Aerospace and Electronic Systems Magazine December 2017 - 40

High-Latitude Application of 3D OTHR Radar
pattern are drawn. The blue rays show a "one-hop" path between the
radar and Target 1, and a "half-hop" path between the radar and Clutter 1. Here, Target 1 represents an aircraft a few kilometers above the
Earth's surface, and Clutter 1 represents backscatter from irregularities in the ionosphere. Target 1 and Clutter 1 are at the same slant
range but appear to the radar to be at different elevation angles. If
we increase the target and clutter slant ranges by equal amounts, we
arrive at the case of the green rays. Here, the ray elevation angles
for Target 2 and Clutter 2 are less than for the corresponding blue
rays, but the rays remain resolvable in elevation angle. The elevation
separation of the target and clutter rays is a consistently exploitable
feature and nulling the half-hop clutter mode is the primary means
by which 3D OTHR can reduce auroral clutter levels.
It is still possible for an azimuth-only resolving array, which
cannot resolve target and clutter in elevation, to modestly improve
signal-to-clutter ratio by reducing the size of the resolution cell in the
azimuth extent. We refer to this azimuthal cell narrowing as a secondary means for achieving auroral clutter suppression. The ideas for 3D
OTHR are discussed in more detail in the next section on mitigations.

MITIGATIONS
Auroral backscatter clutter can be viewed as a spatial-temporal
coloured noise source. As discussed previously, however, the clutter is usually spread fairly broadly in the Doppler domain, due to
high-speed horizontal ionospheric drifts; thus, spatial processing
appears to be a more robust approach than temporal processing.
Space-time adaptive processing may play some role due to the coupling of azimuth and Doppler, as described in the previous section.
Initial experiments [11] show, however, that the coupling is significantly weaker than what one might expect and certainly much
weaker than in the classic scenario of ground clutter observed from
airborne radar. Thus, we set the scope of 3D OTHR to represent
an ensemble of techniques aimed at providing the best possible
spatial processing opportunities. Two types of techniques will be
examined here: joint azimuth-elevation processing [11], [12], and
joint transmit-receive processing [13], [14].

JOINT AZIMUTH-ELEVATION PROCESSING
Many current-generation OTHR systems have wide-aperture linear
receive arrays that can locate targets in range and azimuth. The target and clutter ray paths depicted in Figure 2 suggest, however, that
the elevation resolving capability of 3D OTHR could see application at high latitudes, because the target and clutter echoes could
be separated in elevation, even if colocated in range and azimuth.
Several approaches to elevation control have been realized.
Two approaches will be discussed in this subsection. The first approach is a vertical array. To achieve both azimuth and elevation
resolution, a vertical array normally resembles in appearance a
large wall. Vertical arrays have been used for both shortwave radio
broadcast and OTHR [15]. Heights between about 30 m and 150
m are typical, with lengths of up to about 1 km. The vertical arrays send a low-elevation beam in a direction nearly normal to the
plane of the array, which minimizes interelement electromagnetic
38

coupling and greatly simplifies the operation of the array. The disadvantage is that the arrays are significant civil engineering projects and quickly become expensive with increasing height. There
are also significant challenges to installing these sorts of arrays
at high-latitude locations, where there can be permafrost under
the ground and there can be logistical constraints to creating large
structures at remote locations.
A second approach to elevation control is the horizontal array,
typically lying on the surface of the Earth [16], [17]. This approach
has been used for OTHR [18], [19]. These arrays are structurally
much simpler than vertical arrays, but low-elevation beams result
in severe interelement coupling. Considerable care is required in
operating these arrays to ensure the coupling does not degrade performance, particularly on transmit, where excessive voltage standing wave ratio can overheat the transmitters. To illustrate low-elevation operation, let us say that the mission calls for a certain beam
pattern, which, in turn, requires a particular distribution of feed
point currents across the elements of the planar array to produce
that beam pattern. The array mutual impedance matrix takes the
feed point current requirement and converts it into a feed point
voltage requirement via the Z parameters N-port characterization v
= Zi, where v and i are vectors that represent the N feed point voltages and currents and Z is the mutual impedance matrix. Z can be
determined by calculation or measurement, where the calculation
can be done by using mutual impedance formulas in references
such as [20]. The ratio of the nth (1 ≤ n ≤ N) element feed point
voltage and current is the active feed point impedance, namely, Zn
= vn/in, and if the nth feed point is connected to a transmission line
with characteristic impedance of Z0, then the nth active feed point
reflection coefficient Γn can be computed. Then, it follows that the
forward wave voltage on the transmission line, at the feed point, is
vn+ = vn / (1 + Γ n ), because the boundary conditions require that the
feed point voltage is the sum of the forward and reflected waves.
The forward voltage v+ can then be extrapolated back to the radar
transmitter to specify the correct amplitude and phase at the transmitter. Thus, a given beampattern can be generated with the correct
excitation of the amplifier, provided the calculations can be carried
out, and there is control of each amplifier's amplitude and phase.
Calculations for Zn, Γn, vn, and vn+ were carried out for a 16 × 16
array of 7.5-m monopoles, with 10-m spacing and equal-amplitude
currents phased for end-fire operation (0° elevation) at 10 MHz, to
represent the particularly pathological case of high coupling. Results
are shown in Figure 3. In the figure, we see coloured grids, which
represent the 256 elements in the 16 × 16 array. The beam is directed
to the right in these diagrams. The top two panels show the variation
in impedance and reflection coefficients. The lower panels show the
feed point voltages and the forward wave voltages, normalized to
the means of each of those quantities. Clearly, the element voltages
"pile up" as the wave moves across the array. Increased voltage to
the right side of the array, under the constraint of constant current,
means that the elements on the right must put out more power. Increased voltage for constant current also means increased impedance, so it is not possible to match the impedance to a constant-impedance transmission line across the array, nor does it make sense to
use a variety of impedances for the transmission lines, because the
impedance requirements would change as the beam is steered.

IEEE A&E SYSTEMS MAGAZINE

DECEMBER 2017

Table of Contents for the Digital Edition of Aerospace and Electronic Systems Magazine December 2017