05184858

7/27/2019 05184858

http://slidepdf.com/reader/full/05184858 1/4

IEEE ANTENNAS AND WIRELESS PROPAGATION LETTERS, VOL. 8, 2009 931

Conformal Asymmetric Meandered Flare (AMF)Antenna for Body-Worn ApplicationsDimitris Psychoudakis , Senior Member, IEEE , and John L. Volakis , Fellow, IEEE

Abstract—A conformalbody-worn antenna is presented for com-munications at 300 MHz. Theantenna consists of a thin broadbandflared-dipole element printed on a thin (0.1 mm) FR4 substratewithout metallic backing or other shielding. The letter presentsthe design approach for tuning, matching, and mounting the an-tenna on the human body. Measurements are given for a humanbody phantom, and these are compared to simulations. Differentantenna positions for improved coverage are also presented.

Index Terms—Body-worn antennas, conformal antennas, UHFAntennas, wearable antennas.

I. INTRODUCTION

BODY-WORN antennas are of great interest due to the im-

pending wireless applications. However, body-worn an-

tennas are associated with several challenges due to their low

gain and higher losses. Already, several papers present planar

antennas fabricated on conductive textile material [1]–[6] for

body-worn applications. These designs range from simple body-

worn patch antennas [1]–[3] to dual-band and broadband de-

signs [4]–[6] with electromagnetic band-gap (EBG) substrates

used in some cases [6]. A common theme in all these designsis overcoming the disadvantages of human proximity. In most

designs, the radiating element is shielded from the body by em-

ploying a ground plane or an EBG backing. Such a ground

plane aims to increase antenna efficiency. However, this ap-

proach does not appear as effective since all reported wearable

antennas have measured gains close to 10 dBi ([3] does show

simulations with 5 dBi gain, but no measurements are given).

Additionally, the ground plane backings shield off the radiating

element from the high-permittivity substrate (emulated by the

body). Moreover, such ground planes lead to low-impedance

bandwidth. Therefore, it is desirable to eliminate the ground

plane for increased miniaturization and higher bandwidths (forthe same size antenna). Indeed, it was shown in [7] that without

a ground plane, the body can be used to exploit miniaturization

without significantly reducing efficiency and bandwidth. Addi-

tionally, by excluding the metallic backing, the antenna can be

thinner and more flexible.

Manuscript received May 19, 2009; revised June 18, 2009. First publishedJuly 31, 2009; current version published August 25, 2009.

The authors are with The Ohio State University, Columbus, OH 43212-1191USA (e-mail: [email protected]; [email protected]).

Color versions of one or more of the figures in this letter are available onlineat http://ieeexplore.ieee.org.

Digital Object Identifier 10.1109/LAWP.2009.2028662

In this letter, we present a wearable antenna operating in the

UHF band (300 MHz) that avoids use of the ground plane al-

together. Moreover, the antenna is printed on a flexible FR-4

substrate (0.1 mm thick) enabling ease of mounting conformal

to the body. A human body phantom was also fabricated and

used to verify gain and pattern performance.

II. ANTENNA DESIGN

The geometrical requirements for the subject antenna

for UHF operation are: width 1.5 (3.8 cm), length

( 30 cm), and ultra thin (i.e. conformal to clothing and flex-ible). The operating frequency should optimally be centered at

300 MHz for on-body deployment (ideally 280–340 MHz for

return loss 10 dB). One antenna topology suitable for such

requirements is the printed dipole due to its conformal nature

(it doesn’t require multiple layers), simplicity, and ease of fab-

rication. There is a large selection of printed dipole shapes that

could be considered. Among them are the bow-tie dipole and

the most recently available flare dipole [8]. The latter is 34.8 cm

long 21.3 cm wide and operates from 250 MHz–1.1 GHz.

That is, this design is too large for the subject requirements

outlined above. A length and width reduction for the flare

dipole is expected to reduce bandwidth, but since the originalflare dipole antenna has abundant bandwidth, we hope that the

bandwidth after the reduction will still be satisfactory.

However, the body’s presence could potentially alleviate the

aforementioned bandwidth reduction. As is well known, the

body has a v ery high relative p ermittivity value ,

implying miniaturization (viz. resonant frequency shift). On the

other hand, since the body has losses, the bandwidth is likely to

be retained. Of course, high losses imply lower efficiency (un-

avoidable in any body-worn application).

To start the design, we began with three 12 printed dipoles

(see Fig. 1). These were simulated in free space and also when

placed next to a lossy dielectric rectangular box with

(equivalent ). The latter was intended toprovide an appropriate emulation of the body muscle extracted

from the Debye model referenced in [9]. We remark that the

rightmost geometry in Fig. 1 is an asymmetric meandered flare

(AMF) dipole antenna and is an adaptation of the flare dipole.

Its advantage is that the top meandered section has additional in-

ductance (due to meandering) for miniaturization. The bottom

half is not meandered to allow the feeding cable to run along

it until it reaches the feed location at the middle. In effect, the

bottom half of the dipole serves as a balun.

The loading effect of a lossy dielectric adjacent to the printed

dipoles is depicted in Fig. 2. The free-space simulations imply

that all three antennas in Fig. 1 have almost similar responses

1536-1225/$26.00 © 2009 IEEE

7/27/2019 05184858


932 IEEE ANTENNAS AND WIRELESS PROPAGATION LETTERS, VOL. 8, 2009

Fig. 1. 1 2 long 2 1 : 5 wide printed dipole antennas suitable for body-wornapplications.

(except for the slightly increased bandwidth of the flare). How-

ever, when a lossy dielectric body is placed 1 cm away from

the antennas, the AMF dipole is effectively miniaturized due

to its longer, meandering lower section. This observation sug-

gests that the AMF dipole is a good candidate for body-worn

applications. Furthermore, Fig. 2 suggests that a size increase

of 18% would be necessary to attain a similar performance by

the bow-tie or the flare dipole antennas. Gain and pattern effects

on the dipole performance will be examined in the next section

when the human body phantom is employed.

A. AMF Balun Length

Having selected the AMF for body-worn radiation, we pro-

ceed to optimize its lower half (balun). As mentioned earlier,

the feed cable for the AMF is routed over the lower arm and

serves as a balun. Here, the balun is defined as the portion of

the coaxial cable’s outer conductor that is in electrical contact

(soldered) with the lower dipole arm (see Fig. 3). As shown in

Fig. 3, the balun length, , was varied until the best matching

was achieved. The measured return loss in absence of the body

for five balun length values is plotted and shown in Fig. 3. From

these measurements, it is clear that the optimum (in terms of

miniaturization) balun length is less than 3.4 cm for matchingto the 50- cable.

Fig. 2. Simulatedreturn loss, matched to 50 , of three printed dipoleantennasin (top) free space and (bottom) 1 cm away from a dielectric body with " =

6 2 + j 5 5 (equivalent t a n = 0 : 8 9 ).

Fig. 3. Measured free-space return loss for different balun lengths, x (defined

as the section of the cable attached to the lower half). The inset, on the left,shows a fabricated antenna prototype.

7/27/2019 05184858


PSYCHOUDAKIS AND VOLAKIS: CONFORMAL AMF ANTENNA FOR BODY-WORN APPLICATIONS 933

Fig. 4. (left) Human body FE-BI geometry used for simulations and(right) plastic human phantom used for measurements with antennas mountedon the upper body.

III. ON-BODY CHARACTERISTICS

The AMF antenna performs best (as compared to the others)

when placed near the body. Its performance was assessed viasimulations and measurements of a plastic human phantom

filled with liquid emulating human tissue.

The antenna setup in presence of the human body is depicted

in Fig. 4. The simulation model is shown to the left, and the mea-

surement model to the right (representing a liquid-filled man-

nequin). The height of the human finite element model (FEM)

was 180 cm, its width (shoulder to shoulder) was 46 cm, and its

thickness (front to back) was 18 cm. The volume mesh was cre-

ated for simulations using an in-house finite element-boundary

integral (FE-BI) code presented in [10].

The human body phantom (see Fig. 4, right) was constructed

from a retail plastic mannequin. The mannequin’s wall thick-ness varies between 3–8 mm and is made of plastic having

. To emulate the human body, the mannequin was filled with

a human-tissue-emulating liquid (leading to the pink tint of the

phantom seenin Fig.4).The liquid had a measured and

at 300 MHz. This permittivity is slightly higher than

nominal (for human muscle), but has a slightly lower .

A. Impedance Matching

The effect of the human body on input impedance and the

return loss of the antenna depend significantly on the distance

of the antenna from the skin. This is especially true for dis-

tances between antenna and body on the order of 1 or less.Simulations were carried out to quantify this effect by varying

Fig. 5. Simulated and measured return loss curves for the AMF antenna placed

at different distances from the human body model and phantom.

the distance of the antenna from the human model. Since the

phantom model has a shell thickness that varies and the shape

is not planar, this parametric study is undertaken using simula-

tions. Only one measurement was done for a 1-cm gap between

antenna and body, and this was done for validation (see Fig. 5).

It is clear that the full bandwidth is not covered in this measure-

ment as the dB bandwidth is from 270–320 MHz.

To increase the dB bandwidth, we would need to

increase the dipole width. However, this was prohibited due to

size requirements. The effect of the distance from the body isdemonstrated in the return loss curves in Fig. 5. It is clear that as

the gap decreases, the resonant frequency shifts to lower values,

but the bandwidth remains relatively unchanged. This is mainly

due to the higher losses of the human body.

B. AMF Dipole Efficiency

A major consequence of the presence of the human body is

lower efficiency (and gain). To evaluate the efficiencies of the

derived antenna, the antenna was mounted on the front of the

torso, in the center of the chest area (see Fig. 4), and the bore-

sight gain was estimated via simulations. From the curves in

Fig. 6, it is evident that the closer the antenna comes to the body,the more prominent the losses become. In fact, the pattern ex-

periences a 5-dB gain drop by moving the antenna from 1.5 to

0.25 cm away from the body.

C. Radiation Pattern

The radiation pattern is certainly affected by presence of the

human body. Specifically in Fig. 7, we note the appearance of

nulls and a major lobe in both simulations and measurements.

The agreement between simulations and measurements is note-

worthy, especially since the two setup geometries have major

differences (fewer details for the simulation geometry). Also ob-

served is that the slight asymmetry in the gap between handsand the torso manages to tilt the measured beam toward the

7/27/2019 05184858


934 IEEE ANTENNAS AND WIRELESS PROPAGATION LETTERS, VOL. 8, 2009

Fig. 6. Simulated gain curves for the AMF antenna placed at specific distancesfrom body surface.

Fig. 7. Measured and simulated single antenna patterns with the AMF antennamounted on the chest.

phantom’s left. The cause of the off-center pattern was validated

by removing the arms and remeasuring.

IV. DUAL ANTENNA CONFIGURATION

To alleviate the issue with pattern nulls, we considered two

antennas connected via a hybrid. The hybrid can support two

modes, a common mode where the signal from the two antennas

is added in-phase and a differential mode for out-of-phase recep-

tion of the two signals. Fig. 8 shows the measured patterns for

the above modes when one of the antennas is mounted on the

back and the other on the chest, as before. The single antenna

pattern is also plotted for comparison.

It is not difficult to conclude that the common mode has im-

proved coverage over the single antenna. Even more impor-

tantly, we see that the configuration of the common and differ-

ential modes alleviate reception failure altogether.

Fig. 8. Antenna pattern measurements using two AMF antennas operated incommon and differential modes.

V. CONCLUSION

An AMF antenna with balun was introduced for body-worn

applications. The antenna was compared with alternative de-

signs of same size, and body effects were addressed in carryingout the design. As expected, the antenna showed reduction in

gain and deterioration in coverage due to human body effects.

However, use of multiple antennas showed coverage improve-

ment with simple sum and difference mode reception schemes.

REFERENCES

[1] P. Salonen, Y. Rahmat-Samii, and M. Kivikoski, “Wearable antennasin the vicinity of human body,” in Proc. IEEE Antennas Propag. Soc.

Int. Symp., Jun. 2004, vol. 1, pp. 467–470.[2] P. Salonen andL. Hurme,“A novel fabricWLAN antenna forwearable

applications,” in Proc. Antennas Propag. Soc.y Int. Symp., Jun. 2003,vol. 2, pp. 700–703.

[3] C. Hertleer, H. Rogier, L. Vallozzi, and L. Van Langenhove, “A textileantenna for off-bodycommunication integrated intoprotective clothingfor firefighters,” IEEE Trans. Antennas Propag., vol. 57, no. 4, pp.919–925, Apr. 2009.

[4] T. Yang, W. Davis, and W. Stutzman, “Wearable ultra-wideband half-disk antennas,” in Proc. IEEE Antennas Propag. Soc. Int. Symp., Jul.2005, vol. 3A, pp. 500–503.

[5] P. Salonen, Y. Rahmat-Samii, H. Hurme, and M. Kivikoski, “Dual-band wearable textile antenna,” in Proc. Antennas Propag. Soc. Int.

Symp., Jun. 2004, vol. 1, pp. 463–466.[6] S. Zhuand R. Langley, “Dual-band wearable textile antenna onan EBG

substrate,” IEEE Trans. Antennas Propag., vol. 57, no. 4, pp. 926–935,Apr. 2009.

[7] D. Psychoudakis, C.-C. Chen, and J. Volakis, “Optimizing wearableUHF antennas for on-body operation,” in Proc. Antennas Propag. Soc.

Int. Symp., Jun. 2007, pp. 4184–4187.

[8] S. Koulouridis and J. Volakis, “Minimization of flare dipole via shapeoptimization and matching circuits,” in Proc. Antennas Propag. Soc.

Int. Symp., Jun. 2007, pp. 4785–4788.[9] O. P. Gandhi, B. Q. Gao, and J. Y. Chen, “A frequency-dependent fi-

nite-difference time-domain formulation for general dispersive media,” IEEE Trans. Microw. Theory Tech., vol. 41, no. 4, pp. 658–665, Apr.1993.

[10] R. Kindt, K. Sertel, E. Topsakal, and J. Volakis, “Array decompositionmethod for the accurate analysisof finite arrays,” IEEE Trans.An tennas

Propag., vol. 51, no. 6, pp. 1364–1372, Jun. 2003.

05184858

Documents