Moving Target Indication for Multi-channel Airborne Radar Systems

Moving Target Indication for Multi-channel Airborne Radar Systems

Lidicky, L.

Hoogeboom, P. (promotor)
Ligthart, L.P. (promotor)

Electrical Engineering, Mathematics and Computer Science

International Research Centre for Telecommunication and Radar

2010-10-11

Moving target indication (MTI) using radar is of great interest in civil and military applications. Its uses include airborne or space-borne surveillance of ground moving vehicles (cars, trains) or ships at sea, for instance. Airborne (space-borne) radar offers several advantages when compared to optical imaging, they include: day-andnight, all-weather capability and a possibly greater potential in resolving moving objects in a stationary scene. Recent systems utilise single-channel radars to produce fine-resolution images of stationary scenes by means of the synthetic aperture radar (SAR) and multi-channel arrays to achieve the best possible separation of moving objects from the stationary ones. This technology is currently employed in airborne military and experimental systems such as SOSTAR-X (European), PAMIR (Germany) and MCARM (USA). Also, the Canadian satellite RADARSAT-2 has a special multi-channel mode for radar MTI. The concept of multi-channel MTI alleviates the necessity for high-contrast in the detection of moving objects in the scene and it is very sensitive to low velocities. Its implementation is called space-time adaptive processing or STAP. STAP is capable of detection of moving targets but it offers limited options for their imaging. This thesis is a result of research jointly performed at the International Research Centre for Telecommunications and Radar associated with Delft University of Technology (IRCTR) and the TNO Defense, Security and Safety in the Hague, the Netherlands. The goal was to develop models suitable for general scenarios and to study optimal, yet fast and efficient algorithms capable of processing data created by these models. Theoretical results were to be verified experimentally. The research was to answer several questions: Are there any benefits in combining SAR and STAP? If the task at hand is to perform imaging of moving targets with resolution equal to the resolution of images of stationary targets using SAR, a signal model fully general for the given geometry is required. Inversions of such a model have been found, albeit for some cases only. The approach is mainly suitable for generation of synthetic datasets for numerical evaluation of SAR/MTI processing techniques, but also for the study of new, potentially interesting applications such as MTI with ultra-wideband ground penetrating radar. Is it possible to find a fast yet optimal or near-optimal MTI algorithm? It is suggested that if the channel calibration is done properly, averaging of signals from multiple channels can be used to extract moving targets. Being a relatively inexpensive operation, averaging will yield considerable savings in the computational power necessary to perform MTI. Although the real-time implementation of such an algorithm by means of a high-pass filter was already proposed by other authors, it was not validated using real data. How would such an algorithm perform when applied to measured data? A known STAP algorithm and averaging were applied to two airborne datasets: one based on the MCARM data, the other based on the SOSTAR-X data. It was concluded that a simple channel mismatch model could be assumed. Contrary to the MCARM data, the SOSTAR-X data contained severe channel mismatch. Nevertheless, at least one MTI algorithm worked satisfactorily in both cases. How can we use STAP in a new fashion? A way to detect victims buried in debris during natural disasters or terrorist attacks is proposed. Several experiments with ultra-wideband ground penetrating radar were carried out at the IRCTR. The results were in good agreement with predictions. The main results obtained in this thesis can be summarised as follows: Extended models for multi-channel SAR (MSAR) MTI. The thesis complements the TNO’s expertise in single-channel SAR MTI algorithms. It builds on ideas put forth in open literature and provides more general closed-form expressions for the MSAR transfer function. Heuristic investigations into error sources in real data. A simple model for airborne MTI where so-called internal clutter motion and flight path deviations are neglected, and the mismatch between channels is constant with time is tested on real data from MCARM and SOSTAR-X projects. Although deterministic the STAP techniques used in this thesis have already been applied to the MCARM data, no easy comparison to other published attempts was found. A detailed comparison is provided here. This heuristic validation provided insights into interpretation of SOSTAR-X data. Novel application of STAP to detection of buried victims. A new idea to use STAP for detection of moving objects buried underground or in debris is described. The idea is based on the possibility of detecting a slight motion of the victims produced by breathing. MSAR MTI processing provides fine spatial resolution due to the synthetic aperture concept and slow motion detection capability due to STAP. An alternative side-looking antenna configuration instead of the more traditional downlooking configuration is used. First results with measured data are presented.

http://resolver.tudelft.nl/uuid:46ac3a30-d110-4eb0-a9dd-698727043635

9789461130143

Institutional Repository

doctoral thesis

(c) 2010 Lidicky, L.