You are on page 1of 8

EPJ Web of Conferences 224, 02002 (2019) https://doi.org/10.

1051/epjconf/201922402002
MNPS-2019

Mathematical Modeling of Multi-Element Antenna Arrays with


Chiral Metamaterials Substrates Using Singular Integral
Equations
Dmitriy Klyuev1, Anatoly Neshcheret1, Oleg Osipov1, and Aleksandr Potapov2,3,*
1
Volga State University of Telecommunications and Informatics, RU-443010, Samara, Russia
2
Kotelnikov Institute of Radioengineering and Electronics, Russian Academy of Sciences, RU-125009, Moscow, Russia
3
E JNU-IREE Joint Laboratory of Fractal Method & Signal Processing, Department of Electronic Engineering, College of Information
Science and Technology, JiNan University, Guangzhou, China

Abstract. In this paper, a physical model of the multi-element antenna arrays (MEAA) has been
considered and a self-consistent numerical method for solving the problem of current distribution on the
MEAA surface with chiral metamaterials substrate has been proposed. The algorithm of the input
admittance matrix elements for chiral layer based on the conductive left and right-handed helices has been
developed and elements of the matrix surface impedances for investigated structures have been found. A set
of singular integral equations with a Cauchy kernel for calculating the current density on the surface of the
MEAA has been derived. A numerical solution of the set is a well-posed in the sense of Hadamard.
Calculations of current distribution on the MEAA surface, impedance characteristics of a three-element
antenna array and the dependence of the isolation levels between emitters on the chirality parameter have
been performed and analysed. It has been shown that the use of chiral substrates can substantially increase
the isolation between emitters.

1 Introduction Another approach involves the application of the


moments method for the electrodynamic analysis of
One of the most promising approach to the antennas and antenna array data based on chiral metamaterials [8, 9,
antenna arrays (MEAA) modernization is using of chiral 10]. The disadvantage of this approach is a need to solve
metamaterials in their design [1]. The analysis of the Fredholm integral equations of the first kind, which
literature shows that the use of metamaterials in antenna belong to the class of incorrect mathematical problems in
technology can improve their electrical and mass- the sense of Hadamard.
dimensional characteristics [2]. Research s of multi- The aim of the work is to develop a method of
element antenna arrays based on such metamaterials is electrodynamic analysis of multi-element strip antenna
of great interest because they allow increasing the arrays based on chiral metamaterials thatprovides high
isolation between emitters and, as a result, improving accuracy of calculations and requires relatively small
electromagnetic compatibility (EMC) [3]. The use of computing power.
such antenna arrays also allows increasing throughput in Based on the previous studies of single-element
MIMO systems [4]. In addition, the recent work [5] the microstrip antennas with substrates of chiral
results of a study of reducing the effect between two metamaterials [11,12,13], a system of singular integral
emitters by using metamaterials made in the form of equations (SIE) was obtained for unknown functions of
fractal structures [6,7] have been presented.. the distribution of current densities over emitters of a
However, at present time, there are a number of multi-element antenna array based on chiral
difficulties associated with electrodynamic analysis and metamaterials. The dependences of the input impedance
synthesis of such antenna arrays based on chiral of a three-element antenna array on the normalized arm
metamaterials. Most studies of such structures are length of the emitter were derived , as well as the
carried out by using electrodynamic modeling dependence of the isolation between the emitters on the
complexes, such as FEKO, CST Studio, HFSS. The use chirality parameter. It was shown that for some values
of this approach is typically associatedwith a number of of the chirality parameter, the level of isolation is higher.
difficulties . The main difficulties are the construction of The electrodynamic analysis method of multi-
correct models of the studied antennas and the choice of element antenna arrays based on chiral metamaterials is
the optimal sampling step. In addition, the software of high practical importance , because it allows
modelling these systems is typically very demanding on calculating their impedance, spatial and polarization
computational resources.

*
e-mail: potapov@cplire.ru
© The Authors, published by EDP Sciences. This is an open access article distributed under the terms of the Creative Commons Attribution License 4.0
(http://creativecommons.org/licenses/by/4.0/).
EPJ Web of Conferences 224, 02002 (2019) https://doi.org/10.1051/epjconf/201922402002
MNPS-2019

characteristics with a high accuracy at low x + Δn−1 ∈ [− an , an ] , y ∈ [− bn , bn ] .


computational costs.

3 Surface impedance matrix


2 Formulation of problem

The Fourier transform Tτ = {Tx ,Ty } of the tangential
Figure 1 shows a multi-element strip antenna array 
located on a chiral metamaterial substrate with thickness component of the electric field Eτ and the Fourier

d and with macroscopic parameters ε1 , μ1 , χ . transform of the current density F = Fx , Fy on the { }
Dielectric half-space with parameters ε 2 , μ2 is located 
emitters ηn are also connected through the matrix of
above this chiral substrate. Macroscopic parameters ε1 , surface impedances [Z] on the surfaces of the emitters
ε 2 , μ1 , μ2 , are the relative permittivities and magnetic (in plane z = d ):
permeabilities of the media, respectively.
On the surface of the chiral substrate there is an Ty   Z11 Z12   Fy 
arbitrary number of infinitely thin and perfectly T  =  Z Z 22   Fx 
, (2)
 x   21
conducting rectangular strips (emitters), the length of
where Z ij ( i, j = 1, 2 ) are the elements of the surface
which is 2ln , where n is the serial number of the
emitter. The width of these emitters is 2an and the
impedance, which are functions of the Fourier space
distance between them is Δn . The width of the emitters
variables β , h : Z=ij Z ij ( β, h ) .
gaps is 2bn .
Since it is the Fourier transforms of the field strength
The emitters are excited by harmonic sources of and surface current density that are used, the vectors of
electromotive force applied to the gaps of the emitters,  
the electric field E , magnetic field H and surface
due to which currents are distributed so that the resulting 
current density η were decomposed into Fourier
electromagnetic field satisfies the boundary conditions
(1), Maxwell's equations, and also the radiation integrals by coordinates x and y :
condition at infinity. In this case the functions of the
 
current densities distribution are continuous both on the   x0Tx ( β, h, z ) + y0Ty ( β, h, z ) +  
surfaces of the emitters and in the gaps: η n (x, y ) , where
 
 
∞ ∞
 × 
= E ( x, y, z ) ∫ ∫   + z0Tz ( β, h, z )   d βdh,
x � [− an , an ] , y � [− ln , ln ] . −∞ −∞  
× exp ( −iβx − ihy )
 

 
  x0 M x ( β, h, z ) + y0 M y ( β, h, z ) +  
 ∞ ∞

   ×

= H ( x, y, z ) ∫ ∫   + z0 M z ( β, h, z )   d βdh
−∞ −∞  
× exp ( −iβx − ihy )
 

   ( 2)  (1)
h ( x, y ) =  n0 , H ( x, y, z = d ) − H ( x, y= , z d= )

∞ ∞

∫ ∫ { x F (β, h, z ) + y F (β, h, z ) exp ( −iβx − ihy )}d βdh,


 
= 0 x 0 y
−∞ −∞
Fig. 1. Multi-element strip antenna array (3)
where
We assume that the width of the emitters is much less
than wavelength (2an << λ ) , so that the transverse  
 x0 Ex (x′, y′, z ) + y0 E y (x′, y′, z ) +  
∞ ∞
    ×
T ( β , h, z ) =
1
∫−∞ −∫∞ + z0 Ez (x′, y′, z )
components of the current densities can be neglected
 dx′dy′;
( )
ηxn = 0 . It is also assumed that the tangential 4π 2 
× exp( βx′ + ihy′) 
components of the strengths of external electric fields  
 x0 H x (x′, y′, z ) + y0 H y (x′, y′, z ) +  
n = (0, Eyn ,0 )

have one component Eτ ext ext
. ∞ ∞
    ×
M (z, β , h ) =
1
2 ∫ ∫   + z0 H z ( x , y , z )  dx′dy′;
The boundary conditions are written as follows: ′ ′
4π − ∞ − ∞  
× exp( βx′ + ihy′) 
η ny (x + Δn−1 ,−ln ) = η ny (x + Δn−1 , ln ) = 0,

F (β, h) =
1
a l
[ 
 x0 ηx (x′, y′) + y0 ηy (x′, y′) × ]
 2 ∫ ∫ dx′dy′;
Eτ n (x + Δn −1 , y ) = 0, provided x + Δn−1 ∈ [− an , an ] , 4π − a− l × exp( βx′ + ihy′) 
y ∈ [− ln ,−bn ] ∪ [bn , ln ] (1)   
(4)
In (3) x 0 , y 0 , z 0 are the unit vectors of the
 
Eτ n (x + Δn−1 , y ) = −Eτ ext
n , provided Cartesian coordinate system; {E x , E y , E z } ,

2
EPJ Web of Conferences 224, 02002 (2019) https://doi.org/10.1051/epjconf/201922402002
MNPS-2019

{H x , H y , H z } , {ηη
x, y} are components of the components in the plane z = d of region (z < d) and
 
T x( ) , T y( ) , M x( ) , M (y ) are the Fourier transforms of
2 2 2 2
vectors of the electric field E , magnetic field H and
surface current density η 
, respectively; T x , T y , T z , { } the electric and magnetic fields tangential components of
the in the plane z = d of region ( z > d ) .
{M x , M y , M z } , { F x , F y } are components of Fourier-
  First, we must define the input admittances elements
of region ( z < d ) (chiral layer). We write the Maxwell
image vectors of the electric field E , magnetic field H

and surface current density η , respectively. In (4),
 equations for the complex amplitudes in the Cartesian
considering that the surface density of electric current η
coordinate system:
is different from zero only on the surface of the MEAA:

x ∈ [ −a, a ] , y ∈ [ −l , l ] . In (4): n 0 is the unit normal ∂E z( ) ∂E y
1 (1)
−iωµ 0 µ 1H x( ) ± k χE x( ) ,
1 1
vector to the surface between the first and second region; − =
 1  2 ∂y ∂z
H ( ) , H ( ) are the magnetic field vectors of the in (1)
∂H z( ) ∂H y
1
region 1 ( z < d ) and region 2 ( z > d ) , respectively. iωε 0 ε 1E x( ) ± k χH x( ) ,
1 1
− =
∂y ∂z
To determine the surface impedance matrix [ Z ] , it is
∂E x( ) ∂E z( )
1 1
simpler to find the matrix surface admittances [Y ] : −iωµ 0µ 1H (y ) ± k χE (y ) ,
1 1
− =
∂z ∂x
(9)
∂H x( ) ∂H z( )
1 1
 F y  Y11 Y12  T y  iωε 0 ε 1E (y ) ± k χH (y ) ,
1 1
− =
 =   , (5) ∂z ∂x
 F x  Y 21 Y 22  T x 
∂E (y ) ∂E x(1)
1
−iωµ 0µ 1H z( ) ± k χE z( ) ,
1 1
− =
where Y ij ( i, j = 1, 2 ) are the elements of the matrix ∂x ∂y
surface admittances [Y ] , depending on variables β , h . ∂H (y ) ∂H x(1)
1
iωε 0 ε 1E z( ) ± k χH z( ) ,
1 1
Since the matrix of relations (2) and (5) it follows − =
∂x ∂y
that the matrix [Y ] has an inverse matrix [ Z ] and the
relationships between the elements of these matrices are: where ε 0 is the electric constant of vacuum, µ 0 is the
magnetic constant of vacuum, ε 1 , µ 1 are the chiral
Z11 ( β , h ) = Y22 ( β , h ) ∆( β , h ),
layer relative permittivity and permeability, χ is the
Z12 ( β , h ) = − Y12 ( β , h ) ∆( β , h ),
(6) relative chirality parameter, k =ω ε 0µ 0 is the wave
Z 21 ( β , h ) = − Y21 ( β , h ) ∆( β , h ),
number for a plane wave in vacuum, ω is the frequency.
Z 22 ( β , h ) = Y11 ( β , h ) ∆( β , h ),
Equations (9) are only valid for a chiral media,
consisting of a right-handed (upper arithmetic signs) and
where
left-handed (bottom arithmetic signs) helices. This is
) Y11 (β, h ) Y 22 (β, h ) − Y12 (β, h ) Y 21 (β, h ) .
∆ ( β, h= determined by the sign in the material equations for
chiral media according to the Lindell-Sihvola formalism
The surface admittances matrix elements of the plane [14,15].
z = d are determined through the input admittances We have the following equations system for the
electromagnetic field in the chiral substrate:
matrix of a region ( z > d ) (dielectric layer) Y ( 2 )  and
 
( ) E  2iωµ 0µ 1k χH = 0, (10)
  
( ) ∇ 2 E + k 2 ε 1µ 1 + χ 2
the matrix of a region ( z < d ) (chiral layer) Y  :
1
 
( ) H ± 2iωε 0ε 1k χE = 0,
  
∇ 2 H + k 2 ε 1µ 1 + χ 2
=
( 2) (1)
Y ij Y ij − Y ij . (7)
Substituting (3) into (10) and taking the derivatives
The input admittances matrices are introduced as the of x and y, we obtain the following system of equations
following expressions: { } {
for the Fourier-images T x , T y , T z , M x , M y , M z : }
 M (1,2 )  Y (1,2 ) (1,2 ) (1,2 ) 
Y12  T y  ∂ 2T
( )
 
 x  =  11  , (8) +  k 2 ε 1µ 1 + χ 2 − β 2 − h 2  T  2iωµ 0µ 1k χM = 0,
 M (1,2 )  Y (1,2 ) (1,2 )
Y 22  T x(1,2 )  ∂z 2  
 y   21 
∂ 2M
( )
 
+  k 2 ε 1µ 1 + χ 2 − β 2 − h 2  M ± 2iωε 0 ε 1k χT = 0,
where T x(1) , T y(1) , M ( ), M() ∂z  
1 1 2
x y are the Fourier
(11)
transforms of the electric and magnetic fields tangential

3
EPJ Web of Conferences 224, 02002 (2019) https://doi.org/10.1051/epjconf/201922402002
MNPS-2019

We define T y(1) and M (y1) from equation (11) βh ( sin ( γ R z ) − sin ( γ L z ) ) ±


w − ( β, h ) =
.
solutions with the boundary condition T y(1) (= ) 0:
z 0= ± q ( h ) ( γ R cos ( γ R z ) − γ L cos ( γ L z ) )

T y( ) =
The method for determining the elements of the
±C ( sin ( γ R z ) + sin ( γ L z ) ) ,
1
matrix of input admittances of the plane z = d of the
(z > d)
ε 0ε 1 (12)
M (y ) Ci ( sin ( γ R z ) − sin ( γ L z ) ) , region (dielectric layer) is similar to this
1
=
µ 0µ 1 method. In the derivation of these expressions, the
chirality parameter was assumed to be zero. Here we
where present only the final expressions:

k 2 ( n + χ ) − β 2 − h= k 2 ( n − χ) − β 2 − =
2 2
=
γR 2
, γL ( 2 ) − k ε 2µ 2 − β ; Y ( 2 ) =
2 2h2 , n hβ ε 1µ 1 ,
Y11 = − ;
(13) ωµ 0µ 2 r2 12
ωµ 0µ 2 r2
C is a constant. (15)
( 2) hβ ( 2) k 2
ε µ − h 2
=
We substitute expression (3) to (9) and express the Y 21 = ; Y 22 2 2 ;
(1) (1) (1) (1) ωµ 0µ 2 r2 ωµ 0µ 2 r2
components T x , M x through T y , M y . Using
these expressions and the relations (12), we obtain the
where = r2 k 2 ε 2µ 2 − β 2 − h 2 , ε 2 , µ 2 are the
values of the matrix input admittances elements on the
plane z = d of region ( z < d ) (chiral layer):
relative permittivity and permeability of dielectric layer.
By defining elements of the input admittances
Y11(1) ( β , h ) = −
[
a(h ) iωε 0 ε1ν + ( β , h ) − Y12(1) ( β , h )w + ( β , h ) ] matrices, it is easy to obtain elements of the matrix
sin (γ R z ) + sin (γ L z ) surface admittances and impedances of the surface plane
z = d . Using the matrix equations (2), (5), (8) and
w − (β, h)
Y12(1) ( β , h ) = − (14) performing inverse Fourier transform, we can get
iωμ 0 μ1 ν − ( β , h ) expressions relating the electric and magnetic radiation
a(h )Y22(1) ( β , h ) field with a current density at the surface. By substituting
Y21(1) ( β , h ) = w+ (β, h)
sin (γ R z ) + sin (γ L z )
these expressions them into the appropriate boundary
these conditions are transformed into integral equations
sin (γ R z ) − sin (γ L z )
Y22(1) ( β , h ) = for the current density at the MEAA surface.
iωμ0 μ1ν − ( β , h )a(h )

where
4 System of singular integral equations
−1 Taking into account the expression for the surface
 4k 4 χ 2 ε1 μ1  impedances matrix (2) we obtain the expression for the
a(h ) = 2 1 −  ;
1
(
k ε1 μ1 + k 2 χ 2 − h 2 
 ) (
k 2 ε1 μ1 + k 2 χ 2 − h 2 ) 2 

integral representation of the field of a MEAA on the
surface of the emitters:
2kχ
ζ (h ) = 2
(
k ε1 μ1 + k 2 χ 2 − h 2
;
) E py (x, y ) =
ln Δn −1 + an
2k 2 χ 2 N
s ( h)= 1 − =∑ ∫ ∫ η (x′, y ′)Z (x′, y ′, x, y )dx′dy ′,
Σ

(k )
y
; n np
2
ε 1µ 1 + k χ − h2 2 2
n =1 −ln Δn −1 − an
(16)
Z (x ′ = Δ , y ′, x = Δ , y ) =
Σ
np n −1 p −1
2k 3χε 1µ 1
q (h) = kχ −
∞ ∞

(k ) ∫ ∫ Z ( β , h)e
; 1 −iβ ( x − x′ ) −ih ( y − y ′ )
2
ε 1µ 1 + k χ − h
2 2 2 = e dβdh,
4π 2
11
− ∞− ∞

(
n + ( β, h ) = − s ( h ) ± ( γ R cos ( γ R z ) + γ L cos ( γ L z ) ) + ) ;
where p is a number of the emitter.
+z ( h ) β h ( sin ( γ R z ) + sin ( γ L z ) )
For convenience, we assume that the x-coordinate of
the location of the first emitter coincides with the origin,
i.e. equal to zero (x1 = Δ0 = 0) . Also the abscissa of
n − ( β, h ) = − s ( h ) ( γ R cos ( γ R z ) − γ L cos ( γ L z ) ) ± coordinates of other emitters location will coincide with
;
±z ( h ) βh ( sin ( γ R z ) − sin ( γ L z ) ) the distances between them (xn = Δn−1 ) .
Due to the emitters are rather narrow (a << λ ) the
(
w + ( β, h ) = −βh ± ( sin ( γ R z ) + sin ( γ L z ) )  ) transverse variation of the longitudinal distribution of the
surface current density η y (x ) can be described by the
;
( )
n
 q ( h ) ± ( γ R cos ( γ R z ) + γ L cos ( γ L z ) )
1 − (x a )
2
quasistatic distribution 1 [11,12] and the

4
EPJ Web of Conferences 224, 02002 (2019) https://doi.org/10.1051/epjconf/201922402002
MNPS-2019

distribution functions of the current densities η yn (x' , y ') The normalized system of singular integral equations
has the following form:
will be described as:

f n ( y ′)  ln N
η (x′, y ′) =
y
, (17) − σ p v p (t ) = ∑  ∫ f n′ (t ′)K np (t , t ′)dt ′ +
1 − ((x′ − Δn−1 ) / an )
n
n =1  −ln
2

(21)
1 n f ′ (t ′) 
l
where f n ( y ′) are unknown functions characterizing the (
+ a n ζ Δp −1 − Δn−1 ∫ n dt ′. )
longitudinal distribution of surface current densities at π −l t ′ − t 
n

the respective emitters.


After substituting (17) into (16) we obtain the where
following expression and after applying the integration ∞ lp
l p / an −i ξ (t −t ′ )
procedure in parts y′ to the resulting expression taking K np (t , t ′) = −
1
Δg np (ξ )e an
into account boundary condition f n (− ln ) = f n (ln ) = 0 , a 2π −∞ ξ ∫ dξ ;.

Δg np (ξ ) =
transition to new functions was performed: 1 1
×
f n′( y′) = df n ( y′) dy′ : i C εn, μ , χ
(Δn−1 − Δ p−1 )α
∞   2πa  2 n  i
× J 0 (α )− 
∫  Z11 (α, ξ ) − C εn, μ , χ | ξ | e
ln

( ) dα ;
N n an

∑ ∫ a f ′ ( y ′) ×
1
E py Δ p −1 , y = −   λ  

n n
0
n =1 −ln
(18) μ1 + μ2
Z ( β , h ) iβ (Δn −1 − Δp −1 ) −ih ( y − y′ )
∞ ∞
Cεn, μ , χ = − ;
× ∫∫ J 0 ( βa n ) 11 e e dβdhdy ′. (ε1 + ε2 )(μ1 + μ2 ) − χ 2
− ∞− ∞
ih
( )
Vp
p Δp −1 , t = E 0 p v p (t ); E 0 p =
Ey ext ;
Expression (18) shows that improper integrals over h 2b p
are also divergent. To eliminate the divergence an lp 1 lp 1
asymptotic residue was performed, i.e. in the integrands, σ p = iπ Vp ;
the addition and subtraction of the asymptotic factor λ Z c b p Cεn, μ , χ
Z11∞ (h ) was performed. The expression of the asymptotic vn (t ) are voltage profiles in the gaps of the emitters;
factor is:
Z c is characteristic resistance.
Z11 ( β , h )   → Z (h ) = Cε , μ , χ h , ∞
(19)
h →∞
11

5 Solution of a system of singular


where integral equations
iωμ0 ( μ1 + μ2 ) The obtained system of singular integral equations was
Cε , μ , χ =
(k (ε + ε )(μ + μ ) − (kχ ) )
2
1 2 1 2
2
. solved using by moments method, which with a
sufficient number of members of the approximating
series allows one to obtain correct results of calculating
After a series of mathematical transformations the characteristics of the MEAA with sufficiently high
system of singular integral representations (SIR) of the accuracy [11,12,13].
electric field was obtained, which when the boundary Unknown functions characterizing the longitudinal
conditions are substituted (1), goes over the system of distributions of surface currents on emitters were
singular integral equations (SIE) with Cauchy-type presented in the form of a series consisting of Chebyshev
features with respect to unknown functions polynomials of the first kind:
characterizing the longitudinal current distributions over
AR emitters based on chiral metamaterials: ∞
AknTk (t ′)
f n′ (t ′) = ∑ (22)
 1 ln k =1 1 − t ′2
( ) ∑
N
− Ey ext Δ p −1 , y = − a n f n′ ( y ′) ×

n =1  4π −ln
p
 In order to simplify the procedure of numerical
integration of the exponential in the kernel of expression
∞ ∞ Z ( β , h ) − Z 11∞ (h )
× ∫∫ J 0 ( βa n ) 11 × (21) we decomposed them into Bessel functions and the
−∞−∞ ih (20) 1st and 2nd kind Chebyshev polynomials:
iβ ( Δ − Δ )
× e n −1 p −1 e −ih ( y − y′ ) dβdhdy ′ + ] lp

( )
−i ξt
kJ υ ξl p / an U υ−1 (t )
)∑i
2i −υ
an
=
+
C ε, μ, χ
(
ζ Δ p −1 − Δn −1
ln

)∫ a n f n′ ( y ′) 
dy ′.
e
(ξl p / an υ =1
(23)
π y′ − y  lp


( )
−ln ξt ′

(1 + δ0,ς ) J ς ξl p / an Tς (t )
i
e an
=2 ∑
ς =0

5
EPJ Web of Conferences 224, 02002 (2019) https://doi.org/10.1051/epjconf/201922402002
MNPS-2019

1, ς = 0 at a distance of Δ1 / λ = Δ2 / λ = 1 from each other on a


where δ 0,ς =  is a Kronecker symbol.
0, ς ≠ 0 dielectric substrate, height of d / λ = 0.1 and metallized
After substituting expressions (22) and (23) into the on one side. The permittivity and permeabilities of the
system of SIE (21) and calculating in an analytical form substrate are equals to ε1 = μ1 = 1 .
a series of table integrals we obtain the following system Fig. 2 shows the results of calculating the current
of linear algebraic equations (SLAE): distribution functions for each emitter in common-mode
excitation.
N ∞
Bmp = ∑∑ Akn Ckm
pn
, (24)
n =1 k =1

where
bp / l p

1 − t 2 U m−1 (t )v p (t )dt ,
2
B = σp ∫
p

π −b
m
p /lp

1, m = k 
Ckmpn
= mi k −m+1ζ n, p ,k ,m + an ζ Δp−1 − Δn−1 ×  ( , )
0, m ≠ k 

ζ n, p ,υ,ς = Δg np (ξ )
(
J ς ξl p / an J υ ξl p / an ) ( )
−∞
∫ ξ2
dξ ,

 1 (a)
 2 , | Δp −1 − Δn−1 |< an
(
ζ Δp −1 − Δn−1 ) =  an − Δp −1 − Δn−1 ( 2
) .

0, | Δp −1 − Δn−1 |> an

The matrix of SLAE relative to unknown coefficients


Akn has the following form:

[A ] = [B ][C ]
n
k
p
m
pn −1
km (25)

Unknown functions characterizing the longitudinal


distribution of the surface current density on the emitters
are determined in accordance with the following
expression:

f n (t ) = f n′ (t ')dt ′ =
t


−1
(b)
′ (26)

Akn Tk (t ′) ∞
Akn
∑∫ ∑ U k −1 (t )
t
= dt = − 1 − t 2
−1
k =1 1 − t ′2 k =1 k

The current function is defined as follows:

f n (t )
a an

I yn (t ) = ∫ η y (x, t )dx =
n
∫ dx = an πf n (t ) . (27)
1 − (x an )
2
−a − an

In order to verify this method, we will calculate the


distribution functions of currents over the vibrators of a
three-element AR based on chiral metamaterials and
compare them with the calculation results obtained in the
FEKO software package.
(c)
6 Verification of MEAA analysis method Fig. 2. Current distribution functions for the first (a), second
(b) and third (c) vibrators with in-phase excitation
This MEAA represents three half-wave symmetric
vibrators, the length of which is
l1 / λ = l2 / λ = l3 / λ = 0.25 and the width is
7 Numerical results and discussion
a1 / λ = a2 / λ = a3 / λ = 0.025 . The vibrators are located Fig. 3 shows the dependences of the input impedance on
a three-element strip AR with a chiral metamaterial

6
EPJ Web of Conferences 224, 02002 (2019) https://doi.org/10.1051/epjconf/201922402002
MNPS-2019

substrate made on the basis of left and right spirals. AR


parameters are similar to the previous case. The chirality
parameter is χ = 0.3 . The dependencies are given for
the excitation of only one vibrator.

(a)

(a)

(b)
Fig. 4. Dependence of the isolation between the emitters of a
three-element MEAA on the chirality parameter

As it can be seen from their graphs, the decoupling


(b) values for MEAA with a substrate of chiral
Fig. 3. Dependences of the real (a) and imaginary (b) parts of metamaterials at certain values of the chirality parameter
the input impedance of the first emitter from l / λ upon are significantly higher than those for MEAA with
excitation of one vibrator dielectric.
To conclude, his study presents the method of
We have also calculated the effect of a substrate of electrodynamic analysis of multi-element strip MEAA
chiral metamaterial on the isolation between the based on chiral metamaterials. A system of singular
vibrators of three-element MEAA. Fig.4 4 shows the integral equations with respect to unknown distribution
dependence of the isolation between the emitters of a functions of current density over emitters has been
three-element AR on the chirality parameter for the cases obtained. The method was tested by comparing the
when (a) only one vibrator is excited and two vibrators obtained current distributions with the results calculated
(b) are excited.Here , index 1-2 means that the EMF in FEKO. It has been shown that the use of chiral
source is installed on the vibrator 1, and the load on the substrates in MEAA can significantly improve the
vibrator 2. Similarly, index 1-3 means the presence of isolation between the emitters.
EMF sources on vibrators 1 and 3, and the load on
vibrator 2. Parameters of the AR are equal to: Acknowledgment. The authors prepared this article under the
Δ1 / λ = Δ2 / λ = 0.25 and l / λ = 0.5 . It is important to support of the project "Leading Talents of Guangdong
Province", № 00201502 (2016-2020) in the Jinan University
note that zeros value of the chirality parameter (China, Guangzhou).
corresponds to a substrate made of a dielectric, negative
values correspond to χ substrate based on left-handed
chiral elements, and positive values correspond to right- References
handed chiral elements. 1. N. Engheta, Proc. of the 1988 URSI, 213 (1988)
2. A. Erentok, R.W. Ziolkowski, IEEE Trans. on Ant.
and Propag. 3, 691-707 (2008)
3. V.V. Badalov, S.O. Belyaev, D.A. Kopylov,
A.M. Nescheret, Ant. 11, 31-38 (2017) [in Russian]

7
EPJ Web of Conferences 224, 02002 (2019) https://doi.org/10.1051/epjconf/201922402002
MNPS-2019

4. A.N. Bespalov, A.L. Buzov, D.V. Mishin,


D.S. Klyuev, A.M. Nescheret, Radio Eng. 3, 5-11
(2019) [in Russian]
5. M. Alibakhshikenari, M. Khalily, B.S. Virdee,
C.H. See, R. Abd-Alhameed, E. Limiti, IEEE
Access, DOI 10.1109/ACCESS.2019.2899326.
6. A.A. Potapov, V.I. Grachev, Proc. PIERS 2012 in
Moscow. 1584 – 1587 (2012).
7. A.A. Potapov Proc. 7th Int. Conf. “Neural Networks
and Artificial Intelligence”. 74 – 77 (2012)
8. J. Xiang, S. He, Y. Zhang and G. Zhu, Int. J. of
Appl. Electromag. and Mech. 48(1), 33–46 (2015)
9. C. Zebiri, M. Lashab and F. Benabdelaziz, IET MW,
Ant. Propag. 5(1), 17–29 (2011)
10. C. Zebiri, S. Daoudi, F.Benabdelaziz, M. Lashab,
D. Sayad, N.T. Ali and R.A. Abd-Alhameed, Int. J.
of Appl. Electromag. and Mech. 51, 249–260 (2016)
11. M. A. Buzova, A.N. Dementyev, D.S. Klyuev, et al,
Doklady Physics. 6, 231-234 (2018)
12. A.L. Buzov, D.S. Klyuev, D.A. Kopylov,
A.M. Nescheret, Tech. Phys. Letters, 12, 1065–1068
(2018)
13. A.L. Buzov, M.A. Buzova, D.S. Klyuev,
D.V. Mishin, A.M. Neshcheret, J. of Comm. Tech.
and Elect. 11, 1259–1264 (2018)
14. I.V. Lindell, IEEE Trans Microwave Theory Tech.
40(2), 402–405 (1992)
15. V. Lindell, A.H. Sihvola, S.A. Tretyskov and
A.J. Vitanen, Media Norwood, MA: Altech House.
(1994)

You might also like