Introduction au STAP: 2. Modèle des Signaux et Principe du Filtrage

Page:

35-56

DOI:

https://doi.org/10.3166/TS.28.35-56

OPEN ACCESS

Abstract:

This paper presents the data model for signals received on the antenna array of an airborne radar system. This model is acknowledged and deeply used by the STAP community. To introduce and comprehend basics about STAP signals a simple geometry will be considered for the radar scenario. Finally, principles of STAP filtering will be presented.

**Extended Abstract**

Most radar systems aim at informing the operator about the presence or absence of targets of interest in the observed scene. To do so, electromagnetic pulses are emitted and signal returns are analyzed. Typically, the analysis is performed range cell by range cell. A range cell stands for a portion of the space at a certain distance from the radar and whose dimension is set by the waveform resolution. For each cell, the signal processing chain has to extract correctly the signal of interest (i.e., the target) embedded in noise. The noise is mainly due to thermal noise, possibly jammers, and echoes which differ from the target. These undesired returns are referred to as clutter and are essentially due to the ground.

For airborne radar systems, ground clutter is distributed in angle and in Doppler due to the motion of the platform. More specifically, clutter is situated in the angleDoppler domain in a locus that is mainly determined by the geometry of the scenario. Flight parameters, configuration of the radar antenna, pointing direction and ground elevation are hence parameters that mainly influence the clutter location. For instance, for a perfect sidelooking antenna, the clutter is located on a line in the angle-Doppler domain.

To extract the echo of a slow moving target from ground clutter, it thus necessary to discriminate both signals with respect to their direction of arrival (DOA) and their velocity. In other words, a space-time processing is required. Although such filtering technique was not feasible a few years ago due to technology limitations, it has become nowadays feasible thanks to the introduction of active electronically scanned arrays (Lacomme et al., 2001). Within this context, many algorithms have thus been designed for detecting slow moving targets competing with clutter; these techniques are referred to as space-time adaptive processing (STAP).

To comprehend the essence of STAP, it is necessary to study first the characteristics of the signal received on the antenna array. The first part of the paper presents therefore the conventional signal model that is acknowledged and deeply used by the STAP community. The three main components of the signal, i.e., the target signal, the clutter and the thermal noise, are described.

It is shown that the target echo can be easily described by a two-dimensional cisoid whose frequencies are directly related to the target’s DOA and velocity respectively.

Thermal noise (receiver noise mainly) and clutter are represented by wide sense stationary random processes. Thermal noise is considered to be white spatially and temporally. On the contrary, the clutter component is correlated temporally from pulse-to-pulse and spatially between the array elements. More precisely, for each range cell, the clutter echo is seen as the superposition of a large number of independent clutter sources having zero inherent velocity. Thereby, for each clutter patch, the spatial and Doppler frequencies are linked by a specific relation that determines the clutter locus in the angle-Doppler domain. Canonical radar configurations, i.e., sidelooking and forward looking configurations, are considered to illustrate the impact of the geometry on the clutter locus.

The second part of the paper presents the principle of STAP filtering. Once the received data are sampled in space and time, the presence of targets is usually tested

– for each range cell;

– in the angular sector covered by the mainlobe antenna on transmit;

– for different velocities.

More specifically, a linear STAP filter is applied to the space-time data vector of the range gate of interest. The filter aims at suppressing clutter components while maintaining a gain on the target under test. This operation is tantamount to whitening the received data via the noise covariance matrix (thermal noise plus clutter) and, then, integrating coherently the possible target signal. As the noise covariance matrix is not known a priori, it is usually estimated adaptively in flight with secondary data collected from several range cells near the range gate of interest.

STAP filters are often applied in suboptimal structures where data are first projected into a lower dimension space prior to adaptation. These architectures allow one to decrease both the computational burden and the number of secondary data required to estimate correctly the noise. Two suboptimal techniques and conventional STAP performance metrics are presented in the last sections of the paper

**RÉSUMÉ**

Cet article présente le modèle classique des signaux reçus sur l’antenne réseau d’un système radar aéroporté à bande étroite. Ce modèle est communément reconnu et adopté par la communauté STAP (Space Time Adaptive Processing). Quelques approximations seront faites sur le scénario radar envisagé afin de se concentrer sur les aspects essentiels du filtrage spatio-temporel adaptatif. Le principe du filtrage sera ensuite présenté.

Keywords:

*filtering, space-time adaptive processing, adaptive detection, estimation*

**MOTS-CLÉS**

*filtrage, traitement spatio-temporel adaptatif, détection et estimation radar.*

1. Introduction

2. Modélisation des Signaux

3. Principe du Filtrage STAP

4. Conclusion

References

Barlow E. J. (1949). « Doppler Radar ». Proc. IRE, vol. 37, p. 340-355, April.

Beau S., Marcos S. (2011). « Taylor series expansions for airborne radar space-time adaptive proces-sing ». IET Radar, Sonar and Navigation, vol. 5, n° 3, p. 266-278, March.

Bidon S., Besson O., Tourneret J.-Y. (2011). « Synthèse des traitement STAP pour la détection en environnement hétérogène ». Revue Traitement du Signal.

Billingsley J. B. (1996). Exponential decay in windblown radar ground clutter Doppler spectra : Multifrequency measurements and model. Technical Report n° 997, Lincoln Laboratory, Massachusetts Institute of Technology, July.

Bose S., Steinhardt A. O. (1995). « A Maximal Invariant Framework for Adaptive Detection with Structured and Unstructured Covariance Matrices ». IEEE Trans. Signal Processing, vol. 43, n° 9, p. 2164-2175, September.

Brennan L. E., Reed I. S. (1973). « Theory of adaptive radar ». IEEE Trans. Aerosp. Electron. Syst., vol. 9, p. 237-252, March.

Brennan L. E., Staudaher F. M. (1992). Subclutter visibility demonstration. Technical Report n° RL-TR-92-21, Adaptive Sensors Incorporated, March.

DiPietro R. C. (1992). « Extended Factored Space-Time Processing for Airborne Radar Systems ». Proc. Asilomar Conference on Signals, Systems, and Computers, vol. 1, p. 425-430, October.

Fishbein W., Graveline S. W., Rittenbach O. E. (1978). Clutter Attenuation Analysis. Technical report, Army Electronics Command Fort Monmouth NJ, March.

Kay S. M. (1993). Fundamentals of Statistical Signal Processing : Estimation Theory. Prentice Hall, Englewood Cliffs, NJ.

Kay S. M. (1998). Fundamentals of Statistical Signal Processing : Detection Theory. Prentice Hall, Englewood Cliffs, NJ.

Klemm R. (2002). Principles of Space-Time Adaptive Processing. IEE Press, London, U.K.

Lacomme P., Marchais J.-C., Hardange J.-P., Normant E. (2001). Air and Spaceborne Radar Systems: An Introduction, William Andrew. Le Chevalier F., Principles of Radar and Sonar Signal Processing. Artech House, Norwood,

Melvin W. L. (2004). « A STAP overiew ». IEEE Trans. Aerosp. Electron. Syst., January.

Oudin M., Delmas J. P., Barbaresco F., Lupinski L. (2009). « Siegel distance-based covariance matrix selection for space-time adaptive processing ». Proc. IEEE Radar conf., Bordeaux, October 12–16.

Reed I. S., Mallett J. D., Brennan L. E. (1974). « Rapid Convergence Rate in Adaptive Arrays ». IEEE Trans. Aerosp. Electron. Syst., vol. 10, n° 6, p. 853-863, November.

Ries P., Neyt X., Lapierre F., Verly J. (2008). « Fundamentals of spatial and Doppler frequencies in radar STAP ». IEEE Trans. Aerosp. Electron. Syst., vol. 44, n° 3, p. 1118-1134, July.

Savy L., Degurse J.-F., (2011). « Stop-Band APES : Traitement STAP sur données fortement hétéro-gènes ». Revue Traitement du Signal.

Siegel C. L. (1964). Sympletic Geometry. Academic Press, New-York.

Skolnik M. I. (1970). Radar Handbook. McGraw-Hill.

Ward J. (1994). Space-Time Adaptive Processing for Airborne Radar. Technical Report n° 1015, Lincoln Laboratory, MIT, Lexington, MA, December.