4 elliott r. brown 01

Upload: ishajain4

Post on 05-Apr-2018

217 views

Category:

Documents


0 download

TRANSCRIPT

  • 7/31/2019 4 Elliott R. Brown 01

    1/6

    IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 49, NO. 10, OCTOBER 2001 1357

    On the Gain of a Reconfigurable-Aperture AntennaElliott R. Brown, Fellow, IEEE

    AbstractA full-wave analysis based on the method of mo-ments (MoM) is carried out for a reconfigurable-aperture antennaconsisting of a two-dimensional (2-D) array of filamentarymicrostrip-dipoles interconnected by lossy microelectrome-chanical-system (MEMS) switches. Activation of specific MEMSswitches allows the dipoles to be maintained near the halfwave-res-onant length as the frequency is reduced in octave incrementsbetween 16 and 2 GHz. This keeps the real part of the dipoleself-impedance much higher and the imaginary part much lowerthan in a dipole having a fixed length of at 16 GHz. Hence,the array-antenna gain and aperture efficiency remain muchhigher with frequency than in an array of fixed dipoles. Broadside aperture efficiencies of 3.9, 6.0, 9.5, and 10.6 dB arepredicted for , , , and recap dipole arraysat frequencies of 16, 8, 4, and 2 GHz, respectively, for MEMSswitches having 0.5 dB insertion loss. In contrast, fixed-element

    -separated arrays operating at the same frequencies have

    predicted efficiencies of 3.9, 24.2, 45.0, and 63.0 dB,respectively.

    Index TermsAntenna gain, aperture antennas, reconfigurableantennas.

    I. INTRODUCTION

    SEVERAL research programs have been started recently to

    develop electronic antennas by a new approach called a

    reconfigurable aperture, or recap for short. As suggested by

    the title, a recap is fundamentally different than the traditional

    electronic antennas that have been developed over the years, in-

    cluding the family of single-beam phased arrays, multiple-beam

    aperture antennas (e.g., Rotman lens), and the growing familyof switchable-element smart antennas [1]. The distinguishing

    feature of a recap is its ability to alter the RF current distribution

    within a planar-radiating aperture. In the language of phased ar-

    rays, a recap can change its element pattern in addition to its

    complex-array factor.

    To see this distinction more precisely, recall that the electric

    (far) field from a traditional-phased array can be written [ 2]

    (1)

    where

    ;

    distance (radial unit vector) between the center

    of the array and the measurement point;

    Manuscript received August 5, 2000; revised November14, 2000. This workwas supported by the Defense Advanced Research Projects Agency under theReconfigurable Aperture Program.

    The author is with the University of California, Los Angeles, Los Angeles,CA 90095-1594 USA (e-mail: [email protected]).

    Publisher Item Identifier S 0018-926X(01)06368-2.

    vector between the th element and the measure-

    ment point;permeability of free space;

    propagation constant ;

    element factor;

    array factor;

    total number of elements in the array.

    In traditional phased-array antennas, the element factor is fixed

    by design and cannot be altered electronically or otherwise. The

    far-field pattern is changed only by variation of the complex

    coefficients in phase, amplitude, or both. In a recap one can

    also change the element factor.

    Optoelectronic [3] and MEMS switches [4] have already been

    used to change the length of single-element antennas, such as

    dipoles. Similar devices have been proposed to change the el-ement factor in arrays [5], but it is difficult to tell at this point

    which will work best. However, two benefits of a recap, inde-

    pendent of the core technology, should be the antenna gain as a

    function of beam pointing and as a function of frequency shift.

    The pointing issue has already been addressed through the de-

    velopment of a broad-side/end-fire switchable antenna based on

    an integrated leaky-mode/YagiUda structure [6] and willnotbe

    addressed here. Instead, this paper focuses on the issue of fre-

    quency shift and how a recap array of resonant elements can

    maintain high gain over a wide (reconfiguration) bandwidth by

    varying the element length and interelement separation to stay

    at or near resonance at each frequency. This does not imply an

    increase in the conventional gain-bandwidth product in whichthe bandwidth must be instantaneous. But for some applica-

    tions this distinction may not be so significant if the reconfigura-

    tion time is sufficiently small. For example, with contemporary

    MEMS switches this time would be s, which is adequate

    for many communications systems.

    The pervasiveness of military and commercial satellite com-

    munications in Ku band (1218 GHz) and the explosive growth

    of personal communications services (PCS) just above 2 GHz

    define an interesting application of a reconfigurable aperture as

    an electronically-steerable antenna that can link to space-based

    or terrestrial transceivers. This application will define the spe-

    cific frequency range and other parameters in the simulation de-

    scribed below.

    II. CANDIDATE ARCHITECTURE

    The primary purpose of this paper is to analyze the gain and

    impedance characteristics of a recap antenna in comparison to

    the diffraction limit and to a fixed-element antenna having iden-

    tical architecture and materials properties, but lacking the recon-

    figurability. To facilitate the analysis, the simple recap architec-

    ture shown in Fig. 1(a) was chosen. It consists of a square lat-

    tice of rectangular microstrip elements, each having

    0018926X/01$10.00 2001 IEEE

    Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY KURUKSHETRA. Downloaded on June 5, 2009 at 01:23 from IEEE Xplore. Restrictions apply.

    http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-
  • 7/31/2019 4 Elliott R. Brown 01

    2/6

    1358 IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 49, NO. 10, OCTOBER 2001

    (a) (b)

    Fig. 1. Recap architecture consisting of microstrip elements that can be connected by MEMS switches or to a balanced transmission line to form an array ofplanar-dipole antennas. (a) Maximum-frequency ( ) configuration where the gaps between microstrip elements are alternately opened and connected to abalanced line to form an array of approximately -long, -spaced dipoles. (b) First subharmonic configuration where 50% of the gaps are closed (by aMEMS switch), 25% are opened, and 25% connected to a balanced line to form an array -long, -spaced dipoles at .

    length along the -axis and filamentary width (i.e., )

    along the -axis, and each separated from its neighbors along

    the -axis by an infinitesimal gap (i.e., ). Located at

    each gap is a series-connected MEMS switch and a shunt con-

    nected balanced transmission line. The figure shows a top view

    of this array in the maximum-frequency configuration

    in which all the MEMS switches are left open and every other

    gap is coupled to RF so that the radiating aperture consists of

    a square lattice of microstrip dipoles having length .

    The unit cell of the square lattice is defined by the dashed box

    (width ). In principle, this array can be quite efficient and

    electronically steerable over wide angles under the conditionof half-wave resonance, , where is the

    wavelength in free space, and is the

    permittivity of the dielectric substrate material (assumed loss-

    less). Hence, the maximum frequency configuration is defined

    by .

    Fig. 1(b) shows the array in the first subharmonic

    configuration in which half of the MEMS switches are closed

    and half of the remaining gaps are driven with balanced RF

    so that the radiating aperture consists of -long microstrip

    dipoles lying on a square lattice (unit cell width ). Be-

    cause the length of the dipoles has approximately doubled com-

    pared to those in Fig. 1(a), this configuration should be reso-

    nant at a frequency . Furthermore,if the number of microstrip elements is large, lower subhar-

    monic configurations can be produced by judiciously switching

    some gaps and coupling RF to others. In each case, the recap

    array is reconfigured as a square lattice of half-wave dipoles

    with approximately half-wave center-to-center element sepa-

    ration. The lowest subharmonic frequency that yields an elec-

    tronically-steerable array, and the one called the minimum-fre-

    quency configuration, is the array defined by

    .

    As the recap is configured for subharmonic frequencies, the

    number of activated switches must increase in a manner

    described analytically by the expression .

    TABLE IPROPERTIES OF THE AND THREE SUBHARMONIC CONFIGURATIONS OF

    RECAP ARRAY of 32 16 MICROSTRIP ELEMENTS ON A SUBSTRATEHAVING AND THICKNESS mm

    For example, , 6, 14, and 30 for the 1st, 2nd, 3rd, and

    4th subharmonics, respectively. The switches have a small butsignificant value of insertion and return loss that is ultimately

    an important factor in the useful bandwidth of a recap antenna.

    III. SIMULATION PARAMETERS AND METHODOLOGY

    As suggested above, an interesting application of a recap is

    an electronically-steerable antenna that can link to terrestrial or

    space-based transceivers over a range between an of 16

    GHz and an (3rd-subharmonic) configuration of 2 GHz.

    From the relations given above, an array of microstrip

    elements will cover this range by providing a

    dipole array at 16 GHz and a dipole array at 2 GHz.

    At the intermediate subharmonic frequencies of 8 and 4 GHz,and arrays will be available that also satisfy the

    -length, -spacing of the dipoles. The characteristics

    of the maximum frequency and three subharmonic configura-

    tions are listed in Table I, along with the number of switches

    per dipole required in each configuration.

    The next parameters chosen were the permittivity and

    thickness of the substrate material. Through extensive research

    conducted in the 1980s, it was shown that high- substrates are

    usually deleterious to the performance of microstrip antennas

    (dipoles and patches in particular), because of their propensity

    for surface modes [7], [8]. It was also shown that there is

    always at least one surface mode present (the TM ) and that

    Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY KURUKSHETRA. Downloaded on June 5, 2009 at 01:23 from IEEE Xplore. Restrictions apply.

    http://-/?-http://-/?-http://-/?-http://-/?-
  • 7/31/2019 4 Elliott R. Brown 01

    3/6

    http://-/?-http://-/?-http://-/?-http://-/?-
  • 7/31/2019 4 Elliott R. Brown 01

    4/6

    http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-
  • 7/31/2019 4 Elliott R. Brown 01

    5/6

    BROWN: ON THE GAIN OF A RECONFIGURABLE-APERTURE ANTENNA 1361

    TABLE IIISELF-IMPEDANCE AND NEAREST NEIGHBOR (COLLINEAR) MUTUAL IMPEDANCE VALUES OF PLANAR MICROSTRIP DIPOLES IN FIXED AND RECAP ARRAYS

    effect of the ground plane, which at 8 GHz is now 50% closer

    to the antenna (relative to ) than it is at 16 GHz. These

    trends continue at the lower subharmonics, culminating in a

    fixed element, self-impedance at 2 GHz of 0.01212180 and

    a recap self-impedance of 1.352.69 .

    After obtaining the elements, the current components

    were computed by matrix inversion of (2) using standard Matlab

    routines. Then the antenna gain was computed from (11) using

    (8)(10) as input. Plotted in Fig. 2 is the resulting gain for

    (16 GHz) and each subharmonic configuration in comparison

    to the diffraction-limited gain at 16, 8, 4, and 2 GHz and to

    the fixed-array (dipole length cm) gain at the same fre-

    quencies. The corresponding gain values are listed in Table IV.

    The recap gain is parameterized by the MEMS switch insertion

    loss, ranging between 0.0 and 3.0 dB. For zero MEMS loss, the

    gain of the recap array falls at approximately the same rate with

    frequency as the diffraction limit, namely . Hence, the

    aperture efficiency , remains around dB. At

    first this appears surprising in light of the fact that the real part

    of the recap self-impedance drops steadily from 61.6 to about

    1 between 16 and 2 GHz. However, while the real part drops,

    the mutual impedance elements remain relatively unchanged, as

    displayed through the nearest-neighbor mutual-impedance term

    in Table III. Hence, the mutual coupling between adjacent ele-ments increases, and an increasing fraction of the input power

    to a given element is radiated by its neighbors.

    In contrast to the recap behavior, decreasing the fre-

    quency in a fixed-element array decreases the real part of the

    self-impedance while increasing the (capacitive) imaginary

    part at a comparable rate. This is evident in Table III where

    at all subharmonic frequencies the self-impedance term is

    dominated by a large capacitive reactance. Not only does this

    present a larger return loss to a 50- generator than a recap

    element, but it is much less favorable for mutual coupling. This

    point is also demonstrated in Table III where one sees that the

    self-impedance terms for the fixed-element array dominate

    the mutual impedance terms (collinear nearest neighbors) atall frequencies, making mutual coupling rather ineffective in

    transferring power from one element to its neighbors.

    Another interesting aspect of Fig. 2 pertains to the effect of

    MEMS switch loss on the gain and aperture efficiency. It is

    remarkable that no level of switch loss simulated here can re-

    duce the recap gain to the fixed-element gain. Thus, an inter-

    esting figure of merit is the level of switch loss at which the

    recap gain and the fixedaperture gain become equal. Using

    the definitions given above, the switch loss will be given by

    the expression , where

    is the gain of the recap configuration with zero switch loss.

    Solving for the insertion loss per switch, one finds

    Fig. 2. Antenna gain as a function of frequency for the recap array, afixed-element array, and a diffraction-limited aperture ( ) with a

    cm . The gain for the recap and fixed-element arrays are computedonly at the maximum frequency (16 GHz) and the first three subharmonics (8,4, and 2 GHz). The lines connecting the data points are drawn only as a visualaid. The recap gain is parameterized by the MEMS switch insertion loss ( )that ranges between 0 and 3.0 dB. The MEMS return loss is assumed to benegligible.

    . The resulting values for are 9.6, 6.4,

    and 4.2 dB at 8, 4, and 2, GHz, respectively.

    V. SUMMARY

    This paper has analyzed two-dim arrays of planar microstrip

    dipoles as the frequency is varied in octave increments between

    16 and 2 GHz. Two different array structures were considered:

    1) a fixed-element aperture in which the dipole length is con-

    stant at the maximum frequency (16 GHz)

    and 2) a recap in which the dipole length is maintained near

    over several octaves of bandwidth by judicious activation

    of MEMS switches between the elements. In both structures, the

    inter-element separation is maintained at . For the fixed-el-

    ement aperture, the gain and aperture efficiency are found to de-crease rapidly with frequency because of a rapid increase in the

    return loss arising from impedance mismatch between the gen-

    erator(s) and dipole elements. The impedance mismatch is as-

    sociated with a rapid drop in the real part of the self-impedance

    and a large (capacitive) increase in the imaginary part. And be-

    cause the mutual impedance elements all remain relatively

    small, there is little radiation into free space by either a driven

    element or its neighbors.

    In contrast, for low switch insertion loss, the recap array

    maintains a high gain with reduced frequency that nearly tracks

    the diffraction limit ( ). This is because the switches

    maintain the length near the resonance, which keeps the

    Authorized licensed use limited to: NATIONAL INSTITUTE OF TECHNOLOGY KURUKSHETRA. Downloaded on June 5, 2009 at 01:23 from IEEE Xplore. Restrictions apply.

  • 7/31/2019 4 Elliott R. Brown 01

    6/6

    1362 IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 49, NO. 10, OCTOBER 2001

    TABLE IVANTENNA GAIN FOR THE FIXED ELEMENT AND RECAP ARRAYS VERSUS FREQUENCY AND PARAMETERIZED BY MEMS SWITCH INSERTION LOSS ( )

    real part of the self-impedance relatively high and the imagi-

    nary part relatively low. In addition, the mutual impedance to

    neighboring elements, particularly the collinear term is

    relatively large and approximately constant with decreasing

    frequency. Hence, the recap maintains high efficiency with

    reduced frequency largely because each dipole maintains an

    acceptable return loss and couples increasingly to its neighbors

    in a constructive way, at least for the broadside radiation pat-

    tern. Future research will analyze these effects in the presence

    of electronic beam steering.

    In conclusion, it should be noted that the critical choice ofsubstrate permittivity ( ) in this analysis was driven

    by performance and economic considerations. Higher- mate-

    rials, such as high-resistivity Si or SIGaAs, would likely yield

    inferior performance of the recap antenna compared to that of

    Fig. 2 because of deleterious surface-wave effects. However,

    MEMS switches are now being developed primarily on such

    high- materials. Hence, application of the low- substrate

    simulated here would require that semiconductor-based MEMS

    be bonded by flip-chip or similar packaging technology. This

    approach may be feasible over the simulated frequency range

    where the size of the MEMS die should be a small fraction of a

    wavelength and, therefore, have little effect on the behavior of

    the planar antennas.

    ACKNOWLEDGMENT

    The author would like to thank T. Itoh and Y. Rahmat-Samii

    of UCLA for helpful discussions on this subject.

    REFERENCES

    [1] K. L. Virga and Y. Rahmat-Samii, Low-profile enhanced-bandwidthPIFA antennas for wireless communications packaging, IEEE Trans.

    Microwave Theory Tech., pt. 2, vol. 45, pp. 187988, Oct. 1997.[2] R. J. Mailloux, Phased Array Antenna Handbook. Norwood, MA:

    Artech House, 1994.[3] J. L. Freeman, B. J. Lamberty, and G. S. Andrews, Optoelectri-

    cally reconfigurable monopole antenna, Electron. Lett., vol. 28, pp.15021503, July, 1992.

    [4] J. H. Schaffner, R. Y. Loo, A. E. Schmitz, H. Tsung-Yuan, D. J. Hyman,A. Walston, B. A. Warneke, G. L. Tangonan, and S. W. Livingston, RFMEMS switches for tunable filters and antennas, in Proc. 3rd Int. Conf.

    Micro-Opto-Electro-Mechanical Systems, Mainz, Germany, 1999, pp.2449.

    [5] E. R. Brown, RF-MEMS switches for reconfigurable integrated cir-cuits, IEEE Trans. Microwave Theory Tech., vol. 46, pp. 18681880,Nov. 1998.

    [6] Y. Qian, B. C. C. Chang, M. F. Chang, and T. Itoh, Reconfigurableleaky-mode/multifunction patch antenna structure, Electron. Lett., vol.35, no. 21, pp. 104105, Jan. 1999.

    [7] I. E. Rana and N. G. Alexopoulos, Current distribution and inputimpedance of printed dipoles, IEEE Trans. Antennas Propagat., vol.AP-29, pp. 99105, January 1981.

    [8] P. B. Katehi and N. G. Alexopoulos, On the effect of substrate thick-ness and permittivity on printed circuit dipole properties, IEEE Trans.

    Antennas Propagat., vol. AP-31, pp. 3439, Jan. 1983.[9] N. G. Alexopoulos, P. B. Katehi, and D. B. Rutledge, Substrate

    optimization for integrated circuit antennas, IEEE Trans. MicrowaveTheory Tech., vol. MTT-31, pp. 550557, July 1983.

    [10] R. S. Elliott and G. Stern, The design of microstrip dipole arrays in-cluding mutual couplingPart I: Theory, IEEE Trans. Antennas Prop-agat., vol. Ap-29, pp. 757760, Sept. 1981.

    [11] D. M. Pozar, Analysis of finite phased arrays of printed dipoles, IEEETrans. Antennnas Propagat., vol. AP-33, pp. 10451053, Oct. 1985.

    [12] A. Malczewski, S. Eshelman, B. Pillans, J. Ehmke, and C. L. Goldsmith,X-band RF MEMS phase shifters for phased array applications, IEEE

    Microwave Guided Wave Lett., vol. 9, no. 12, pp. 51719, Dec. 1999.[13] Matlab 5.3. Natick, MA: Mathworks, Inc.[14] D. M. Pozar, Improved computational efficiency for the moment

    method solution of printed dipoles and patches, Electromagn., vol. 3,

    pp. 299309, 1983.[15] R. C. Hansen, Phased Array Antennas. New York: Wiley, 1998.

    Elliott R. Brown (M92SM97F00) received theM.S. and Ph.D. degrees in applied physics from Cali-fornia Institute of Technology, Pasadena, in 1985 and1981, respectively.

    He is a Professor of Electrical Engineering at theUniversity of California, Los Angeles (UCLA) and iscurrently conducting research projects in RF powerelectronics and thermal management, RF recon-figurable antennas, MEMS ultrasonic transducersfor biomedical imaging, shot noise suppression

    in semiconductor devices, and THz electronics and optoelectronics. Beforejoining UCLA, he was a Program Manager at DARPA in Arlington, VA. Prior

    to DARPA, he was with Massachusets Institute of Technology (MIT), LincolnLaboratory, Lexington, MA, where he conducted and managed research insolid-state science and technology.

    Dr. Brown is a member of the American Physical Society.