IEEE - Aerospace and Electronic Systems - April 2022 - 28

Cross-Platform Radar Resource Management for Coordinated Search and Tracking
the load between the AC. Another option is to first assign
the most important tracking tasks to the most suitable AC
and then use the remaining radar resources for covering as
much as possible of the search volume.
IMPLICATIONS FOR MULTIPLATFORM MULTISENSOR
MANAGEMENT ARCHITECTURE
Figure 9.
Number of unassigned tracks for each sensor management
algorithm.
A major advantage of the macro/micromanager approach
is that the macromanager does not need any specific
knowledge about sensors carried by other platforms. It can
just ask them which tasks they can execute sufficiently
well and at which cost or it can estimate their answers
based on assumed performance models. This is a major
benefit when performing missions with a variety of platforms
with varying degrees of performance and willingness
to share details about their capabilities.
Via simulations it was discovered that a fully optimal
solution might not be significantly better than a solution
achieved by a combination of a naive partial solution followed
by an optimization step, see the discussion of the
simulated results. This results implies that even computationally
cheap algorithms can provide significant operational
benefits while having a low gap toward optimality.
Another important aspect is that the proposed architecFigure
10.
Computation time ofeach sensor management algorithm.
COMPUTATION TIME ANALYSIS
Figure 10 shows the computation time needed by each
sensor management algorithm (for micromanager and
macromanager) averaged over all Monte Carlo runs.
It can be observed that the computational effort
increases significantly as more " optimal " algorithms are
used, but this does not result necessarily to proportionally
better search and tracking performance. This can be
explained by the following.
The optimum region of the goal function might be
rather broad, and therefore, the optimal task-to-sensor
assignments might not be significantly better
than the ones in a broad region around them.
The load imposed by tracking tasks can vary significantly,
and therefore, it can be used for balancing
the uneven load imposed to the radars by a simple
Az./El. split of the search volume.
The aforementioned points indicate that a set of heuristics
can be found that can lead to a task assignment that
is very close to the optimal one. An example would be to
split the search volume in a simple and intuitive (to the
pilots) way and then use tracking tasks in order to balance
28
ture favors intelligent and capable sensors, such as AESA
radars with intelligent built-in radar managers.
The integration of new sensors is simplified as a unified
management scheme across all sensors is proposed,
which at the same time allows these sensors to optimize
their resources based on their own local sensor
micromanager.
CONCLUSIONS
This article has demonstrated the benefits of coordinated
radar resource management for multiple platforms using a
collaborative two-ship search and track scenario as reference.
The main benefits are the improved search and
tracking performance: enhanced tracking accuracy and
shorter time to detect targets.
In order to achieve these results, a two-level approach
was used. At the first level there is a micromanager
located close to or in the sensor. The micromanager is
tasked with finding the best sensor settings for executing a
task that might be assigned to it and also reporting what
resources each task needs. At the second level, there is a
macromanager who decides which tasks will actually be
executed and by which sensor/platform.
Remaining open challenges are as follows.
Would it be better to evaluate the optimal coordination
in a decentralized manner? If yes, how and
what are the communication needs involved?
IEEE A&E SYSTEMS MAGAZINE
APRIL 2022
IEEE - Aerospace and Electronic Systems - April 2022

Table of Contents for the Digital Edition of IEEE - Aerospace and Electronic Systems - April 2022
