Experiments with cognitive radar

Abstract

This paper presents experimental results from a cognitive radar system. In the cognitive approach used, radar parameters are constantly self-monitored and self-adjusted so that radar performance remains at a level specified by the user. This is in stark contrast to traditional radar where parameters are fixed and the performance varies according to the encountered scenario. The experiments were conducted with a specially designed system, the "Cognitive Radar Experimental Workspace" or "CREW". The CREW was designed so that it can incorporate cognitive processing concepts that require adaptive feedback, memory generation, and memory exploitation. These operations can be accomplished at pulse repetition frequencies (PRFs) of up to 30 kHz, enabling real-time time adaptation. The experiment results demonstrate that a cognitive approach can improve radar performance, evaluated through tracking error, over and above that achievable with a conventional fixed parameter radar system. Although the demonstration provided is based upon a tracking example, the lessons learned apply across a wide range of radar applications.

Experimentation of an adaptive and autonomous RF signalling strategy for detection

“…The experimental realization of schemes, such as that described in this work, requires the availability of highly capable, highly flexible hardware and software systems, such as the CREW [9]. The system consists of 4 channels enabling arbitrary waveform generation, clock synchronization, a baseband converter responsible for up and down conversion to/from the intermediate frequency centred on 5.5 GHz, and the RF frequency of 92.48 GHz with 4 channels of signal digitization, and the computer which controls the system.…”

Advanced cognitive networked radar surveillance

“…These offer opportunity both to investigate hardware design as well as providing a source of gathering networked data observations of targets and clutter, both in scarce supply. The CREW radar system [8] was developed as a networked radar capable of cognitive processing both in terms of hardware adaptation and processing adaptation. This is the first example of the specific design of a cognitive networked radar system.…”

Cost-Reference Particle Filter for Cognitive Radar Tracking Systems with Unknown Statistics

“…However, there is no further study on the optimization of the cost function, especially when the tracking structure can be adaptive. Cognitive radar designs the optimal estimator for Bayesian framework in [ 32 ], but the majority of the existing Bayesian tracking methods for cognitive radar applications are based on the Kalman filter for linear systems [ 33 , 34 ], and Kalman-like solutions, e.g., cubature KF (CKF) and continuous-discrete (CD)-CKF for nonlinear problems [ 35 ]. Few researches use PF.…”