R F

2008 IEEE INTERNATIONAL RF AND MICROWAVE CONFERENCE PROCEEDINGS December 2-

2-4, 2008, Kuala Lumpur, MALAYSIA
M 08

Smart Antenna:Weight Calculation and Side-lobe Reduction by

Unequal Spacing Technique

M. N. Md Tan, S.K. A.Rahim, M.T. Ali and T.A.Rahman

Wireless Communication Center (WCC),

University Technology Malaysia (UTM),
Skudai Johor.
mnor1408@yahoo.com, sharulkamal@fke.utm.my, mizi732002@yahoo.com, tharek@fke.utm.my

Abstract - This paper is divided into two parts. The

first part presents the study of smart antenna beam-
forming where a single beam is shaped by weighting
and summing the antenna outputs. In order to calculate
the weight of the array elements, the conventional
beam-former (known as delay and sum) has been
chosen. MATLAB software is used to calculate the
weights, and to obtain the array factor of 8 elements
linear array antenna. The second part illustrates that,
the side-lobe can be reduced by using unequal spacing
technique, but the number of elements is remained Figure 1: Block diagram of smart antenna system.
unchanged. The comparison between symmetric and
asymmetric arrangement for unequal spacing is also In Figure 1, the received signals, x(t), are multiplied by
presented. The results show that by using the a complex weight in the processor. These weighted
symmetrical arrangement, the side-lobe obtained is elements are then summed to form the array output,
much better than asymmetrical arrangement. In fact, y(t).
the side-lobe produced by asymmetrical arrangement is
worse than the equal arrangement. Array output,
M
Keywords: Smart antenna; Beam-forming; symmetrical; y(t ) = ∑ Wi*xi (t ) (1)
asymmetrical i =1

Where M = number of element

1. Introduction W=complex weight
x(t ) = antenna array output
Smart antenna is considered as an active area of
research since the last few decades. This field can be The weight vector,
applied into different areas such as radar, sonar,
W = [W1,..........,WM ]T (2)
medical imaging and also communication, especially
in wireless communications. Smart antenna system
exploits the spatial diversity among different users and Signal induced (signal vector) on all elements,
with proper signal processing techniques, this system
has shown a convincing quality in terms of increasing x ( t ) = [ x1 ( t ),......, x M ( t )]T (3 )
the coverage and capacity, mitigate multi-path fading
and enhance security and privacy. [1]. So, the output of the array become
The smart antenna term is based on the system in
which a system using an antenna array together with a
y (t ) = W H x (t ) (4 )
processor that can adjust the radiation pattern to the
desired user [2]. In other words, smart antenna is
capable to direct the maximum radiation of the antenna Where T and H denotes transposition and complex
pattern towards the signal of interest (SOI) and at the conjugate transposition.
same time, place nulls towards the signal not of For adaptive antenna, the values of these weights
interest (SNOI), thereby increasing the performance of depends on the environment changes such as the
the wireless system. A block diagram for smart location of the user, the signal to noise ratio and etc.
antenna system is shown in Figure 1. These weights are important to ensure that the main
beam is points to the correct direction.

978-1-4244-2867-0/08/$25.00 ©2008 IEEE 441

The assumption made during the simulation was that Table 1: Weight values at different angle of users with 8
the location (angle) of the desired user is already array elements (equal spacing)
determined by on of the Direction of Arrival (DOA)
algorithm at receiving site. In order to calculate the Weight Angle of desired users
(W) θ = 00 θ = 300 θ = -300
weight and obtain the radiation pattern, the user
location is entered randomly in the programming using W1 0.1250 0.1250 0.1250
MATLAB. W2 0.1250 0.0000 + 0.1250i 0.0000 - 0.1250i
W3 0.1250 -0.1250 + 0.0000i -0.1250 - 0.0000i
W4 0.1250 -0.0000 - 0.1250i -0.0000 + 0.1250i
2. Beamforming Modelling W5 0.1250 0.1250 - 0.0000i 0.1250 + 0.0000i
W6 0.1250 0.0000 + 0.1250i 0.0000 - 0.1250i
In smart antenna system, a lot of different W7 0.1250 -0.1250 + 0.0000i -0.1250 - 0.0000i
processors (beam-former) can be used in order to study W8 0.1250 -0.0000 - 0.1250i -0.0000 + 0.1250i
the performance of the system. The processors include
conventional beam-former, null-steering beam-former,
optimal beam-former and side-lobe beam-former. Table 2: Weight values at different angle of users with 8
Beam steering is the simplest form of beam-forming arrays elements (unequal spacing)
that can be achieved by a delay and sum beam-former.
It is also well known as conventional beam-former.
The weight of the beam-former are all made equal in Weight Angle of desired users
magnitude, whereas the phases are selected to steer the (W) θ = 00 θ = 300 θ =-300
main beam of the array in a particular direction θi [3]. W1 0.1250 0.1250 0.1250
The array weights (W) can be obtained by W2 0.1250 0.0000 + 0.1250i 0.0000 - 0.1250i
W3 0.1250 -0.0386 - 0.1189i -0.0386 + 0.1189i
1
W = [ a (θ i )] (5 ) W4 0.1250 0.0735 + 0.1011i 0.0735 - 0.1011i
M
W5 0.1250 -0.1011 - 0.0735i -0.1011 + 0.0735i
where M = the number of a linear array antenna W6 0.1250 0.1189 + 0.0386i 0.1189 - 0.0386i
a(θi) = steering vector at direction θi W7 0.1250 -0.1250 + 0.0000i -0.1250 - 0.0000i
W8 0.1250 -0.0000 - 0.1250i -0.0000 + 0.1250i
2πd 2πd
−j sin θi − j ( M −1) sin θ i
a(θi ) = [1, e λ ,....e λ ]T (6)
Figure 2 shows the array factor pattern for 8
where d = distance between the array elements elements linear array antenna with equal spacing
(optimum spacing is 0.5 λ ). The result showed that the
The far-zone field of a uniform linear array of identical 0
beam can be directed correctly up to ± 60 only. The
elements is equal to the product of the field of a single pattern alters when the angle approaces 900 or 2700 [5].
element and the array factor of that array [4]. That is This is the disadvantages of using linear array since the
coverage is less than 1800.
Etotal= E (single element x Array Factor (AF) ) (7)
90
1
where AF = W '*a(θi ) (8) 120
0.8
60

0.6
150 30
The results for weight calculation, the radiation pattern 0.4
and their performance in term of the Half Power 0.2
Beam-width (HPBW) and First side-lobe (FSL) for
equal and unequal spacing are presented. Table 1 and 2 180 0

shows the weight values for each element at different

angles for both equal and unequal spacing array
elements respectively. These tables show that the 210 330

weight values will change automatically when the

angle of users or the spacing between the elements are 240 300
changed. 270

Figure 2: Radiation pattern of 8 elements at different

angle of users with equal spacing

442
3. Unequal Spacing Array Element Analysis of elements for equal spacing

Number of First side-lobe (FSL) Half power beam-

If the array elements are arranged on the y-axis Elements (dB) width (HPBW)
with a uniform spacing between elements, d, this 8 -6.3991 170
arrangement is called uniform linear array as shown in 16 -6.5871 90
Figure 3. 32 -6.6452 30

The arrangement of the elements for unequal

spacing basically can be made into two techniques,
either in symmetrical arrangement or asymmetrical
arrangement. These arrangements are shown in Figure
5.

Figure 3: Uniform linear array elements

Where I = excitation current amplitude

d = spacing between element

The side-lobe level of this arrangement can be (a)

reduced by changing the arrangement in Figure 3 into
non-uniform linear array. There are two techniques to
form non-uniform linear array antenna. The first
technique is by varying the spacing between elements
but excitation amplitude is uniform. The second
technique is by varying the excitation amplitude but (b)
the spacing between elements is uniform. Unluckily, Figure 5: The arrangement of the element for unequal
the second method requires a complicated feed system spacing; (a) Symmetric (b) Asymmetric
and at the same time the beam-width also increases [6].
In this paper, the first method is chosen. The side
level will be compared between equal and unequal 3.1 Equal spacing Vs Unequal Spacing
spacing elements. The optimum equal spacing is (asymmetrical arrangement)
0.5 λ . The radiation pattern for 8, 16 and 32 elements
with equal spacing array element at 0 degrees user First comparison made is between equal spacing with
angle is shown in Figure 4. Table 3 shows the details unequal spacing with asymmetrical arrangement. The
of the differences between different numbers of results shown in Figure 7 were based on the
elements for equal spacing. It shows clearly that when asymmetrical arrangements shown in Figure 6.
the number of elements is doubled, the beam-width is
reduced by almost 50%. The side-lobe also shows
some improvement when the number of elements
increased.
0
8 elements
16 elements
(a)
-5 32 elements
Normalized radiation pattern, dB

-10

(b)
-15

Figure 6: Different unequal spacing arrangements for

-20 asymmetrical technique: (a) design a (b) design b

-25 The radiation pattern in Figure 7 shows that for

both asymmetrical arrangements, the side-lode
-30 obtained is worse than the side-lobe produced by equal
-100 -80 -60 -40 -20 0 20 40 60 80 100
Angle of Users,deg spacing array, but the HPBW for both designs a and
Figure 4: Radiation pattern with different number of
design b are reduced from 170 to 110 and 130
elements with equal spacing at 00. respectively.
Table 3: Comparison between different numbers

443
 0
0 equal spacing
equal spacing unequal spacing
unequal-design a design 2
-5
-5 unequal-design b

Normalized radiation pattern, dB

Normalized rdiation pattern,dB

-10
-10

-15
-15

-20
-20

-25
-25

-30
-30 -100 -80 -60 -40 -20 0 20 40 60 80 100
-100 -80 -60 -40 -20 0 20 40 60 80 100
Angle of Users,degrees
Angle of Users,degrees

Figure 7: Radiation pattern of 8 elements for equal Figure 9: Radiation pattern of 8 elements for equal
and unequal asymmetrical arrangement and unequal spacing at 00

3.2 Equal spacing Vs Unequal Spacing

(symmetrical arrangement)

Figure 8 and Figure 9 shows the comparison made

between equal and unequal (symmetrical arrangement) (a)
spacing for 8 elements. The arrangements of the
elements were based on the even number arrangement
as given in Figure 5(a) and equation (6) in [6]. The
graph plotted in Figure 8 and Figure 9 were based on
the position given in Figure 10 below. (b)
Figure 10: Symmetrical arrangement for even number
of elements: (a) design 1 (b) design 2
0
equal spacing
unequal spacing design 1
It can be seen obviously that between Figure 9 and
X: -89
Y: -6.243
Figure 10 if the arrangement is made carefully, the
-5
side-lobe of the linear array antenna can be enhanced
Normalized radiation pattern,dB

easily. The HPBW and first side-lobe (FSL) for 8

-10 elements with equal spacing is 170 and -6.399dB
respectively. Meanwhile, the HPBW and FSL for
-15 design 1 are 150 and -9.0693dB and for design 2, the
HPBW is 160 and FSL is -7.7073dB.
FSL in design 1 is better than design 2 where the
-20
main difference is that the side-lobe other than FLS in
design 1 is higher than design 2. In fact, the ambiguity
-25 level in design 1 is improved from -6.248dB to
-8.199dB in design 2. It is also shown that the result
-30 obtained in design 2 is better than the result obtained in
-100 -80 -60 -40 -20 0 20 40 60 80 100
Angle of Users,degrees
[7]. In [7], the FSL showed a tremendous improvement
but again, the side-lobe other than the FSL is much
Figure 8: Radiation pattern of 8 elements for both higher than the equal spacing.
equal and unequal spa

444
 Table 4 shows the comparison made between the [4] C.A. Balanis, “Antenna theory analysis and
equal spacing array element and unequal spacing for design”, John Wiley and Sons, 1997. pp. 251
both design 1 and design 2. It can be seen clearly that [5] Iclia Villordo-Jemenez, Ignacio E. Zaldivar-
although unequal spacing array required longer size, Huerta, G.M Galvan-Tejada, “An Overview of
the FSL for both designs are improved and the beam- SDMA in Communications Systems”, Circuit and
width reduced slightly. System, MWSCAS06’, 49th IEEE Conference,
Midwest Symposium, 6-9 August 2006, pp.168-
Table 4: Comparison between equal and unequal 171.
spacing element [6] R.F.Harrington, “Side-lobe Reduction by Non-
uniform Element Spacing”, IRE Transaction. On
Equal Unequal spacing
Antenna and Propagation, Vol. Ap-9, pp. 187,
spacing Design 1 Design 2 March, 1961.
Beam-width 170 150 160 [7] Cheng-Cheh Yu, “Side-lobe reduction of
-9.0693dB -7.7073dB asymmetric linear array, Electronic Letter, Vol 33,
FSL -6.399dB
(41.73%) (20.45%)
No 9, 24th April 1997.
4.7 λ 4.1 λ
Array length 3.5 λ
(34.28%) (17.14%)

4.0 Conclusion

A study of the linear array antenna in term of their

weight for each of the elements, the radiation pattern in
both the equal and unequal spacing has been carried
out. The first part of this paper shows that it is
important to calculate the weight of each element
correctly in order to direct the beam to the desired user.
The second part shows that considerable
improvement of side-lobe level can be obtained by
using symmetrical unequal spacing arrays arrangement
in comparison with equal spacing arrays having the
same number of elements and uniform excitation
amplitude. In asymmetrical arrangement, the HPBW is
reduced but the side-level is about the same level and
in fact some results shows that the side-lobe obtained
are worst than the symmetrical arrangement.

Acknowledgement

The authors would like to thank to everyone for

their helps and supports in completing this project
especially to University Technology Malaysia (UTM)
and also to Wireless Communication Center (WCC).

References
[1] Roy,R.H, “An overview of smart antenna
Technology: The next wave in Wireless
Communication”, Vol 3, Mac 1998, pp. 339-345
[2] Godara,L.C., “ Smart Antenna”, CRS Press, 2004.
pp.4
[3] Godara,L.C., “Application of Antenna Arrays to
mobile Communications, Part II: Beam-forming
and Direction of Arrival Considerations”
Proceeding of the IEEE, vol 85, No 8, August
1997, pp. 1195-1245.

445