Performance analysis of a simulated distributed multi-radar system for target detection

The size and number of vessels in the maritime domain has grown significantly over the past decades. Hence, it has become more and more important to obtain reliable information about the current traffic situation. There cannot be more than one traffic situation at a certain time in a specific area. This unique traffic situation picture is seen differently and never in total by each individual vessel or monitoring station. Therefore, it is a reasonable idea to share and merge their corresponding sensor data in a cooperative traffic situation assessment scheme. The radar is considered to be the primary sensor for traffic situation awareness and collision avoidance in the maritime domain. In fact, nearly every vessel is equipped with at least one radar system composing a distributed sensor network, which is spatially sampling the unique situation picture. In this work, we characterise the performance of such a multi-radar system in terms of number of sensors and aspect angle diversity. The results of simulations of different radar configurations are analysed and compared to the performance of a single radar. First, a static case with an ideal (semi-circular) configuration for radars is investigated. It is shown that a small number of radars with the assumed characteristics in a multi-radar system can significantly improve the probability of detection and the detection accuracy compared to a single radar with the chosen fusion strategy. A high probability of detection can be achieved with at least 4 radars. Then the simulation is extended to a realistic scenario with a mobile target and realistic (static) radar locations. It is shown that the fusion of multiple radars yields significant improvement of the detection probability of more than 60% in the best case and a target can be detected about 1NM earlier. An increase of the accuracy by a factor of 3 on average is achievable as well.