT-S.