IEEE - Aerospace and Electronic Systems - December 2019 - 26

Cognitive Radar Principles for Defence and Security Applications
Classiﬁcation:. Regions of interest found by the
"segmentation" will either be shown to the pilot or operator for his own assessment and/or fed to an appropriate
classifier system. Today's state-of-the-art classifiers [17]
are able to distinguish between a dozen target types, presuming they are well known from the training process.
Those classifiers may fail completely in cases where
the target type is not known from the training process;
the target type is changed to a certain extent compared
with the known target from the training process;
the target type is only partly visible due to occlusion
by other objects nearby.

ADAPTATION TO A CHANGING ENVIRONMENT
A radar has to operate in a complex environment composed of the wanted and unwanted objects, clutter, terrain,
propagation conditions, and the electromagnetic environment. Depending on the radar platform (stationary, moving, naval, land based, airborne), the environment is more
or less dynamically varying. Cognitive methods open a
wide field of sensing and learning about the environment
and adapting and optimizing radar operation. This section
gives an overview of the environmental adaptation, in particular by supplementing the sensed environment with
data from the external sources.
The challenges for radar operation are the following:
adapt the scan strategy and search volume to the
terrain;
adapt detection thresholds to the clutter background,
suppress detections and tracks of clutter, and unwanted
targets;
optimize the detection and track performance of the
wanted targets;
reduce propagation effects or utilize propagation
effects for improving the performance.
In addition to the sensed environment, external data are
available for better adapting the radar to the environment.
Examples of the external data are:
OpenStreetMap (including OpenSeaMap) with
coastlines, roads, wind turbines, buoys, etc.;
shuttle radar topography mission (SRTM) data with
the digital elevation models of the earth;
weather data providing actual and local information
about wind speed and direction, sea state or
precipitation.
In the following, some examples of the smart techniques
for adapting the radar to the environment are discussed.
26

TERRAIN ADAPTATION
Depending on the geographical own sensor position, the
beam patterns and search volume can be adjusted to the
local terrain by the use of a digital terrain map. The terrain
information can be augmented by the actual ground clutter
measurements, which allow for keeping the clutter level
inside the dynamic range of the radar receivers, respectively, inside the clutter improvement limit of the radar.

ADAPTATION TO CLUTTER ENVIRONMENT
In order to adjust the detection thresholds to the actual
background, the clutter level has to be sensed and estimated. This estimation can be performed on a single look
basis or overtime by a learning process of the clutter characteristics. The clutter features can be characterized by its
spatial composition, amplitude statistics, and Doppler
spectrum. This allows for a reliable adjustment of the
detection thresholds to the clutter characteristics while
maintaining the false alarm rate low and ensuring the target detection performance over the clutter background.
Adjusting signal processing to the clutter estimation,
e.g., by adjusting the Doppler filter characteristics to the
estimated clutter spectrum, clutter suppression and, thus,
target detection can be optimized.
By adjusting waveform to the estimated clutter, the
target detection performance can be improved further. An
example of this application could be the selection of the
clutter fill range based on the estimated clutter range or
adjusting the pulse repetition interval for the dedicated target illuminations with respect to the a priori known target
range, and radial velocity and clutter background.

SENSING AND ADAPTATION TO PROPAGATION CONDITIONS
Multipath propagation on a sea surface is typical in naval
radar applications. Superposition of the direct and
reflected path at the sea surface leads to zones with fading
and loss of target detection as well as the errors in elevation measurement. By sensing fading situations, tracking
can be made more robust against the detection misses and
elevation errors by appropriate changing the transmit frequency and, thus, avoiding a fading situation.
Depending on the weather and atmospheric situation,
ducting can occur, which may result in wave propagation
over the sea surface up to distances well beyond the typical radar horizon. Ducting conditions cause multiple times
around clutter and target echoes, resulting in ghost targets,
false alarms, or reduction of the target detection performance. Ducting conditions are rather common and
expected for naval radars. Upon sensing ducting conditions, the degrading effects can be prevented by adapting
the processing to these situations. Moreover, the radar can
benefit from ducting by switching to waveforms, which

IEEE A&E SYSTEMS MAGAZINE

DECEMBER 2019

IEEE - Aerospace and Electronic Systems - December 2019

Table of Contents for the Digital Edition of IEEE - Aerospace and Electronic Systems - December 2019