Stochastic modelling of target detection and tracking in surface wave HF radars


Sevgi L.

INTERNATIONAL JOURNAL OF NUMERICAL MODELLING-ELECTRONIC NETWORKS DEVICES AND FIELDS, cilt.11, sa.3, ss.167-181, 1998 (SCI-Expanded) identifier

Özet

In this study, a software package is described which has been developed for the simulation of detection and tracking in surface wave high frequency (SW-HF) radars. The aim is to investigate the problems related to detection and tracking of surface targets at beyond the horizon ranges. In SW-HF radars target detection and tracking involves stochastic features such as target RCS fluctuations, atmospheric, galactic and/or man-made ambient interference components, and the sea clutter with the dominant resonant Bragg returns which affect target detection are all incorporated in the package. In the model, first, terrain data are fed into the simulator by means of a specially designed graphical user interface. Then, I scenario is prepared where the radar's location, coverage and operational parameters can be defined, together with different targets and their sailing routes. The radial propagation paths for angular resolution cells are extracted from the terrain data with the lengths of (possible) sea-island transitions. The surface wave path losses are calculated over the smooth spherical earth's surface with finite conductivity. Surface roughness and mixed-path propagation effects are also included in these calculations. The target detection is performed in the frequency domain after calculation of the noise floor, signal to noise (SNR) and clutter to noise (CNR) ratios. Since SW-HF radars yield coarse range and azimuth but accurate velocity measurements, different Kalman filter techniques are applied for the target tracking, and algorithms are added to improve the track-measurement data correlation. Various simulation tests are performed and the results are presented. (C) 1998 John Wiley & Sons, Ltd.