Lecture 27
Array Antennas
27.1 Linear Array of Dipole Antennas
Antenna array can be designed so that the constructive and destructive interference in the far field can be used to steer the direction of radiation of the antenna, or the far-field radiation pattern of an antenna array. The relative phases of the array elements can be changed in time so that the beam of an array antenna can be steered in real time. This has important applications in, for example, air-traffic control. A simple linear dipole array is shown in Figure 27.1.
Figure 27.1: Schematics of a dipole array. To simplify the math, the far-field approximation can be used to find its far field.
First, without loss of generality, we assume that this is a linear array of Hertzian dipoles aligned on the x axis. The current can then be described mathematically as follows:
J(r0) = zˆ I l [A0δ(x0) + A1δ(x0 − d1) + A2δ(x0 − d2) + · · ·
+ AN−1δ(x0 − dN−1)]δ(y0)δ(z0) (27.1.1)
270 Electromagnetic Field Theory
27.1.1 Far-Field Approximation
The vector potential on the xy-plane in the far field, using the sifting property of delta function, yield the following equation, to be
A(r) ∼= zˆ
µIl
4πr
e−jβr ∫dr0 [A0δ(x0) + A1δ(x0 − d1) + · · · ]δ(y0)δ(z0)e jβr0·rˆ
= zˆ
µIl
4πr
e−jβr[A0 + A1e jβd1 cos φ + A2e jβd2 cos φ + · · · + AN−1e jβdN−1 cos φ ] (27.1.2)
In the above, we have assumed that the observation point is on the xy plane, or that r = ρ = xˆx + yˆy. Thus, rˆ = xˆ cos φ + yˆ sin φ. Also, since the sources are aligned on the x axis, then r0 = xˆx0, and r0 · rˆ = x0 cos φ. Consequently, e jβr0·rˆ = e jβx0 cos φ.
If dn = nd, and An = e jnψ, then the antenna array, which assumes a progressively increasing phase shift between different elements, is called a linear phase array. Thus, (27.1.2) in the above becomes
A(r) ∼= zˆ
µIl
4πr
e−jβr[1 + e j(βd cos φ+ψ) + e j2(βd cos φ+ψ) + · · ·
+ e j(N−1)(βd cos φ+ψ) ] (27.1.3)
27.1.2 Radiation Pattern of an Array
The above (27.1.3) can be summed in closed form using
∑N−1 n=0 xn = (1 − xN)/(1 − x) (27.1.4)
Then in the far field,
A(r) ∼= zˆ
µIl
4πr
e−jβr 1 − e jN(βd cos φ+ψ)
1 − e j(βd cos φ+ψ)
(27.1.5)
Ordinarily, as shown previously, E ≈ −jω(θˆAθ + φˆAφ). But since A is zˆ directed, Aφ = 0. Furthermore, on the xy plane, Eθ ≈ −jωAθ = jωAz. Therefore,
|Eθ| = |E0|
| 1 − e jN(βd cos φ+ψ) / 1 − e j(βd cos φ+ψ) |, r → ∞
= |E0|
| sin N/2 (βd cos φ + ψ) / sin 1/2 (βd cos φ + ψ) |, r → ∞ (27.1.6)
The factor multiplying |E0| above is also called the array factor. The above can be used to plot the far-field pattern of an antenna array.
Equation (27.1.6) has an array factor that is of the form |sin Nx| / |sin x|. This function appears in digital signal processing frequently, and is known as the digital sinc function. The reason why this is so is because the far field is proportional to the Fourier transform of the current. The current in this case a finite array of Hertzian dipole, which is a product of a box function and infinite array of Hertzian dipole. The Fourier transform of such a current, as is well known in digital signal processing, is the digital sinc.
Plots of |sin 3x| and |sin x| are shown as an example and the resulting |sin 3x| / |sin x| is also shown in Figure 27.2. The function peaks when both the numerator and the denominator of the digital sinc vanish. This happens when x = nπ for integer n.
Figure 27.2: Plot of the digital sinc, |sin 3x| / |sin x|.
In equation (27.1.6), x = 1/2 (βd cos φ + ψ). We notice that the maximum in (27.1.6) would occur if x = nπ, or if
βd cos φ + ψ = 2nπ, n = 0, ±1, ±2, ±3, · · · (27.1.7)
The zeros or nulls will occur at Nx = nπ, or
βd cos φ + ψ = 2nπ / N, n = ±1, ±2, ±3, · · · , n 6= mN (27.1.8)
For example,
Case I. ψ = 0, βd = π, principal maximum is at φ = ±π/2. If N = 5, nulls are at φ = ±cos−1(2n/5), or φ = ±66.4
, ±36.9
, ±113.6
, ±143.1
. The radiation pattern is seen to form lopes. Since ψ = 0, the radiated fields in the y direction are in phase and the peak of the radiation lope is in the y direction or the broadside direction. Hence, this is called a broadside array.
271 Array Antennas
Figure 27.3: The radiation pattern of a three-element array. The broadside and endfire directions of the array is also labeled
Case II. ψ = π, βd = π, principal maximum is at φ = 0, π. If N = 4, nulls are at φ = ±cos−1(n/2 − 1), or φ = ±120
, ±90
, ±60
. Since the sources are out of phase by 180
, and N = 4 is even, the radiation fields cancel each other in the broadside, but add in the x direction or the end-fire direction.
272 Electromagnetic Field Theory
Figure 27.4: By changing the phase of the linear array, the radiation pattern of the antenna array can be changed.
From the above examples, it is seen that the interference effects between the different antenna elements of a linear array focus the power in a given direction. We can use linear array to increase the directivity of antennas. Moreover, it is shown that the radiation patterns can be changed by adjusting the spacings of the elements as well as the phase shift between them. The idea of antenna array design is to make the main lobe of the pattern to be much higher than the side lobes so that the radiated power of the antenna is directed along the main lobe or lobes rather than the side lobes. So side-lobe level suppression is an important goal of designing a highly directive antenna design. Also, by changing the phase of the antenna elements in real time, the beam of the antenna can be steered in real time with no moving parts.
27.2 When is Far-Field Approximation Valid?
In making the far-field approximation in (27.1.2), it will be interesting to ponder when the far-field approximation is valid? That is, when we can approximate
e−jβ|r−r0| ≈ e−jβr+jβr0·rˆ (27.2.1)
to arrive at (27.1.2). This is especially important because when we integrate over r0, it can range over large values especially for a large array. In this case, r0 can be as large as (N − 1)d.
To answer this question, we need to study the approximation in (27.2.1) more carefully. First, we have
|r − r0|2 = (r − r0) · (r − r0) = r2 − 2r · r0 + r02 (27.2.2)
273 Array Antennas
27.2.1 Rayleigh Distance
Figure 27.5: The right half of a Gaussian beam [74] displays the physics of the near field, the Fresnel zone, and the far zone. In the far zone, the field behaves like a spherical wave.
When a wave field leaves an aperture antenna, it can be approximately described by a Gaussian beam [74] (see Figure 27.5). Near to the antenna aperture, or the near zone, it is approximately a plane wave with wave fronts parallel to the aperture surface. Far from the antenna aperture, or in the far zone, the field behaves like a spherical wave, with its typical wave front. In between is the Fresnel zone.
Consequently, after using that β = 2π/λ, for the far-field approximation to be valid, we need (27.2.8), or that
r ≫ π/λ r02 (27.2.9)
If the aperture of the antenna is of radius W, then r0 < rmax0 ∼= W and the far field approximation is valid if
r ≫ π/λ W2 = rR (27.2.10)
If r is larger than this distance, then an antenna beam behaves like a spherical wave and starts to diverge. This distance rR is also known as the Rayleigh distance. After this distance, the wave from a finite size source resembles a spherical wave which is diverging in all directions. Also, notice that the shorter the wavelength λ, the larger is this distance. This also explains why a laser pointer works. A laser pointer light can be thought of radiation from a finite size source located at the aperture of the laser pointer. The laser pointer beam remains collimated for quite a distance, before it becomes a divergent beam or a beam with a spherical wave front.
276 Electromagnetic Field Theory
In some textbooks [31], it is common to define acceptable phase error to be π/8. The Rayleigh distance is the distance beyond which the phase error is below this value. When the phase error of π/8 is put on the right-hand side of (27.2.8), one gets
β r02 / 2r ≈ π/8 (27.2.11)
Using the approximation, the Rayleigh distance is defined to be
rR = 2D2/λ (27.2.12)
where D = 2W is the diameter of the antenna aperture.
27.2.2 Near Zone, Fresnel Zone, and Far Zone
Therefore, when a source radiates, the radiation field is divided into the near zone, the Fresnel zone, and the far zone (also known as the radiation zone, or the Fraunhofer zone in optics). The Rayleigh distance is the demarcation boundary between the Fresnel zone and the far zone. The larger the aperture of an antenna array is, the further one has to be to reach the far zone of an antenna. This distance becomes larger too when the wavelength is short. In the far zone, the far field behaves like a spherical wave, and its radiation pattern is proportional to the Fourier transform of the current.
In some sources, like the Hertzian dipole, in the near zone, much reactive energy is stored in the electric field or the magnetic field near to the source. This near zone receives reactive power from the source, which corresponds to instantaneous power that flows from the source, but is return to the source after one time harmonic cycle. Hence, a Hertzian dipole has input impedance that looks like that of a capacitor, because much of the near field of this dipole is in the electric field.
The field in the far zone carries power that radiates to infinity. As a result, the field in the near zone decays rapidly, but the field in the far zone decays as 1/r for energy conservation.