Aerospace and Electronic Systems Magazine December 2017 - 26

Design of a Network of Skywave Over-the-Horizon Radars

Figure 4.

Figure 5.

least the resultant number of hours. This clearly identifies the seasons and solar activity levels that do not satisfy a particular system
requirement of target coverage, or more generally speaking, are outside the system performance expectations. This performance metric
for the scenario is displayed in Table 1 (red indicates all the conditions that produce mostly-at-least three or fewer hours of coverage).
Clearly, BNE to SYD is a challenging flight to track, and the winter
and spring seasons (low solar activity in the Earth-based seasonal
sense) with low solar activity are difficult for all flight missions.
This table allows us to investigate these poor performing parameter
sets and understand what is causing the behaviour: in this case, target-radar geometry and low levels of ionisation in the ionosphere.
Although informative, the concise form of Table 1 provides no
insight into the specific time of day that coverage is attained. We
use an additional representation, shown in Figure 6, for the percentage of spatial track coverage as a function of hour of the day.
This allows the radar network designer to identify the key hours of
operation for the network.
For example, from this figure, we see that each of the candidate
flights in our mission have similar hours of operation in the range
of approximately 0 to 15 Universal Time (UT). We note that from
19 UT (sunrise at the ionospheric control point), there is a short
increase in performance followed by a quick decline. Inspection of

the propagation look-up tables shows that this behaviour is caused
by the optimal propagation frequency increasing outside of the design range of operating frequency of each radar in the network,
which causes suboptimal frequency selection. We do not see this
behaviour later in the day, as the well-developed ionosphere supports a much greater range of frequencies.

Map of eastern Australia. Shows flight routes colour coded with the
predicted hours of radar network coverage for inbound to SYD flights
in December with low solar activity for a radar network designed to the
base-system design. Red translates to no track coverage, while green is
10+ hours of coverage a day, and yellow patches are around 5 hours of
coverage.

Map of eastern Australia. Shows flight routes colour coded with the
predicted hours of radar network coverage for inbound to SYD flights
in December with low solar activity for a radar network designed to +6
dB of the base-system design. Red translates to no track coverage, while
green is 10+ hours of coverage a day, and yellow patches are around 5
hours of coverage.

RADAR SYSTEM SENSITIVITY BENEFIT AND STABILITY
ANALYSIS
The radar network design methodology discussed makes many
assumptions and simplifications (for example, the use of climatological models of the ionosphere [6]). We note that there are a
class of variations to performance that are difficult to use in a predictive model, such as shortwave fade-outs caused by solar flares.
However, there are the day-to-day variations in the noise levels
and the solar activity level driving the ionosphere. We consider it
important to include an assessment of the robustness of our radar
network design to these non-impulsive fluctuations in sensitivity.
For instance, will a small change in radar sensitivity mean overall
network performance is compromised, or is there some margin for
error?

IEEE A&E SYSTEMS MAGAZINE

DECEMBER 2017

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