IE1:E TRASS.4CTIOIiS O S dSTEIiPi.%SAXD PROP.4GATIOS.

JAKVARI' 1975 137

Communications

Trees Performing as Radio Antennas

KURT IKRATH, ~ I L L 1 ~ 4 1KENYEBECK,
~1 AND
ROBERT T. HOVERTER

Abstract-Radio transmission and receptionexperiments con-

ducted in a tropical jungle are described. The performance of
conventional whip antennas is compared with the performance of
trees utilized as antennas in conjunction with hybrid electromag-
netic antenna couplers (HEMAC's). The trees were found to out-
perform the whip in some cases by up to 20 dB.

I. ISTROL)I-CTIOX
"It, would seem that living vegetation ma)- play a nlore important
part in electrical phen~nnenaThan hm been generally supposed.. --
If! as indicnted above in these experiment.~, the eart.h's surface i?
ahead>-generously provided with efficient,antennae, which m-e have
but, to utilize for cntnmunicatior~a, - -" These words were written
in 1904 b>- AIajor George 0. Scluirt, [Y.S. A h m ySignal Corps,. in a
report to the 1-.P. 1)epartnwnt o f \\-a:. in ronnection \\-it11 m i h a r y
maneuvers in the Pacific Division [I!.
In 1969, p e ~ ~ o n n of
e l the 77.8.Army Electronics Command again
employd T~W.G a. anter1n:t5. I I I thih ('&+e, the trees were used as
transmitter antennm for frequenc-iesranging from medium to short
R-avelengths. In the experiments t o be described, the t.ree t.runk was
used LLS a single-turn secondarl- winding in a resurlant toroid-type
transformer! wherein the primary winding was a flexible t.oroida1
spiral wrapped aluUnd the tree trunk.n'henstretchedout conl-
pletely, the toroid hesome.; a 24-fr long elect.rical wire antenna;
R-hm pushed together. it herrme- a miled ruugnetic I<mpantenna
of about &in di:tnwter. Hecause r d it> intrinhic electrir:d and nlagnetir
properties: the toroid \\-as given the 11ame HElIAC, an acronyn
f o r hybrid electromagnetic. antenna coupler. With 12-JF7 RF power
and n t freyuenl.ie>between 4 and 5 lJHz, signal tran.;.n~kions rang-
ing from T to 11 mi were achieved u,ing HEliAC coupled oak and
pine forest t r e e for transmis.;ion and a vertical whip antenna for
Fig. 1. H E M A C toroid coupled t.ree andPRC-74set a t jungle hole
reception. With 35-JT' R F pnwer and at frequenries of 425 kHz site. Chiva Chiva Area, Panama Canal Zone. Sept. 1971.
and 460 kHz, signal t.ransmision ranges from 30 to 35 mi were
attained using very large oak trees coupled by a HEMAC toroid
designed for the medium frequency range. Furt,hermore, as demon- 11. MEASUREMENTS
strated by H F radiation patterns from different.ly oriented natural
A. The Decays wiih Distance o j 4.650 N H z Signals Emitted from the
tree loops [2] and by %IF radiat,ion patterns from large oak trees W h i p and from Jungle Trees (rlzcy. 26, 1971)
near swampy water-filled gullies [3], the int.eraction of a toroidal
HEMAC coupled t.ree with adjacent, trees and features of the local Setup: Jungle trees located within dense underbrush vegetat,ion
terrain can be exploit,ed to launch H F and MF signal emissions into were coupled by HEhIAC's toa PRC-74 set. Used with its nhip,t,he
dEired geographic directions. set, was placeda few feet away from the same trees. S'egeetation
However, 1l1e deciduous forestsin Ken- Jersey are a poor sub- wit.hhin 1 ft of t,he whip was cut away (Figs. 1 and 2).
stitute for dense tropical jungle forests in which ferns and palms Through adjustment of the matching and tuning controls on t,he
grow rn tall as trees, and which present a great obstacle to tactical PRC74, maximal available power was delivered to t.he whip a t each
radio conlmunications by conventional whip ant.ennas. In order to location.Similarly, wit.h theHEMAC toroid coupled trees con-
evaluat,e the abi1it.y of an HElL-1C to overcome these obstacles, an nected by a series tuned matchbox to t.he P R C 7 4 set,, the tuning
impedance matchbox was designed t.0 connect a standard PRC-74 capacitor of the matchbox and the output. controls of the P R C 7 4
H F t,ransceiver with the HEXIAC coupled tree. This impedance were adjusted for relat.ive maximal RF current flon- t.hrough t.he
matchbox provided a mat.ch t o the empirically der ermined equivalent toroid; e.g., 1.00 A on the first. t.ree, 0.75 A on the second (more
series tuned load impedances of pine and oak trees ranging from distant) t,ree, and 1.00 A on thethird fmost distant) tree. The
-1.5 C2 t.o -5 C2 [4]. corresponding maximal readingx on the PRC-74 out.put. meter were
about. 50 to 70 percent of full scale.
Manuscript received January 24, 1971;.resised September 1. 1974. Signals radiated from the whip and from the t,oroid coupled jungle
The authors are mit.h the Communlcat.lons/Sutomat.ic Dat.a Process- t.rees were received with a horizontal wire dipole ant.enna and a n
ingLaboratory, U . S. Army ElectronicsCommand, Fort. hlonmouth,
S . J. 07703. HRO-500 receiver. The st,rengths of the received CR7 signals as
, .-

138 JEEE TILIXSACTIONS OK AKTENNAS A ~ PROPAGATION,

I J.LXVARY 1975

RECEIVER

c ( N 3 R T n l PiPPROX DISTANCE IN FEET c _

Fig, 3. Relative levels of 4.650-MHz CW signals radiated from re-

spectively, PRC-74 whip and HEXIBC toroid coupledjungle frees
powered bF identical PRC-74 transceiver set and received by hori-
zontal w i r e dipole antenna. Clliva Chiva Test Area. Panama Canal
Zone. Bug. 26. 1971.

Fig. 2.PRC-74 set and whip at jungle hole site, Chiva Chiva Area,
Panama Canal Zone, Sept. 1971.

displayed on the HRO-500 S-meter were measured using an HF

signal generat.or as substitution standard.
ResuZfs: Using 130 p\- as t.he zero dB reference, the signal levels
produced bJ- the toroid coupled t.rees andthe whip radiator are
plot,ted in Fig. 3 as functions of the distance from t.he horizontal
dipole receiver antenna.
Considering the spat.ial limitations in the Chiva Chiva Test Area
and t.he corresponding small number of data points, the lower at-
tenuation of signals from the toroid coupled trees could be dismissed
as a stat,ist,ical coincidence. Hoxever, previousmeasurements in
New Jersey [L']and the subsequently described measurements in
the GamboaJungle Area prove the deterministic nature of t.he Fig. 4 . Transmitter and receiver locations. R I t o Rs. Gamboa Area,
divergency between signal decays as a funct.ion of distance as meas- Panama Canal Zone. Sept. 6. 7 , and 8, 1971.
uredn-ith whips and ~ i t htrees. Similar divergencies were also
observed with different types of antennas in microwave modeling
experiments in which grasses, herbs, and shrubs mere used to model whip on Lhe road or to toroid coupled jungle trees at theedge of the
R F scat.ter in jungle forests [SI. road. -4t,ypiml receiver set,up is seen in Fig. 5, which shows t,he
toroid coupled tree on the rightside of t.he road behindthe Landrover
B. Gamboa Jungle Area Experiments vehicle.
Measurements were made at intervals of 0.5 mi up to3 mi dist.ance
Xeasurements in the Gamboa Jungle Area were carried out to along the road, i.e., from 0.4 to 2.5 mi air line distance from the
determine to Khat extent the relat,ive dryness and wetness of t.he transmitter site.
jungle vegetation affects t.he signal-versus-distance decays as sensed I n t,he initial Sept. 5 measurements when t.he vegetation was
by whip and by trees. dripping wet, after heavy rains, the toroid t,ree transmit.ter output
Xemurement setups: TheXMTRandthe receiver locat.ions was adjust,ed to produce at. a distance of 0.4 mi about the samefield
(R1 to Re) are marked on t,he map of the Gamboa Test, Area in st.rength as t.hat, produced by the PRC74 Whip Transmitker. For
Fig. 4. The 4.630 MHz signals were transmitted from t,he PRC-74 this purposet,hesignal field strength was measured with theURM-85
whip on the jungle road and from a t,oroid coupled jungle tree about whip.
30 ft. away at the side of t.he road. The signal decays were measured The sameset,tings of t,he controls of t.he toroid coupled t.ree trans-
by CW t.ransmissions from whip to whip, whip t.o trees, tree to Khip, mitter were reproduced again on Sept. 8 in hot, sunny weather, when
and t.rees t.o trees. the veget.ation was relatively dry.
Similarly, as in t,he Chiva Chiva Test Area, the radiated signals Results: The drastically differenttypes of signal decays-with-
were received and measured with t.he URM-85 WI SnalJrzer-Field distance as sensed with t.he whip and with toroid coupled trees in
Strength Meter connected t.0 either the tripod-mount,ed URM-85 the dripping wet and in the dry jungle is quantified by the corre-
COJlMUNICATIONS 139

50L TREETO TREES

i 9
WHIP TO TREES

40:- TREE TO WHIP

S+N W H I P TO WHIP
x-)

THEORETICAL r -

I , , ,I ,I I
05 I I5 2 25
AIR MILEAGE-

produce and to sense the dominant horizontal polarization, i.e., the

polarization thathasthegreater survival rate in thedominant
vertically struct,ured roughness of terrain and vegetation.

TREE TO TREES C . Qualitative Tesis

WHIP TO TREES
40- WHIP TO W H I P
Different foreign radio stat.ions were received using toroid coupled
cv, trees on the Las CruzesJungle Trail andvoice communicat,ions tests
were carried out between differentlocations on t,he LasCruses
dB
30- Jungle Trail in theGamboa Area and a location in the Chiva Chiva

1 XI-
Jungle Area. Additional det.ails of tests and related measurement
data can be found in [4].

111. CONCLUSIONS

10 - T h e phenomena that govern radio emission and reception at

ground locationsin jungle forestsare easily recognized by considering
t,he jungleas a maze of aperture-coupled screen rooms. In thejungle
I I I I 1 1 1 1 I I I I I I case, the screens, in the form of vert,ical tree and fern trunks, and
05 I I5 2 25 3 4 5 the horizontal forest canopy are of variable t.hickness, have variable
AIR MILEAGE-
shaped apertures, andare composed of diveme substances that
Fig. 6. 4.650-XIHz signal-plus-n~ise/noiseversus distance.from XMTR cont.ain mostly water.
site,junglevegetatlondrlppmg wet afterheavy ramfall, cloudy
weather, Gamboa A r e a , Panama Canal Zone, Sept.. 5 , 1971. The local forest structure determines the directivity of radiation.
However, the directional radiat.ion bias of local forest st.ructures a n
be overcome by employing phased twin tree arrays, [SI and [7].
sponding signal-plus-noise t.0 noiseratio-versus-distance curves in
Figs. 6 and 7.
These curves conform to and exp ( - 0 . 5 ~ ~ )type lawswithin ACKNOWLEDGMEKT

with thewhip than with HEMAC toroid coupled t.rees.

Considering the results of t,he previously ment,ioned modeling REFEFWXCES
- 7 .
’ 140 IEEE TRANSACTIONS ON AXTE?JJ7AS .4Am PROPbGATION,JdNU.4RY 1975

communications in dense forestcovered t.errains (Synopsis of Lebanon

State Forest X. J. experiments) U S. 4rmy Electronlcs Command
Fort 3tonmbuth. X. J., RBED ?eel;. R;p. ECOJI-3508. 9 o v . 1 9 i i
(AD i3i308).
[3] K. Skrivseth, “Signal propagationa t 400 kHz using an oak tree wit.h
a HEJIAC as an antenna.” U. S. Army Elect,ronics Command. Fort
Monmouth, X. J., R&D Tech. Rep. ECOM-3504. Nov. 1971.
41 I<. Ikrath, K . Kenuebeck, and R.T.Hoverter,“Performance of
trees as radioantennasintropicaljungleforests(PanamaCanal
Zone Experiments).” U. S. 4 r m y Electronics Command. Fort JIon-
~ ~ ~ tSen-
J ~ Jersey,
, , R&D Tech. Rep. ECOM-3534. Feb. 1972 (AD
[5]
.i-€ZZ3U).
Is.
I h a t h and G. LeXIeune, “Modeling of HF-VHF radio transmis-
.=I \ ‘ /
sion in jungle forest covered terrains,” U. S. Army Research and
Derelopment Laboratory. Fort Monmouth, X. J.. R&D Tech. Rep.
ECOXI-3512, NOT.1971 (AD 735335).
161 I<. Ikrath. K. J. Murphy. and W. Iiennebeck. “Utilization as RF-
antennas of live and of lifeless st,ructures in natural and in man made
jungles,” C . S. Army Elect.ronics Command. Fort. Monmouth,
X . J., RBED Tech. Rep. ECOM-4133. June 197? (AD 563887).
[i]K . Ikrath. W. Iienneheck. and Ii. J. Murphy. High frequenry rad!o
emission andreceptlon by treesandbyhelicopter using hybrld
electromagneticantenna couplers.” U . S . ArmsElectronics Com-
mand, Fort 3Ionmout.h. N. J., RBD Tech. Rep. ECON-41i3. Der.
1953.

Prediction of ActiveArray Impedance fromSimulator

Measurements Using Rounded Ramp Function
Interpolation
*E
x
< x
Id.x

AYDERS G. DERNER.1-D

Abstract-In order to extend the active reflection coefficient of a

phased array to general scan angles from measurements on a multi-
element phased array waveguide simulator, a newinterpolation
technique using rounded ramp functions (RRl?s) is presented. The
interpolation formulais testedon an array, simulated by a 25-element
simulator, with stripline-fed slots as radiatingelements. Contour
plots showing the magnitude and the phase of the active reflection
coefficient are included. The described interpolation scheme gives
a much smoother behavior than a Fourier series expansion pub-
lished earlier.

ISTI~OI~UCTIOX
The waveguide simulator is a powerful tool in designing phased
array antennas. Recently Gustincic[l] described a new mult.ielement.
waveguide simulatortechnique where a single simulat,or can be
used to determine the active array impedance at. anumber of
discrete scan angles. In t.he same paper an interpolation scheme to
predict the array performance at. more general scan angles is pre-
sent,ed. The scheme is given by an expansion of the array excitation
into Fourier series. With this interpolation formula the magnit.ude Fig. 2. Magnitude of active reflection coefficient.
and the phase of the activereflection coefficient of an infinite phased
array, simulated with a 2jelement waveguide simulator, have been
calculated andplotted. For scan angles more than 45” from the The function has t.he asymptotes
H-plane scan axis, an increased ripple can be observed irl the contour
plots between the measured values. Furthermore, when the array is
scanned along the E-plane axis, t.he fornlula breaks down.
In the present communication, a new interpolation scheme is
presented, where a summation of suitably shifted rounded ramp
functions (R.R.F’s) is used. The formula is then tested on t.he same
frr(x,a) =
I
z,

0,
ff-x- c3

a . x + -03.

The parameter a determines the rounded transistion between these

(2)

asymptotes. By a composition of suit.ably shifted RRF’s, the follow-

set of measurements as those used for the previously mentioned
ing interpolationformula is achieved in t,he two-dimensional direction
simulator. This new interpolationtechnique gives in the whole
cosine space:
space a much smoot.her behavior than t,he Fourier expansion formula.
N
X - bk
RRF INTERPOLATIOS k=1

T h e R R Fis defined by

where

Manuscript received June 10 1974. revised July 11 19i4.

Theauthor is with t h e Di&ion’ofNetworkTdeory.Microwave
Antenna Laboratory, Chalmers University of Technology, Got.henburg.
Sweden.