Knowledge-based AIS Shadow Zone Identification in VTS Area


  • Sang-Lok Yoo
  • Keon-Myung Lee


The current lack of reliability for AIS messages is a key disadvantage for their trustworthiness as a navigational device. AIS messages are frequently lost in the VTS area. This paper is concerned with identifying the AIS shadow zone in the VTS area. We proposed a knowledge-based algorithm to extract sequential messages from raw data. After extracting data for more than 2 knots from raw data for an underway vessel, loss rate is calculated from the data within 10 minutes of the reporting intervals. The first row in each group is discarded to use sequential messages. The loss message is visualized by density contour plots with two-dimensional histograms. Two and three shadow zones were identified for AIS Class-A type and B-type, respectively, in Wando VTS. It is possible to check which zone is more frequently lost visually. Furthermore, the analysis shows that the loss rate of the Class-B type was 33 times that of the Class-A type. It was found that the loss rate of AIS is mainly caused by fishing vessels with lower antenna height than cargo ships and passenger ships. In particular, it was found that the loss rate of fishing vessels equipped with Class-B type was about 10 times that of those with Class-A type. We also found a weak correlation between number of message and loss message. Therefore, the VTS operator should recognize the characteristics of AIS message loss rate by ship type through these findings, and rate trustworthiness of AIS predicted messages accordingly. If there are any obstructions such as islands or mountains in the VTS area, the proposed method allows for easy identification of relative AIS shadow zone.