Assignment 1

Q. 1 Discuss various methods of antenna array pattern synthesis.

The synthesis of planar antenna array has many practical applications. First method is
the Taguchi’s method to synthesize and optimize radiation patterns. The antenna array
synthesis occupies a prime order in the desired radiation pattern creation, than these
are realized with calculation of the excitations and positions of all antenna elements.
The planar antenna array is particularly useful in many applications such as phased
array radar, communication with satellite. We have been a new numerical global
optimization method (Taguchi`s method) based on the orthogonal array(OA) which
proved a great performance at the linear antenna array synthesis, in order to minimize
the side lobe level and eliminate the interference with computation time very quick
contrarily in the other global optimization methods. Based to this efficient method we
can use their optimized procedures in the synthesis of planar antenna array. Taguchi
method effectively reduces the number of tests required in an optimization process.
The steps taken in Taguchi's optimization can be illustrated as follows.

Ini
Desi
R Con tial
Ide
NgnT
e duc ize
ntif
inpue
d Otythe Ini Fig. 1 Flow chart of Taguchi method
texpr R De Co
tia
Ide
u opti pr Nsig T
parm nd
c eri obl e Onnti
mal
ame
liz
e
mein d uct fyer
e leve em inp
tersa u ex
th
op
t ntslSel ut m
usin
Taguchi’s method ishused andti in
ect
valu
c pe the
e synthesis of planar antenna array, A set of phase shift weights are
tim
par i
g o e ri
generated in order to
e steer
buil
OAan
es n the
aln beam towards any desired direction. Figure.2 presents the planar
ampr
me
O d a t ete ob
lev
antenna array geometry. ort
andC ntsa
p res ho h rs
con elle
ti
Consider a planar tiarraygo composed
pon
duc
ri e valanm by M × N equispace isotropic antenna elements with
t O usi dSe
ue
o
m se tnal ng n
e bui
iz tablarr p OA
con sleC
ri t an ldct
a efirm ay ri
a i aan
dt
ti atioDe
o nsig m m res or
co
e i nd po e
n exp na th
t nsri
R eri z uct og
fit
? ea
 ntss e
g lear
i rm
fun
Y t
e o ati
ra
cti
E ?
n on y
on
S
R exD
a pe
esY
E n riig
E
n g me Sn
d e nta
fitE
ne
n
ss
fu
d
nc
ti
on

Fig. 2 Geometry of a M×N array of isotropic sources for field analysis

Interelements spacing dx along the x-axis and dy along the y-axis. The planar array antenna can be
seen as a linear array of linear arrays. The total field of a M×N elements is the multiplication
of the element radiation ¯E¯o in the coordinates by the sum of the normalized excitations
currents on the array elements where AF m and AFn are the array factor for the linear array in the x-
direction and the array factor of the linear array in the y-direction, respectively.

The second method for synthesising antenna array patterns with controlled side lobes and
enhanced directivity is proposed. It relies on the relationship between the array factor and its
corresponding element excitations through Fourier transform properties. Unlike the standard array
synthesis- based Fourier transform methods, the proposed method enforces some specific
constraints to achieve the desired goals. It involves the computation of the radiation pattern from
initial element excitations of a linear array using inverse fast Fourier transform (FFT). Once the
radiation pattern has been computed, theN constraints are applied to modify it by forcing a wide null
within the whole main beam region and preserving all the side lobes unchanged. By doing this, the
M
side lobe structures of both original and modified patterns are exactly kept same. Then, a new array
pattern with ideally no side lobes can be obtained by subtracting the original array pattern from the
modified pattern. Consider a linear array of N isotropic elements which are separated uniformly by a
distance 𝑑. The relation between antenna's radiation and its corresponding element excitations can
be written as

𝐴𝐹 (𝑢) = ∑𝑁−1 𝑎𝑛𝑒𝑗𝑛𝜓

𝑛=0

where 𝜓 = (2𝜋𝑑⁄𝜆)𝑢, 𝑢 = 𝑠𝑖𝑛(𝜃), 𝜃 represents the angle with respect to the direction normal to the
array axis, and 𝑎𝑛 is the excitation coefficients of the array elements. The above-mentioned equation
forms a finite Fourier series that relates the antenna's radiation pattern and its corresponding element
excitations 𝑎𝑛 through a discrete inverse Fourier transform (IFT). Alternatively, the element
excitation can be obtained by applying a discrete Fourier transform on the given array factor (AF).
These two Fourier transform relationships are repeatedly exploited with some constraints to
construct the required cancellation pattern which is used for side lobe cancellation or reduction.
The construction process starts with the calculation of the array factor (AF), according to above
equation using an initial set for the N element excitation coefficients (for example, uniform
amplitude distribution). This calculation is carried out with a K-point inverse FFT. To achieve
better resolution, the value of K is usually chosen to be much larger than the number of the array
elements N. In order to perform such calculations correctly, zero padding with the initial element
excitations, should be used. For example, a vector of size K should be generated with its first N
values being the initial excitations for n=1...N, and the rest of the elements are chosen to be zero
(zero padding) for n=N+1...K. By this way, an exact cancellation pattern is computed in K points
with the inverse FFT. This is followed by the first-null-to-null-beam- width (FNNBW) calculation
and enforcing the magnitude of the cancellation pattern to be zero only within the whole range of
the FNNBW, while the side lobe levels are left unchanged. After notching out the main lobe, a
direct K-point FFT is performed on the resultant pattern of notched main lobe to get a new set of
excitation coefficients. Figure 1 shows the exact cancellation pattern and the corresponding
element. excitations for used parameters N=20 and K=512. For comparison, the uniformly excited
array pattern, i.e., initial pattern, under the same parameters is also shown in Figure 3. Then, the
overall array pattern can be obtained by subtracting the original array pattern from the approximate
cancellation pattern as follows

Figure 3.

The third method of antenna array pattern synthesis algorithm is the arbitrary arrays based on
coordinate descent method (CDM). With this algorithm, the complex element weights are found to
minimize a weighted L2 norm of the difference between desired and achieved pattern. Compared
with traditional optimization techniques, CDM is easy to implement and efficient to reach the
optimum solutions. Main advantage is the flexibility. CDM is suitable for linear and planar array
with arbitrary array elements on arbitrary positions. With this method, we can configure arbitrary
beam pattern, which gives it the ability to solve variety of beam forming problem, e.g. focused
beam, shaped beam, nulls at arbitrary direction and with arbitrary beam width. CDM is applicable
for phase-only and amplitude-only arrays as well, and furthermore, it is a suitable method to treat
the problem of array with element failures.

Enhanced Practical Swarm Optimization (EPSO) algorithm can also be used to pattern synthesis of
linear arrays. Updating formulas of global best particle position and velocity can be modified to
improve the convergence accuracy of classical Practical Swarm Optimization. The developed EPSO
is tested and compared by one of the researchers with a standard benchmark to be validated as an
efficient optimization tool for beam forming applications. Different numerical examples are
presented to illustrate the capability of EPSO for pattern synthesis with a prescribed wide nulls
locations and depths. Collective multiple deep nulls approach and direct weights perturbations
approach are considered to obtain adaptive wide null steering subject to peak side lobe level and
minimum main beam width constraints. Starting from initial Chebyshev pattern, single or multiple
wide nulls are achieved by optimum perturbations of elements current amplitude or complex weights
to have either symmetric or asymmetric nulls about the main beam. Proper formation of the cost
function is presented for all case studies as a key factor to include the pattern constraints in the
optimization process.