Finding a Needle in a Haystack: Satellite Detection of Moving Objects in Marine Environments

There is a growing need in maritime missions to monitor moving vessels with satellite sensors, in order to detect vessels that may mislead about their identity and transmit wrong identification parameters. In order to provide an efficient and cost-effective solution, vessel behavior prediction is a necessary ability. We present three models for vessel behavior prediction: Min-Max, Uniform-Walk and Normal-Walk. We use real marine traffic data (AIS, Automatic Identification System) to compare the performance of these models and their ability to predict vessel behavior in a time frame of 1--11 hours.