Pairing AIS and RF data improves maritime vessel detection accuracy

Pairing AIS with RF: Analysis of a Markovian Approach for Maritime Data Fusion

Machine LearningCryptography and Security

Summary

Tracking ships at sea is important for safety and spotting illegal activities, but it's hard because of the vast ocean area. The authors studied how to combine two kinds of data: AIS signals that ships send out but can be faked or missed, and radio frequency (RF) detections from space that don’t identify ships directly. They used a special statistical model to link these two data types and found their method gets close to simpler methods for matching ships but doesn't always beat them. Their approach helps understand errors and biases when combining these data sources.

What this means in practice

  • For maritime surveillance teams: Improve vessel monitoring by aligning AIS signals with independent RF detections to detect vessels that obscure their identity.
  • For satellite data analysts: Use a statistical model to fuse AIS data with RF signals from space for better tracking of ship movements and identifying suspicious behavior.

Tested on simulated data.

Authors

Daniel Grimm, Kyra Jongman, Renato Cordeiro Ferreira, Giuseppe Cascavilla, Alfredo Cuzzocrea, Willem-Jan Van Den Heuvel

Abstract

Monitoring maritime activities is important for safety and security, particularly for detecting illegal activities, but remains challenging due to the size and complexity of the maritime domain. The Automatic Identification System (AIS) is commonly used for this purpose, despite its known vulnerabilities. In contrast, space-based radio frequency (RF) observations provide an independent source of vessel activity but lack ship identification and therefore cannot identify vessels on their own. Following the CRISP-DM process, this study applies a Pair Hidden Markov Model (PHMM) to align individual RF detections with AIS trajectories, using temporal and spatial information while modeling noise and timing discrepancies. We evaluate it on NOAA AIS data with synthetic RF detections against nearest-neighbor baselines and analyze its errors with bias tests on trajectory length, sampling density, stationarity and AIS--RF distance. The length-corrected Forward score reaches a top-1 precision of $0.7776$ on the validation split of multi-candidate detections, where errors are possible, and an AUC of $0.637$ for detecting dark vessels among the detections that pass the prefilter. With a single RF detection, the score mostly reflects the AIS trajectory itself, and its errors favor longer and more stationary trajectories. A log-odds score, which compares the full model with a null model without match transitions, raises precision to $0.8580$ and AUC to $0.852$. Its length-corrected variant no longer shows a significant bias towards stationary trajectories and short inter-message distances, but still favors candidates farther from the RF detection. A time-weighted nearest-neighbor baseline without fitting remains ahead (precision $0.9013$, AUC $0.904$), so this single-detection, hand-designed PHMM formulation approaches but does not beat it.