IEEE Aerospace and Electronic Systems Magazine - June 2020 - 8

Feature Article:

DOI. No. 10.1109/MAES.2019.2957847

The Development From Adaptive to Cognitive Radar
Resource Management
Alexander Charlish, Folker Hoffmann, Christoph Degen, Isabel Schlangen,
Sensor Data and Information Fusion Department of Fraunhofer FKIE

INTRODUCTION
Cognitive radar can be defined as [1]:
A radar system that acquires information, knowledge, and understanding about its operating environment
through online estimation, reasoning and learning, or
from datasets comprising context. A cognitive radar then
exploits the acquired information, knowledge, and understanding to enhance information extraction, data processing, and radar management.
Implicit in this definition is the implementation of perception-action cycles that are predominantly recognized as a
defining feature of cognitive radar. As is evident from the
variety of living species, the extent to which information,
knowledge, and understanding is acquired and exploited can
vary greatly. Consequently, a single feature qualifying a radar
as cognitive does not exist. Instead cognition can be visualized as a continuous spectrum with varying degrees of cognitive capability [2]. A cognitive radar ultimately finds that the
information present in a single radar dwell is insufficient to
effectively control the available degrees of freedom. Therefore, a cognitive radar attempts to preserve information,
knowledge, and understanding exploited from any available
source, be it processed radar data acquired over extended
time periods, context, or learning.
Multifunction radar (MFR) is a particularly attractive
recipient for cognitive radar techniques. An MFR, ideally
enabled by arbitrary waveform generation and electronic
beam steering, is capable of executing many tasks that support differing radar functions. Radar resource management is
a vital enabling technology for an MFR, as it enables an

Authors' current address: A. Charlish, F. Hoffmann,
C. Degen, and I. Schlangen, Sensor Data and Information Fusion Department of Fraunhofer FKIE, 53343,
Wachtberg, Germany (e-mail: alexander.charlish@
fkie.fraunhofer.de).
Manuscript received March 30, 2019; accepted June 12,
2019, and ready for publication December 3, 2019.
Review handled by D. Crouse.
0885-8985/19/$26.00 ß 2019 IEEE
8

effective control of the transmitter degrees of freedom, thus
unlocking the full potential of the system. Management
becomes even more crucial for the current trend of multifunction RF systems, as the additional functions place additional
demands on the shared resources. Since radar resource management techniques have implemented perception-action
cycles for many decades, it is a topic that is very closely
related to cognitive radar.
In the "Cognitive Radar Architecture" section, a cognitive radar architecture is presented, within which cognitive
radar techniques can be realized. Then, in the "Development
of Adaptivity in Radar Resource Management" and "Drive
to Cognitive Radar Resource Management" sections, an
overview of adaptive and cognitive radar resource management techniques is presented. As this series of techniques
represents increasing adaptation to information or knowledge, the reader is taken on a walk along the cognitive spectrum in the context of radar resource management. The
"Challenges and Drawbacks" section discusses drawbacks
and challenges that should be considered in the cognitive
radar research and development process.

COGNITIVE RADAR ARCHITECTURE
Before describing the adaptive to cognitive radar resource
management techniques, it is valuable to consider an architecture within which the techniques can be realized. Various
different architectures have been proposed in the literature,
which generally comprise components for perception, action,
and memory that are split over multiple hierarchical levels. A
suitable cognitive radar architecture is described in [1] and
illustrated in Figure 1, which is an adapted version of the cognitive radar architecture by Smits et al. [3].
In this architecture, the hierarchical levels are based
on information abstraction levels. Abstraction levels for
sensor data and information processing have been widely
studied, most notably by the Joint Directors of Laboratories Data Fusion Model [4] and its revised versions [5].
Based on these studies, information abstraction levels relevant for a cognitive radar system can be identified as the

IEEE A&E SYSTEMS MAGAZINE

JUNE 2020

IEEE Aerospace and Electronic Systems Magazine - June 2020

Table of Contents for the Digital Edition of IEEE Aerospace and Electronic Systems Magazine - June 2020