Performance assessment of HF-radar ship detection

Recently, significant efforts have gone into developing more reliable Integrated Maritime Surveillance (IMS) systems. The main purpose of these systems is to detect and track, and identify and classify cooperative and non-cooperative targets crossing maritime national boundaries. In such a context, each kind of sensor (i.e. shipborne or airborne, on-shore or off-shore) has its own tasks. Great attention has been paid to High-Frequency (HF) Surface-Wave (SW) radars as a long-range early-warning tool for applications in open waters. This paper presents the results from the NURC BP09 HF-radar experiment which took place in the Ligurian Sea (Mediterranean Sea), between May and December 2009. The effectiveness of the 3D constant false alarm rate (CFAR) detection algorithm, developed at the Institute of Technology of the University of Hamburg (TUHH), is investigated by means of ground-truth data sent from ships carrying Automatic Identification System (AIS) transponders. Under conservative assumptions, some preliminary results about the probabilities of detection and false alarm are presented and discussed. In the last part of the paper, foundations about future investigations are provided as well.