Está en la página 1de 10

Orfanidis tiene una funci

on llamada yagi, que calcula:


1. Las corrientes de entrada de los dipolos, se consideran dipolos el reflector, los directores y el
dipolo propiamente dicho,
2. La directividad en veces
3. La relaci
on adelante atr
as en veces
y cuyos par
ametros de entrada son tres arreglos unidimensionales:
1. La Longitud de los dipolos en lambdas
2. El radio de los alambres en que se construyen los dipolos, en lambdas
3. La ubcaci
on de los dipolos a lo largo del eje x, seguramente en lambas pero no dice nada la
documentaci
on, probar esto y consultar c
odigo fuente para estar seguros
Si con los resultado obtenidos se utiliza la funci
on gain2d se puede obtener un gr
afico de la directividad, ojo que la documentaci
on habla de la funci
on array2d pero no es as.
Debe presentar un documento .m donde:
1. Sustente la parte del c
odigo fuente de yagi.m donde se aclare si la distancia entre los dipolos
est
a dada en metros, lambdas o que unidad.
2. Realice el ejemplo
ultimo d
gito de su c
edula %3 + 1 del artculo: Cheng, D.; Chen, C., Optimum element spacings for Yagi-Uda arrays. Antennas and Propagation, IEEE Transactions on
, vol.21, no.5, pp.615,623, Sep 1973. Los ejemplos se encuentran en la secci
on VII Numerical
Examples.
Nota: % es el operador que calcula el residuo de una divisi
on entera. Por ejemplo si el u
ltimo
dgito de su cedula es 6 entonces 6 %3 = 0 + 1 = 1, debe hacer el ejemplo 1.
3. Compare los resultados obtenidos por el autor en 1973 con las herramientas que tenemos yagi.m
y gain2d.m. Si piensa que no se pueden obtener resultados con estas herramientas justifique.
4. grafique los patrones de radiaci
on, para el ejemplo que le correspondi
o de los arreglos inicial y
optimizado.
5. Bono obtenga los gr
aficos que se presentan en el artculo: Fig. 2. y Fig. 4. El ejemplo 3 no
est
a graficado.

IEEE TRANSACIIONS ON ANTENNAS AND PROPAGATION, VOL.

AP-21, NO. 5, SEPTEMBER 1973

615

Optimum Element Spacings for Yagi-Uda Arrays


DAVID K. CHENG

Abstract-A method is developedfor the maximization of the


forwardgainof
a Yagi-Udaarray by adjusting the interelement
the mutual coupling
spacings. Theeffects of a bite dipole radius and
between the elements are taken into consideration. Currents in the
array elements areapproximated by three-termexpansionswith
complex coefficients which convert the governing integral equations
intosimultaneousalgebraicequations.The
array gain is maximized by the repeatedapplication of aperturbationprocedure
which converges rapidlyto yield aset of optimum, generally unequal,
element spacings. This method eliminatesthe need for a haphazard
trial-and-errorapproachorforinterpreting
a vast data collection.
Illustrative examples are given.

I. INTRODUCTION
ECAUSE of t,heir simplicity and versatility, YagiUdaantenna a.rrays [l], [a] have found many
important practical a.pplicat,ions. A conventional YagiUda array consists of a row of para.lle1straight cylindrical
dipoles, of which only the second one is driven by a source
and a.11 others are parasitic. Fig. 1 represents a typical
arrangement. The driven element no. 2 is norma.llytuned
to resonance. Element no. 1 is a reflector which is usually
longer than thedriven element, while elements no.3 to N
are directors and are shorter than the driven element.
An important performance index ofYagi-Uda arrays
is the gain, or directivity if element losses are neglected.
Arra.y directivity depends on the radius and length of the
dipole element,sas well as on the totalnumber of elements
and the spacingsbetween them. Theoretically, each of
those pa.rameters can be varied individually, making the
problem of finding an absolute optimumcombination
practically impossible. Attemptshave been madeby
various investigators t,o maximize the gain of Yagi-Uda
However, the result,s havenot
been
arrays [3)-[5].
meaningful on account of inherent rough approximations.
Ehrenspeck and Poehler [SI examined experimentally
and systemat,ically,a methodfor obtaining maximum ga,in
from a Yagi-Uda array with equally spaced directors of
equal length. They concluded that the phase velocity of
the surface wa,ve traveling along the row of directors
could be used as a. design criterion. This phase velocity
depends, of course,on the element length and spacing
paramet.ers in a very complicated way. The surfa.ce-wave
concept has also been used to calculate the phase velocity
ofinfinitelylonguniformdipole
arrayswith assumed
current distribut,ions [7], [SI, t o determine cutofffrequencies [9] and choose design parameters [lo 1, and t o
analyze long Yagi-Uda, arrays [ll].

ManuscriptreceivedDecember27,1972;revisedFebruary
23,
1973.
The aut6ore are with the Departmentof Electrical andComputer
Engineering, Syracuse Univermty, Syracuse,N.Y. 13210.

AND

C. A. CHEN

U
1

Fig. 1. Typical Yagi-Uda array.

Recently, Bojsen et al. [12] used a numerical approach


to obtain curves showing -the variation of maximum gain
with element spacing for endike half-wave dipole arrays,
of which only oneelement is excited and the rest are
parasitics center-loaded with reactances. Sinusoidalcurrent distributions were assumed and the effects of dipole
radius were neglected. Their results appear to dispute the
propriety of basing the optimumdesign solely on
travelingwave considera,tions.The nonuniformity in both the
amplitudes and the progressive phase shifts of the currents in
the elemenk of a Yagi-Uda array has been pointed out
by Thiele [13].
Morris [14] used a three-term approximation for antenna currents to solve the governing simultaneous integral equations for Yagi-Uda arrays. He obtained a large
amount of interesting da$a for typical arrays nith 2, 4,
and 8 equally spaced directors of equal length. Of particular sigficance is the evidence that the arraygain drops
sharply when the director spacing is larger than about
0.4X, whichconfirms
the findings of Ehrenspeckand
Poehler [SI.
The purposes of this paper are to demonstrate that the
forward gain can be increased by spacing the directors .
nonuniformly and to present a method for determining
the spacings needed for gain optimization. The method
employs a spacing perturbation technique [15] which
converges quite rapidly. The three-term theory developed
by King and his associates [lS] is used t o convert the
integra.1equations into a set of algebraic equations. Typical
numerical results are presented, and radiation patterns
and current distributions on the arrayelements are plotted.
The spacing perturbation technique could be used in
conjunction with the method of moments [17] which also
converts integral equations into simultaneous algebraic
equations. However, in subsectioning the array elements,
matrices of much larger dimensionswould have to be
manipulated. This limits the number of array elements
that can be handled, Furthermore, the currents on the
pa.rasitic elements depend much morecritically upon those
on other mutually coupled elements. The effects of small
errors multiply and there would be convergence problems

616

IEEE TRANSACTIONS ON ANTEXNASAND

unless more subsections than those normally required for


driven elements are taken.In using the three-term theory
the largest matrices t o be handledfor an N-element array
are of a dimension N X N and no convergence problems
are encountered.

11.

CURRENT

DISTRIBUTIONS
IN YAGI-UDAARRAY

We shall s u m m h e in this section the integral-equation


N elements of an
formulation for the currents in the
Yagi-Uda array using athree-termapproximation
for
the driven element and two terms with complexcoefficients for the para.sitic elements [E]. The N simultaneous
integral equations t o be solved are

2
N

hi

/-hi

li(Xi')

Kkia ( z k , z i ' ) dzf

PROPAGATION, SEPTEWBER

1973

and two simultaneous matrix equations for the column


) {A(3)}:
matrices of complex coefficients ( A ( 2 )and

+
[XPd(2)]{A(2)}+ [ 9 d ( 3 ) ] [ A ( 3=
) } - {\E&)}Az(~).(13)
[ W ) ] { A ( 2 ) ) [@~(3)]{A(~)j
= - ( @ ~ ( l ) J A 2 ( l ) (12)

The expression for *=a(') and the elements of the AT X 1


column matrices { @z(l) ) and {9z,J1) ) as well as those for
the N X N square matrices [ W ] , [ W ] , [9d(2)],
and
[ @ P ] a,re rather involved. They can be found in [16)
and [ls]. It is noted that when the geometrical dimen1, and
sions are known,the complex coefficientsAd'), {A(?)
{ A @ ) )can be eva,luated from (11)-(13). From (7), the
current distributions in all the elements of a Yagi-Uda
array can then bedetermined.
When the driven element (no. 2) in a Yagi-Uda array
is a half-wavedipole,
as it often is, &hz = r/2 and
preceding
cos @oh2 = 0. Some of the quantities in the
formulation mill become indeterminate. Although they
yield definite values in the l i t i n g process, an alternative
formulation is preferred in order to avoid computational
dif6culties. The integral equation (1) for the driven halfwave dipole becomes [18]

5
N

hi

1i(2i')~2S(e,Zl)

dz;

j
- 30
[iVoz(~inBo I z 1 - 1) + Oz COS 602)

(14)

with
N

(3)
Rki ( z k )

[( z k - z / ) 2

Rki(hk)

[(hk

- Z()z

+
+

bk?]'"

(4)

bki2]1'2

(5)

0 2 = j30
i=l

/
h

(6)

= a.

I n solving the simultaneous integral equations (1), the


current distributions l i (2) are assumed t o have thefollowing form:

I ; ( z < ) K 2 j ( O J x dz;.
l)

(15)

-hi

It is convenient to modify
form:
S2(1)( 2 )

bkk

( z ) in (8) t o the following

sin Bo I x

(16)

butretain Si(2)( x ) and


( x ) as they were in (9) and
(10). Using an overbar t o denote the quantities for this
modified formulation, we have, instead of (11)-( 13),

a
li(2) =

Ai(m))Xi(m) ( 2 )

(7)

m=l

with
Xi("(2) =

sin/9O(hi- I z

I)

(8)

&hi

(9)

Xi'') ( z )

= COS / 3 ~ COS

Xi(3) ( x )

= cos +/3& - cos +pohi

(10)

m d AJl) = 0 for i # 2 (parasitic elements). Substitution


of (7) in (1) and use of certain approximate relations for
the integrals involved yield

.[l - cos

(31) / Ices r+)- r+>


C)Il.

+ [1 - cos

cos

1 - sm -

(20)

CHENG AND

617

CHEN: SPACINGS FOR YAGI-UDA AFtRAYS

k P 2

k = 2

small amounts Adk and Adi, respectively. The perturbed


distance can be obtained from (4)

--here

RkiP

= [(zk

- zi')2 f

(bx-i i- A& -

Adi)2]1'2

<< Rk?.Withthis change in distance a


typical term in the integrals contained in (1) and (14)
can be mitten as

if (Adk -

h;

SiCmn)
(2:)

KkiP(Zk,Zi!)

dzl

-I& i

k
The elenlents of [ ! @ d @ ) ] and [ + d c 3 ) ] in (19) are thesame
as those for [ \ k J m ) ] in (13) for m = 2 and 3. The elements
of [6(2)]and [ & ( 3 ) ] are the same as those for [+c2)] and
[W] except when k = 2. For k = 2, we have
&(2)

IpZi(2)

(0)- ip2id'(2)

where f P 2 i ( 2 ) ( 0 ) and

=
\k2i(3)

qrti(3)

(26)

(0)- q r 2 & p 3 )

(27)

[?Pdm)-Jp

Si(m)(zi')Kzi(O,z()
dzi',

m = 2,3.

(28)

-hi

The current distributions in the e1ement.s of a YagiUda array m<th a half-wave driven element (no. 2) can
then be determinedby solving forthe complex coefficients
{ . i . ( 2 ) ) , and { A ( 3 )from
)
(17)-(19) and using (7)
with Sz(l)
( 2 ) in ( 16) replacing S P )(2) of (8).
&

111. SPACING
PERTURBATION
With a view to adjusting the elementspacingsin
a
Yagi-Uda array for ma,ximum gain, we assume that the
posit.ions of the kth a.nd the ith elements be displaced by

= [@(m)]

{ @2c1)

)P =

{Ip#}P

(0) are

hi

ezi(")(0)
=

i, m

= 1,2,3.

(30)

We note that the additional term due to spacing perturbation is proportional to the difference (Adk - A&). As
a consequence, we write the new, perturbed matrices in
(12) and (13) as
[@(m)]p

&(3)

[9?dm)]

+ [A@m)],
+ [AfPdm)],

nt = 2,3 (31)
m = 2,3 (32)

1 + { Aa2") ]
{ q az(1) 1 -!- { A*zd(') 1

(33)
(34)

{ +2(l)

where the kith elements of the square deviation matrices


and the Mh elements of the column deviation matrices
are defined with the aid of the Kronecker delta 6ki as
follows :
[A@'")]ki

[A\ka'"']ki

{ A@2(')}k

I,+ =

(Adk -

(1 - &i),

( A & - A d j ) $ k ~ ( ~ )-( l6ki)

.m = 2,3

?n.

(35)
= 2,3

(Ad&- Adz) $kzC1) (1

(36)
(37)

(Ad&- Adz) $b%(')

(38)

- 8k2)
(1 - &e).

618

s m m 1973
~

IEEE TF~ANSACIXONSON ANTENNAS AND PROPAGATION,

The expressions for the newly defined complex coefficients we have


Ijkkid(m), 4 k Z ( l ) , and yiM(1) are given in Appendix I.
{ Ai4- (1' }
The coefficients { A @)) and { A ( 3 ) }for the current terms
in (7) will also be changed.
We
write
Iid(3?1)

4,Jm),

{ A @ )+
} {bA(Z)}

(39)

{AW'jp = {A(3))+ { A(46)


i4(3)).

(40)

{A(Z)jP

Substitution of the perturbed matrices (31)-(34), (39),


and (40) into (12) and (13) yields,when second-order
deviation terms areneglected:

[O(Z)]{AA(2)}

[9(3)-J{

AA(3)}

(42)
I n view of (35)-(38), the kth element of the right-hand
side of (41) can be written as
1y

where [Pz] is an N X N square matrix defined in Appen..*


11. Similarly, a.nother N X N matrix [P3] can be
defined (also in Appendix 11) from the right-hand side
of (42), and (41) and (42) become

[o(z)-J p

[*d")]

From (43),
inversion.

( 3 q

[\Ed(3)ll
{AA(2)}

1 1{
{ A A ~

{a@))
=

['@')]

[@d@)ll

( Adi) /di << 1, exp [jfh(di


can be approximated by

}p
The perturbed current coefficients { A @ ) } pand ( ~ - l ( ~ ) can
then be determined from (39) and (40).
The preceding formulation holds for @oh;f ~ / as
2 long
as it is less than 5 ~ / 4 .In case the driven element is a
haU-wave dipole Dohz = ~ / 2 ,the matrices [O(')], [ W ] ,
[\Ed(')],
and [ \ E d ( 3 ) ] in (43) should be changed to [@)I,
[&3)],
[$d'"],
and [%d(31], respectively, as developed in
Section 11, and

IA4

(45)

{ Ad}

PEETURBED
ARRdy

+ A&) sin 8 cos 41

+ jSoAdi sin e cos 4) *exp ( jSdi sin e cos 4).

Using the perturbed current coefficients in (39) and (40),


we obtain from (47)
6 O N
E'(e,#,) 'v E(@,,#,)
+j exp ( jsdi sin g, cos 4)
Bo i-I

{ j&,( Adi) sin e cos 4


[Mi(') (e) A ;()'
Mi(3)(e) A p ]

and { A A ( 3 ) } can be found by matrix

[Fz]
ICFJ

For small Ad;, that is; for

+[ ~ p ( e )

[p34

-1

The radiation field of a spacing-perturbed linear array


at a distance Ro from a reference origin is

CPZl
{Ad). (43)

[6(3)]

IV. R A D I A ~ O NFIELDFROM

(1

[PzIddi

i=l

[@)]

1tli(3)((e)

. A A ~ ( ~ ) ] (48)
)

where E(B,4) is the radiation field of theunperturbed


array, a,nd

CaENG AND CHEN: SPACINGS FOR YAGI-UDA

619

ARRAYS

With (51), we can m i t e (48) as

Using (52), (61) ca.n be expressed ax

+ a(e,$)
= E(4+) + DIT{Adl

E'(4dJ) = E(&+)

Pi,' = Pi,
(52)

+ 2{Ad}T{B3}4- {Ad}T[ReCz]{Ad}

(62)

where the radiated power of the unperturbed array

where the superscript T denotes transposition. Equation


(52) is useful because the deviation field AE due t o spacing perturbation is expressed explicitly in terms of {Ad}.
We are now ready Do consider the problem of gain optimization of Yagi-Uda arrays by spacing perturbation.

r2a

r2*

V. GAIN OPTIMIZATION
BY SPACING
PERTURBATION
and
The gain of an a.rray in the direction (eo,&) is

rz

(53)
where Pi, is the t,ime-average input power. With spacing
perturbation E becomes E', Pi, becomes Pin', and the
perturbed gain becomes
(54)

From (52),

I Evo,do)12 = (E* + AE*) (E' + AE)


= I E 1' + 2{Ad)T{B~]
+ {Ad]'[Cl]{Ad]

{ B I }and [Cl] have previously been defined, respectively,


in (56) and (57) ;and [CZ]is a positive definite Hermitian
matrix.
The objective of gain optimization by spacing perturbation is t o find the small changes in theelement spacings
such thak the array gain in a. given direction is increased
a.nd t o repeat the process until further increases in gain
are negligible. Hence, it is essential that
AG(eo,do) = G'(eo,b) - G(@o,+o)

(66)

be positive. Substitution of (53)- (62) in (66) yields

(55)
where

{BI)= Re (E'{D*})

(56)

{D}*{D}T.

(57)

and
CCll

I n (56), Re ( E {D*)
) = real part of the product of E
a,nd the complex conjugate of the column matrix ID}. Note that thenegative sign in (68) for { B }in the numerThe AT X N square matrix [C,] is positive semi-definite, ator of AG(Bo,dJo) in (67) implies that the array gain d
and, since {Ad] is a real mat,rix, [Cl] in the last term of decrease for an improper choice of { Ad}.
( 5 5 ) can be replaced by m e Cl]. Pi,,' in (54) is
In order t o be certain that AG(&,&) slcill be positive,
we make use of a known relation in the theory of matrices
Pin' = $ R e [Voz*IzP(O)] = Pi,
(Ad)T{B2) (58)
[15], [19]. Applied to the present problem, the relation
where
asserts that if m e Cz] is positive definite, then
p In
. - -:Va Re [A2(')&(')(0)
({BITIRe C2T1(B])({AdIT[Re CZ]{A~})
A z ( ~ ) S (0)
Z ( ~ ) A z ( ~ ) X(~0()1
~ ) (59)
2 ({Ad)T(B])2. (69)

and the kth element of the column matrix (B2}


is

I n (69), the equality sign holds when

{ A d ) = a[Re CZ]-~{B}
(70)
For a.lossless array,theinput power equals the total
power ra,diated, a.nd Pin' can be mitten in an alternative where CY is a positive consta.nt. Hence, if t>hespacing
changes in { A d } are chosen, such that,
form :

{Ad] = a[Re C2]-1(2{B~] - 60G{B2))


then

(71)

620

IEEE TRANSACTIONS
AND
ON ANTENNAS

TABLE I
GAINOPTIMIZATION
FOR SIX-ELEMENT
YAGI-UDA
~
OF

(PERTURBATION

DECECMR
SPACINGS)

2h1 = 0.51X, 2h9 = 0.50% 283 = 2h4 = 2h5 = 2hs = 0.43X,


a = 0.003369X

InitiSlArray
Optimizedhray

0.250 0.310 0.310 0.310


0.310
8.06
0.250 0.336 0.398 0.310 0.407 11.81

PROPAGATION, SEPTENBER

1973

/:h,,
,[<h,
A')?\,
\

-6 - 5 4 - 3 - 2 -I 0 I 2 3 4
10-2A/V

0.15

0.05

jI
I

0 I 2 3 4 5

-6-5-4-3-2-1

-5-4-3-2 -I 0 I 2 3 4

-2-1 0 I 2 3 4

lO-'A/V

Fig. 3. Currents in optimizedsix-elementYagi-Udaarray


by
perturbation of all element epacings. 2hl = 0.51X, 2 b = 0.5OX
2h3 = 2h4 = 2h5 = 2hs = 0.43XJ u = 0.003369X. - Re(l), - -

0.5-

- 0.4-

Im(I).

0.3 0.2 -

0.1 0

15

30

45

60

75

90

105 120 135


150

165 160

DEGREES

Fig. 2. Normalized patterns of six-element Yagi-Uda arrap


(Example 1).

TABLE II
GAINOPTWATTIONFOR SIX-ELEMENT
YAGI-UDA
ARRAY
(PERTURBATTON
OF ALL ELEXENT
SPACINGS)

b2dX
Initial Array
Optimized
Array

bdX

b43/X

bdX

0.280 0.310 0.310 0.310


0.354

0.250 0.352
0.355

bdX

Gain

0.310 7.53
0.373
11.85

15

30

45

60

75 90 105 120 135 150 165 180


rp DEGREES

Fig. 4. Normalized
patterns
of six-element
Yagi-Uda
arrays
(Example 2).

The a in (71) should be sufficiently small to sat,isfy the


condition ( Adi) / d i << 1.
Equation (71) represents the requisite fmt perturbation, with which AG and G' are determinedfrom (72) 0.310X. The director spacings are to be adjusted for gain
and (66) respectively. A second perturbation can then maximization (reflector spacing fixed).
Since the driven element is a half-wave dipole, the
be performed using the new G' as the initial gain, andthe
formulation
outlined in (14)-(28) is used to determine
process repeated until further increases are negligible. It
the
current
distributions
in the array element.s, the radihas beenfound that. this iterative procedure converges
ated
field,
and
the
gain.
The ga.in of t,he initia.1 array is
rapidly a.fter a fewcycles. We t.hen have an unequally
found
t
o
be
S.06
(9.06
dB),
which checks well with the
spaced Yagi-Uda a.rray which yields a maximum gain.
result obtained by Morris [14]. Now we keep the relative
positions of the reflect.or and the driven element fixed
VII. NUblERICaL EXAMPLES
(bzl = 0.250X) and perturb the positions of the directors
I n this sect,ionwe present the computed results of three in accordance with (70) in order to increase the gain. Six
examples which illustzate the effectiveness of increasing iterations yield the result,s shown in Table I. The optithe gain or directivity of Yagi-Uda arra.ys by spacing mized array is unequally spaced and has a gain of 11.81
pcrturba,tion.
(10.72 dB), an increase of 46.5 percent (1.66 dB). The
E n m p E e 1: Six-element Yag-Uda array with a half- computed relative field intensities inthe direction of
wave driver (2h2 = 0.5OX; one reflector, 2hi = 0.51X; four maximum radiat,ion are 0.910 and 0.958, respect,ively,for
directors, 2h3 = 2h.j = 2h.5 = 2h6 = 0.43X; 0, = 0.003369X). the initial and op t,imized arrays. The normalized radiation
In the initial array, bB1= 0.25OX, ba2 = b43 = b s = bS5= patterns are shown in Fig. 2. It is seen t.hat the pattern
)

CHENG AND CHEN: SPACINGS FORYAGI-UDA

621

ARRAYS

TABLE III
GAINOPTIMIZATION
FOR TEN-WNT
YAGI-UDA
AILRAY
(PERTORBATION OF D ~ ~ E C TSPACINGS)
OR
2hl = 0.51X, 2hZ = 0.50X, 2hi = 0.43X (i = 3, 4, .**, lo), ~a = 0.003369X

Initial Array
Optimized Array

bl/x

bdX

b,,/X

bdX

b6dX

b16/h

bm/X

bpdX

blO.O/X

Gain

0.250
0.250

0.330
0.319

0,330
0.357

0.330
0.326

0.400

0.330

0.330
0.343

0.330
0.320

0.330
0.355

0.330
0.397

12.36
16.20

for t.he opt,imum array has not only a narrower main


beam but also lowersidelobes, a fact. which has been
noted previously [20].
Example 2: Six-element,Yagi-Uda array 1it.h a halfwave driver (2h2 = 0.50X; one reflector, 2hl = 0.5lX; four
directors, 2h3 = 2hl = 2h5 = 2$ = 0.43X; a = 0.003369h).
In the initial a,rray, b ~ 1= 0.280X, b32 = b.13 = b s = bG5=
0.310X. All element spacings a.re to be adjusted for gain
optimization.
The reflector spacing bP1in the initial srray is arbitrarily chosen to be 0.280X, and all ot,her element spacings are
given as 0.310X. The gain of thisinitial array is 7.53
(8.77 dB). Now all element spacings are adjusted simultaneously in t,he optimization procedure t,o increase t.he
gain. The results are summarized in Table 11. The gain
of the optimized array is 11.85 (10.74 dB), an increase of
57.3 percent (1.97 dB).
The real and imaginary pa,rts of t.he currents in the
elements of t.he optimized array withunequal spacings
are plot,ted in Fig. 3, which includes t,he effects of mutual
coupling and finite dipole radius. It is interest,ing to find
that, thedirector spacing for t,he optimized array is 0.2501,
which confirmswith what hasbeen foundby other investigators [lS]. The normalized radiation patterns for both
the initial and t,he optimized arrays are given in Fig. 4.
Again, the pat.t>ernfor t,he optimized a.rray has a narrower
mainbeamas
well as lowersidelobes. The computed
relative field intensities in the direction of ma.ximum
radia,t,ionare 0.920 and 0.976, respectively, for t,he initial
a.nd optimized arrays.
Example 3: Ten-element Yag-Uda a.rray xith a halfwa.ve driver (2ha = 0.5OX; one reflector, 2hl = 0.5lX; eight
directors, 2hi = 0.43X, i = 3,4,.. -,lo;a = 0.003369h). In
the init,ial array, bPl = 0.250X, bB = b43 =
= b10,9 =
0.3101. The director spacings are t.0 be adjusted for gain
maximization.
Wit,h t.en elements in a Yagi-Uda array, it mould be
impract,icalto use the moment method wit.h subsectioning
for numerical solut-ion. However, only 10 X 10 matrices
a.re involved in t,he present formulation. The results for
the opt,imized array are summarized inTable 111. The
calculated gain for t.he array with eight equa.llyspaced
directors is 12.36 (10.92 dB) which checks very closely
n-ith the result of hIorris [14]. The gain of the optimized
array is 16.20 (12.10 dB, an increase of 31 percent (1.18
dB). Even for this example, the t,otalcomput,ingtime for
seven iterat,ions on an IBM 370/155 computer t,ook only
about 5 min.

VIII. CONCLUSIOK
A method has been developed for the maximization of
t.he forward gain of a I'agi-Uda array by adjusting the
interelement spa.cings. The effects of a finite element
radius and the mutual
coupling beheen t.he array elements
are taken intoconsideration. A three-term expansion with
complexcoefficientsisused
to approximate the current
distribution in the elements and to convert the governing
integral equations int.0 simultaneous algebraic equations.
The array gain is ma.ximized by the repeated application
of a perturbation procedure which converges rapidly to
yield a set, of opt,imum, generally unequal, element spacings. 1llust.rative examples are given to show typical gain
increases that a.re attaina.ble with this technique.
Although the formulat,ion using t,he three-term theory
a.ppears tedious, t.he end result is in a fairly simple form.
The matrix equations need not be reformulated for different a.rrays once they have been obt.ained.The formulation
itself is on firm grounds and has been expounded in many
research a.rticles and several books. As far a.s its application to t,he present problem is concerned, the only numerically tedious part is the evaluation of definite integrals
of the type given in (24), (25), (A-7), and (A-8). The
method of momentswith subsectioning cannot convenient.ly be used here because the critica.1 dependenceof the
currents in the parasitic elements on mutual coupling
demands h e subsectmiorlingand the consequent manipulat,ion of complex matrices of very large dimensions.
The largest matrices encountered in the spacing perturbation technique using the t,hree-term theory are of a
dimension X X N for an LY-element array. Theconvergent
iterat.ive procedure yields the opt.imum spacings for maximum gain nit.hout the needfor a haphazard trial-anderror approach or for interpreting a vast dat,a collection.
APPENDIX

Expressions for 4 d r n ) 4kk2(1),


,
$ d m ) , and
(38) :
[ ; ; y l L k ) - $kid'(%)

$kBd(l)

in (35)k # i

COSPOh.k,

k = i

$kZ(')
+&)

(h.k)

- (1 - 6k2) $ k ~ a ' ( ' )

COS p&,

(-4-1)
k #2

k = 2

(A-2)

622

AND PROPAGATION, SEPTEMBER

ANTENNAS
IEEEON
TRANSACTIONS

tities represent

14)

A2 =

[l

- COS B&k]

[COS

k#2

r$) r$)] r$)


- COS

- [COS

cos B o h . k ] [ 1 - cos

1973

(A-

r*)]

(A-6)
(A-7)

and

IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATIOR, VOL.

-21,

NO.

5,

SEPTEMBER
6231973

ing maximum gain for Yagi antennas, IRE Trans. Antennas


Prupagai., vol. AP-7, pp. 379-386, Oct. 1959.
[7] D. L. Sengupta, On %hephase velocity of wave propagation
along an i d b i t e Yagj structure, f R E Trans. Antennas
Propagai., vol. AP-7, pp. 234239, July 1959.
[SI F.Serracchioli and C. A. Lev& The calculated phase velocity
of long end-fire dipolearrays, IRE Trans. Antennas Propagat.,
_ ..
vol. AP-7, pp. S42443434, Dec. 1959.
[9] L.C. She: Characteristm of propagat.ingwaves on Yagi-Uda
structure. IEEE Trans. Micruwave T h w w Tech.. vol. MTT19, pp,. 536-542, June 1971.
, Directivity and bandwidth of single-band and doubleband Yagi arrays, IEEE .Trans. Antennas Propagat. (Commun.), vol. AP-20, pp. 778-780, Nov. 1972.
R. J. Mailloux, The longYagi-Uda array, IEEE Trans.
Antmnm Propagat., voL AP-14, pp. 128-137, Mar. 1966.
J. H. Bojsen, H. Schaer-Jacobsen, E. Nilsson, and J. B.
Andersen, Ivlax1IIO(um gam of Yagi-Uda arrays, Electron.
Lett., vol. 7, no. 18, pp. 531-532, Se t. 9, 1971.
G. A. Thiele, Analysis of Yagi-&a-type antennas, IEEE
Trans. Antennas Propagat., vol. AP-17, pp. 24-31, Jan. 1969.
I. L. Morris, Optimizationofthe Yagi.Ama.y,Ph.D. dissertation,
Harvard Univ., Cambridge, Mass., 1965.
F.I. Tseng and D. I(.Cheng, Spacingperturbation techniques
for array optimization, R d w Sn., vol. 3 (New Series), pp.
__
451-457; &fay 1968.
.
R. W. P. King, R. E. Mack, and S.
S.
Sandler, Arrays of
Cylindrical Dipoles. New York: Cambridge U&v. Prees, 1968.
R. F. Harrineton. Field Comvutation bu Mument Meulods.
New York: Ivl~cmihn,1968.
D. K. Cheng and C. A. Chen, Optimum element spacings for
Yagi-Uda arrays, Syracuse Univ.,Syracuse,N.Y.,Tech.
Rep. TR-72-9, Nov. 1972.
D. K. Cheng, Optimization techniques for antenna arrays,
Proc. IEEE, vol. 59, pp.. 1664-1674, Dec. 1971.
D. L. Sengupta, On d o r m and linearly tapered long Yagi
antennas, IRE Trans. Antennas Propagat., vol. AP-11, pp.
11-17, Jan. 1960.

REFERENCES
c11 H. Yagi, Beam transmission of ultra shortwava, Proc. IRE,
vol. 16, pp. 715-741, June 1928.
c21 S. Uda and Y. Mushiake, Yagi-Udu Antenna. Tokyo, Japan:
Maruzen Co., 1954,
c31 W. Walkinshaw, Theoreticaltreatment of short Yagi aerials,
J . Inst. Ekc. Eng., vol. 93, pt. IIIA, no. 3, pp. 598-614, 1946.
c41 D. G. Reid, The gain of an idealized Yagi array, J . Inst.
Eke. Eng., vol. 93, pt. IIIA,no. 3, pp. 564-566, 1946.
C51 R. M. Fishenden and E. R. Wiblin, Design of Yagi aerials,
Proe. Inst. Elee. Eng., vol. 96, pp. 5-12, Mar. 1949.
C6 1 H. W.Ehrenspeck and H. Poehler, A new method for obtain-

A New Method for Calculating Correction


Factors for Near-Field Gain Measurements
ARTHUR C. LUDWIG

AND

Absfract-A new method is presented for calculating near-field


antenna gain correction factors directly from measured far-field
pattern data by using a spherical waveexpansion of the pattern.
This eliminates the need for any assumptions regarding antenna
aperture field distributions. The only significant assumption in the
new method is to neglect multiple scattering between the antennas.
The method is applied to the case of a horn antenna. Calculated
results are compared to direct measured results, demonstrating
The method is also compared to the
agreement to within 0.03
method of Chu and Semplak, with similar agreement. The sensitivity of the results to truncation error and noise in the data is also
investigated and contrasted to sensitivity of prior methods to errors
in the assumed field distribution.

RICHARD A. NORMAN

I. INTRODUCTION
T IS welllrnonm t,hat, t,he apparent gain of t.wo ant,ennas separated by a finitmedistance differs from t,he
gain in the limiting case of infinite separat.ion, and many
authorshave dealt m-it,h the problem of correcting for
Allof t,hese prior techniques are
this effect [1]-[7].
analytical
and
typically
involve an assumption that the
dB.
fields are known to have a specific analytic form on some
surface. Even t,hough t,he results generally agree very well
with experimental dat,a, it, is difficult to assign an exact
tolerance to the computed correction factors due to the
various assumptions used [8], [9].
It is the purpose of this paper to present an a.lt,ernate
Manuscript received January 29, 1973;revised April2, 1973. This
work was supported by NASA under Contract NAS 7-100.
a,pproach
based on the use of experimental data, rat,her
The authors are with the Jet Propulsion Laboratory, California
than an assumed field distribut,ion. This method will be
Institute of Technology, Pasadena, Calif. 91103.

También podría gustarte