Abstract. A comparative experimental study of two algorithms for the automatic classification of surface targets using passive radar data is presented: two-stage probabilistic classification and multiple-hypothesis probabilistic classification. The study is motivated by the need to improve the reliability of passive target recognition in airborne, coastal, and distributed passive radar surveillance systems for maritime situational awareness, where decisions must be made under incomplete, ambiguous, and partially contradictory information. The objective is to compare these algorithms in terms of classification accuracy and relative computational complexity in typical radio-frequency scenarios of interest for passive surveillance systems. For a valid comparison, both algorithms were tested within a unified computational framework on a common synthetic database modeling platform composition, onboard emitter suites, and their operating ranges. All features were mapped to a normalized feature space, which made it possible to eliminate the effect of specific signal parameterization and to focus the analysis on the algorithmic differences between the methods. The experiments were conducted under a full-factorial variation of key scenario parameters, including emitter operating probability, observed emission coverage, the average number of emitters per platform, the average number of platforms per emitter, and operating range width. It is shown that multiple-hypothesis probabilistic classification makes more complete use of available information and is safe more robust to uncertainty, which in a number of scenarios yields an absolute accuracy gain of up to 57 percentage points compared with the two-stage algorithm. At the same time, this gain is achieved at the cost of a substantial increase in computational burden caused by the growth in hypothesis-tree depth. The two-stage algorithm, in contrast, while being less accurate in ambiguous situations, retains substantially higher computational efficiency. The practical significance of the results lies in supporting the choice of classification architecture for passive radar and related surveillance systems used for maritime situational awareness, as well as in confirming the feasibility of adaptive or hybrid use of the compared algorithms depending on the observation mode and available computational resources.

Keywords: surface target classification, radar emitter classification, passive radar systems, maritime situational awareness, two-stage probabilistic classification, multiple-hypothesis probabilistic classification, comparative algorithm analysis

For citation: Bocharov P. V., Libman M. S. Performance of surface target classification algorithms using passive radar data: a comparative experimental study of two-stage and multiple-hypothesis probabilistic classification methods. Aerospace Engineering and Technology. 2026. Vol. 4, no. 1, pp. 144–165. (In Russ.) https://doi.org/10.52467/2949-401X-2026-4-1-144-165; https://elibrary.ru/WGOAPY

References
1. Mironenko M. V., Pyatakovich V. A., Vasilenko A. M. Integrated system for control and monitoring of the surface and underwater situation in ensuring the safety of the marine areas of the state Strategicheskaya stabil’nost’ [Strategic Stability]. 2017. Vol. 81, no. 4, pp. 41–46. (In Russ.) https://elibrary.ru/ZTDBYH
2. Bocharov P. V., Libman M. S. An algorithm for classifying surface targets using passive radar data. Aerospace Engineering and Technology. 2024. Vol. 2, no. 4, pp. 118–129. (In Russ.) https://doi.org/10.52467/2949-401X-2024-2-4-118-129; https://elibrary.ru/BCDHLK
3. Liu C., Liu Y., Zhang Q. A two-stage classification algorithm for radar targets based on compressive detection. EURASIP Journal on Advances in Signal Processing. 2021. Vol. 2021, art. 23. https://doi.org/10.1186/s13634-021-00719-5
4. Greco M. S., Gini F., Farina A. A multiple hypotheses testing approach to radar detection and pre-classification // Proceedings of the 2004 IEEE Radar Conference (IEEE Cat. No.04CH37509), Philadelphia, PA, USA, 2004, pp. 463–468. https://doi.org/10.1109/NRC.2004.1316469
5. Bishop C. M. Pattern recognition and machine learning. New York: Springer, 2006. 738 p.
6. Hoi S. C. H., Sahoo D., Lu J., Zhao P. Online learning: A comprehensive survey. Neurocomputing. 2021. Vol. 459. P. 249–289. https://doi.org/10.1016/j.neucom.2021.04.112
7. Wang Z., Sun J., Li Q., Ding G. A new multiple hypothesis tracker integrated with detection processing // Sensors. 2019. Vol. 19, № 23, art. № 5278. https://doi.org/10.3390/s19235278
8. Kvasnov A. V. Methodology of classification and recognition of the radar emission sources based on Bayesian programming. IET Radar, Sonar & Navigation. 2020. Vol. 14, no. 8, pp. 1175–1182. https://doi.org/10.1049/iet-rsn.2019.0380
9. Revillon G., Mohammad-Djafari A., Enderli C. Radar emitters classification and clustering with a scale mixture of normal distributions. IET Radar, Sonar & Navigation. 2019. Vol. 13, no. 1, pp. 128–138. https://doi.org/10.1049/iet-rsn.2018.5202
10. Hall D.L., Llinas J. An introduction to multisensor data fusion. Proceedings of the IEEE. 1997. Vol. 85, no. 1, pp. 6–23. https://doi.org/10.1109/5.554205
11. Xu T., Yuan S., Liu Z., Guo F. Radar Emitter Recognition Based on Parameter Set Clustering and Classification // Remote Sensing. 2022. Vol. 14, no. 18. Art. 4468. https://doi.org/10.3390/rs14184468
12. Wang K. N. Illumination waveform design for non-Gaussian multi-hypothesis target classification in cognitive radar : PhD thesis. Monterey, CA : Naval Postgraduate School, 2012. 125 p. urn: handle:10945/7427
13. Blackman S. S. Multiple hypothesis tracking for multiple target tracking. IEEE Aerospace and Electronic Systems Magazine. 2004. Vol. 19, no. 1, pp. 5–18. https://doi.org/10.1109/MAES.2004.1263228
14. Oncu S., Karakaya M., Dalveren Y., Kara A., Derawi M. Real-Time Radar Classification Based on Software-Defined Radio Platforms: Enhancing Processing Speed and Accuracy with Graphics Processing Unit Acceleration. Sensors. 2024. Vol. 24, no. 23. Art. 7776. https://doi.org/10.3390/s24237776
15. Reid D. B. An Algorithm for Tracking Multiple Targets. IEEE Transactions on Automatic Control. 1979. Vol. 24, no. 6, pp. 843–854. https://doi.org/10.1109/TAC.1979.1102177