Zobacz książki o podobnej tematyce

Powiadomienie o dostępności

Więcej informacji o Multistatic Multihypothesis Tracking Techniques for Underwater- and Air Surveillance Applications

• Dane szczegółowe
• Opis
• Ulubione w innej kategorii

Opis

We developed and discussed diverse concepts of multistatic target tracking based on the multihypothesis tracking approach. When suiting a tracking technique to the specific application, it is necessary to specify models to describe the physical phenomena. Thus, the tracking systems must reflect the available knowledge about the sonar or radar system. The analysis of simulated and real data sets showed that the advantages of multistatic sonar can only be exploited if the measurement model is correct. That is, we need to interpret correctly the measurement information and the events of false alarms and missed detections. However, the precision of modeling knowledge for underwater sound is limited by the stochastic fluctuations/uncertainties. Therefore, in order to provide a robust algorithm it is important not to overestimate the accuracy of the modeling. In the passive radar application a precise measurement modeling is not only useful but essential in order to distinguish between ghosts and targets. We discussed two Deghosting approaches, which are both based on a hard constraint logic: The LR ap- proach displays tracks that are confirmed by at least d illuminators, the IP approach finds the best multi-target interpretation in which every R/D track is used exactly once. Taking into account hard constraints is necessary to handle the ghosting prob- lem. However, this causes sensitivity of the tracking approach, which needs to be compensated by the robustness of the other components of the tracking approach. To develop a robust and efficient tracking approach we have to consider different aspects that have decisive influence on the successful operation of the complete sys- tem. The non-linear Kalman Filter update and the Gaussian sum initialization from measurements of two illuminators were therefore discussed separately and analyzed by detailed numerical simulations. When developing a tracking strategy we need to find a compromise between high per- formance and robust target tracking. As discussed in Sec. 5, starting from the aspect of multi-sensor fusion a high precision sensor model is necessary to correctly associate measurements from different sensors. The tracking task in the passive radar appli- cation is even more sensitive, since the sensor model is exploited for distinguishing between ghosts and targets. In this thesis we investigated several tracking techniques for application to the track- ing task in passive radar exploiting DAB/DVB-T transmission. The thesis provides analysis and discussion of their performance in various scenarios; however, we can not deliver a final statement, if DAB and DVB-T transmissions can deliver a sufficient situation picture with appropriate accuracies. In future the algorithms have to be tested thoroughly with real radar data to provide better indication of the expected performance capabilities. Other important aspects may be the analysis of benefits from combination of DAB/DVB-T systems with other passive radar systems like on the basis of FM signals or even active systems. The deghosting performance strongly correlates with the number of detections avail- able from different illuminator and receiver combinations. What was not discussed here is the exploitation of additional receivers and finally as stated in Sec. 3.4 the exploitation of a more accurate azimuth measurement or an additional measurement of a phase difference, which may have a significant impact on the performance of the total system.

Szczegóły książki

• Pełny tytuł: Multistatic Multihypothesis Tracking Techniques for Underwater- and Air Surveillance Applications
• Autor: Martina Brötje
• Język: Angielski
• Oprawa: Twarda
• EAN: 9783898632553
• ISBN: 3898632555
• ID: 12770196
• Wydawca: Gca Ges.F.Computeranw.
• Waga: 327 g
• Wymiary: 208 × 146 × 20 mm
• Data wydania: 02. May 2013

---

---

---