Constrained Passive Sensor Placement for Bearings-Only Tracking under Uncertainty
2026 11th International Conference on Recent Advances in Air and Space Technologies (RAST), İstanbul, Türkiye, 13 - 15 Mayıs 2026, ss.1-6, (Tam Metin Bildiri)
- Yayın Türü: Bildiri / Tam Metin Bildiri
- Doi Numarası: 10.1109/rast69551.2026.11672367
- Basıldığı Şehir: İstanbul
- Basıldığı Ülke: Türkiye
- Sayfa Sayıları: ss.1-6
- Orta Doğu Teknik Üniversitesi Adresli: Evet
Özet
We study constrained sensor placement for target localization and tracking with passive angular measurements. Sensors are selected from a discrete candidate set under minimum-separation and keep-out constraints. Placement quality is evaluated via an Extended Kalman Filter (EKF) covariance objective over a Monte Carlo ensemble of target trajectories, using Joseph-form updates for numerical stability. We propose a computationally efficient greedy selection scheme with geometric pre-screening, and benchmark it against random placement, uniform-circle placement, K-means, and a discrete genetic algorithm. Across M=50 scenarios at σθ=2° and N=4 sensors, greedy achieves the lowest mean uncertainty objective (e.g., 650.7 ± 27.8 vs. 689.6 ± 25.1 for K-means) while requiring substantially fewer objective evaluations. Ablations over the screening budget K′ show a clear quality–runtime trade-off, and a robust multi-regime design improves worst-case performance under higher bearing noise. Finally, extending the measurement model to range–bearing sensing yields additional uncertainty reduction, demonstrating the flexibility of the framework.