Array Antennas Introduction Athens 2009 - Tuesday
Array Antennas Introduction Athens 2009 - Tuesday
Outline
• Definition:
• An array antenna is a spatially extended collection of N similar radiating elements, and
the term "similar radiating elements" means that all the elements have the same
radiation p
patterns,, orientated in the same direction in 3D space.
p
• The elements don't have to be necessary spaced on a regular grid, but it is assumed
that they are all fed with the same frequency.
– Group of individual radiating elements
– Feed from a common terminal
– By linear networks
Array types
Arrays
y types:
yp application
pp
• Depending on it application
– Communications
• mobile
• Satellite
– Radar
– …
• D
Depending
di on it geometry
t
– Linear
– Planar
– Conformal
» Cylindrical
» Spherical
This classification depends on the position where the different elements are
p
placed:
Linear (elements in a line)
Planar (elements in a plane): rectangular (elements in a rectangular shape),
triangular
g (elements
( in a triangle
g shape,
p , circular (elements
( on concentric
circumferences)
Conformal (elements in a 3D-surface): cylinder, sphere...
• Satcom antenna
-airborne
airborne radar technology for
satellite communications
placed on the F16
• Cobra Dane
– A big antenna formed
of 34769 radiating
elements
– works at 1200 MHz
– part of the security
radar in USA
Omnidirectional Circularly
Polarized Slot Antenna Fed by a Circularly Polarized Omnidirectional
Cylindrical Waveguide for Millimeter Wave Monopole with
Identification Friend or Foe Parasitic Strip Elements for
System in the 37GHz band Identification Friend or Foe System
ANTENNA DESIGN AND MEASUREMENT TECHNIQUES - Madrid (UPM) – March 2009
Arrays
y types:
yp network (I)
The active arrays are antennas with variable phase, that allow beam steering in a variable
direction (very useful in Radar systems).
» frequency
f A/D
» time
» code ….
– Simultaneous variables patterns
LPF
IF
IFA A/D
Dig
LNA
Filter
A/D
The adaptative arrays are kind of antennas that works with active feeding modifying
instantaneously the radiation pattern depending of the signal that it receives (These antennas are
very useful in communication systems)
ANTENNA DESIGN AND MEASUREMENT TECHNIQUES - Madrid (UPM) – March 2009 16
Big arrays
Array theory
• F
Formulation
l ti condition:
diti
– Equal elements
– Equal oriented elements
x y
G G G G
E A ( r ,θ , φ ) = ∑ Ei = Ee ( r ,θ ,φ ) ∑ Ai e jk0rrˆ i
G
FA (θ ,φ ) = ∑ Ai e jk0rrˆ i
G G
E A ( r ,θ , φ ) = Ee ( r ,θ ,φ ) ⋅ FA (θ ,φ )
In function of:
The radiated field can be expressed as the product of the element field, ¾Element positions
situated in the origin,
origin by the ARRAY FACTOR FA(θ,φ). (θ φ)
¾Excitation Ai
¾Frequency
ANTENNA DESIGN AND MEASUREMENT TECHNIQUES - Madrid (UPM) – March 2009 20
Multiplication patterns principle (I)
E A (r , θ , φ ) = Ee (r , θ , φ ) ⋅ FA (θ , φ )
• The radiation pattern of an array is the product of the radiation pattern of the single
radiating element and the array factor.
•The
Th total
t t l radiated
di t d field
fi ld polarization
l i ti depends
d d only
l on the
th usedd radiating
di ti element
l t (FA is
i a
scalar value).
• The array
y factor allow to analyze
y how is the influence of the geometry
g y and the excitation
law on the radiation without considering what kind of radiating element we use.
•In big arrays, FA(θ,φ) varies more than Ee(θ,φ) does, and we can approximate the total
radiation pattern as the array factor.
factor
• Example:
p
E A (r , θ , φ ) = Ee (r , θ , φ ) ⋅ FA (θ , φ )
G
FA (θ , φ ) = ∑ An e jnk0rrˆ n = ∑ An e jnk0d cosθ = ∑ An e jnψ
n n n
DFT −1 { An }!!
•As we can see in this expression, the array factor FA is the DFT of the excitation law of the array.
•We can establish a parallelism between the studied concepts in signal processing with the arrays concepts.
g pprocessing
•While in signal g we ppass from time domain to frequency
q y spectrum,
p , in arrays
y theory
y we ppass from
spatial domain (position and excitation law) to angular spectrum (radiation pattern).
•Some of the properties that we obtain studying the DFT are:
¾The arrayy factor is a periodic
p function,, of period
p 2π,, in the ψ variable.
¾If the array is longer (electrically o in function of λ), the main lobe of it array factor will be narrower, the
same that occur with the relation between narrow pulses and narrow frequency spectrum.
¾Excitation with laws that decrease from the centre to the edges
g have lower side lobe levels in the
radiation pattern, although we have a wider main lobe in the radiation pattern.
G
FA (θ , φ ) = ∑ An e jk0rrˆ n N −1 N −1 N −1
FA (θ , φ ) = ∑ An e = ∑ an e = ∑ an e jnψ
jnk0d cosθ jn( k0 d cosθ +α )
jnα
An = an e n =0 n =0 n =0
G G ψ = kd cos θ + α
rn = ndzˆ, rˆ ⋅ rn = nd cos θ
Radiating elements with progressive phase:
⎛ Nψ ⎞
jNψ N −1 sin ⎜ ⎟
Example of a broadside array: N −1
e −1 ψ
⎝ 2 ⎠
FA (ψ ) = ∑ e jnψ =
j
omnidirectional
idi ti l antenna
t
mobile communications
used
d in
i An = 1 n=0 e jψ − 1
=e 2
⎛ψ ⎞
sin ⎜ ⎟
1.0
⎛ Nψ ⎞
⎝2⎠
⏐FAN(ψ)⏐ sin ⎜ ⎟
1 ⎝ 2 ⎠
0.8 FAN (ψ ) =
N ψ
N=2 sin
3 2
0.6 4
5
04
0.4 6
7
8
0.2
9
10
0
0 30 60 90 120 150 180
± ψ (º)
Width between nulls of principal lobes: Width of the main lobe at -3dB:
λ λ
ψ 1N = ±2π / N ⇒ k0 d cosθ1N = 2π / N ⇒ cosθ1N = Δθ −3dB = 0.886 [rad ]
Nd Nd
⎛π ⎞ λ
⇒ sin ⎜ − θ1N ⎟ = Side lobe levels SLL:
⎝2 ⎠ Nd
⎛ 3π ⎞ 1/ N
FAN ⎜ ⎟ =
⎛π ⎞ ⎛ λ ⎞ ⎝ ⎠ sin ⎛ 3π ⎞
N
Δθ = 2 ⎜ − θ1N ⎟ = 2arcsin ⎜ ⎟ ⎜ ⎟
⎝ 2 ⎠ ⎝ Nd ⎠ ⎝ 2N ⎠
2λ 2π 2
If Nd >> λ ⇒ Δθ nulls = = [rad ] SLLN ,big = ⇒ −13.46dB
Nd L 3π
ANTENNA DESIGN AND MEASUREMENT TECHNIQUES - Madrid (UPM) – March 2009 25
¾ The new variable α allow to adjust the steering direction of the main lobe of the
array.
¾The progressive phase between elements α allow to compensate for a determined
di ti off the
direction th space, the
th phase
h difference
diff associated
i t d to
t the
th propagation
ti between
b t
the generated waves for the different elements, positioning the maximum radiation
in this specified direction.
0 ≤ θ ≤ π ⇒ − k0 d + α ≤ ψ ≤ + k0 d + α
⎛ ψ⎞
sin ⎜ N ⎟
1 ⎝ 2⎠ Visible margin
FA (θ ) =
N ⎛ψ ⎞
sin ⎜ ⎟
⎝2⎠
⎛ α ⎞
Maximum in ψ=0, θ = cos ⎜ −
−1
⎟
Periodic of period 2π in ψ⎝ k0 d ⎠
Nulls in ψ=2π/N
π θmax 0 θ
• If we observe this figure, we can deduce some arrays properties that can be generalized to equispace
linear antennas:
¾ If the excitation coefficient An (except the progresive phase component) are reals and positives,
the maximum of the array factor is in ψ = 0,0 because it is in this direction where the contribution
of all the arrays elements will sum in phase.
¾ If the visible margin include ψ = 0, the maximum will be in the direction θmax= arccos(-α/k0d).
So varying the difference phase α, it allow to change the steering direction of the array main
l b This
lobe. Thi method
h d is
i actually
ll usedd in
i electronical
l i l exploration
l i radars,
d controlling
lli digitally
di i ll the
h
phase shift between the elements.
• The arrays are divided depending on it steering direction in these followings types:
¾ Broadside array: has it radiation maximum in the perpendicular plane of the array (θmax
= π/2), where α = 0 and the visible margin is –k0d < ψ < k0d.
¾ Exploration array: steer at a variable direction θmax fixed by the difference constant
phase α. The visible margin is the general one:0 ≤ θ ≤ π ⇒ − k d + α ≤ ψ ≤ + k d + α
0 0
⎛ −α ⎞
ψ = k0 d cosθ max + α = 0 ⇒ θ max = arccos ⎜ ⎟
⎝ k0 d ⎠
¾ Endfire array: has the radiation maximum in the array axis (θmax = 0 or π).
¾ For the case θmax = 0, α = -k0d and the visible margin is -2k0d < ψ < 0. in this case the
radiaton pattern is pencil beam type with the same beamwidth in the two principal planes.
planes
¾ If the distance between elements d (and/or the phase shift α) are bigger than determined
values (as d > λ , can appear “grating lobes” or diffraction lobes (emerge similar of the
) So this why
main lobe). y d = 0.6 – 0.8λ for broadside arrays
y and d = 0.4 – 0.45λ for enfire
arrays.
¾ When the elements are fed with the same amplitude |An| = 1, the array factor FA continue to
be a periodic sinc and the diagram in ψ (slide(22)) is still valid.
Broadside array
Uniform feeding
g in amplitude
p and pphase: The visible margin
g is
Maximum: ψ =0⇒ θ max = π 2
1
d=λ/2
0.8
06
0.6 d=0.75λ
d 0.75λ
0.4 d=λ
02
0.2
0
-2π -π 0 π 2π ψ
Exploration array: steer at a variable direction θmax fixed by the difference constant phase α.
0.8
0.6
2πd 2πd
− +α <ψ < +α
04
0.4 λ λ
0.2
0
ψ
-300 -200 -100 0 100 200 300
Steering direction:
⎛ λ ⎞
θ 0 = cos −1 ⎜ −α ⎟
⎝ 2π d ⎠
d=λ/2
180 0 0
180
210 330
210 330
α=120 α=180º
90
=120º 120 60 120
90
60
150 30 150 30
180 0 180 0
The endfire array is characterized to achieve a pencil beam type main lobe
Depending on the array axis:
−4π d
• Visible margin:
g <ψ < 0
λ
• Required progressive phase:
d
α = − k0 d = −2π
λ
• Width between the nulls of the main lobe:
2λ
• If Nd >> λ ⇒ BWnulls = 2 rad
Nd
• Width of the main lobe at -3dB:
• If Nd >> λ ⇒ BW 0 88λ
0.88
−3 dB = 2 rad
Nd
ANTENNA DESIGN AND MEASUREMENT TECHNIQUES - Madrid (UPM) – March 2009 33
• FA(ψ) is always a periodic function with period ψ =2π: the valid margin of
the radiation pattern is the margin with possible values of θ: between 0 y π
Graphic
ψ ≡ kd cos θ + α
representation
FA (θ , φ ) = ∑ Ai e jjiψ
Broadside
– Phase: α = 0 ψ = kd cos θ Endfire
Exploration
θ = 0, ψ (θ = 0 ) = kd + α = 0
Uniform phase
– Visible ψ = kd cos θ + α
− kd < ψ < kd α = − kd = − 2πd λ
margin: g
Progressive phase
p
– Maximum:ψ = 0 ⇒ θ max = π 2 − kd + α < ψ < kd + α − 4 πd λ < ψ < 0
• Triangular
g feeding:
g when An=[1-abs(-(n-1)/2+i)/(n/2))]
[ (( ) ) ( ))] ; for n=0 to i-1