Clutter Simulation And Three-dimensional STAP Techniques For Airship-borne Radar

Stratospheric airships have many advantages, such as either stationary or maneuver deployment, long duration, huge loading capacity and recyclable use, and thus is widely used in ground survey, monitoring, missile early warning and defense. Research and development of stratospheric airships have become a most active field in many countries. In order to promote the application of airship-borne radar, this paper deals with the modeling and simulation of three-dimensional(3D) terrain clutter and three-dimensional Space-Time Adaptive Processing(3D-STAP) techniques. The major work in this paper is outlined as follows:1. The signal model for airship-bome three-dimensional array radar is built. Airship-borne 3D antenna array model and radar coordinate system are given. Based on the uniformly spaced linear array signal model proposed in the Ward’s report, the transmitted and received signals, noise, jamming and clutter components of 3D antenna array are analyzed in detail. Explicit expressions for their space-time snapshots and covariance matrices are derived in terms of three-dimensional coordinates of each element and each clutter patch. With elevation, azimuth and Doppler information incorporated, these expressions are suitable for arbitrary 3D antenna arrays.2. Three-dimensional terrain clutter is simulated at a background of an airship-borne three-dimensional array radar. Typical statistical models of radar clutter are introduced. Emphasis is put on three essential aspects in 3D terrain clutter simulation:clutter patch dividing, clutter signal simulation and terrain shielding evaluation. A whole flow chart of clutter returns simulation is presented. Computer simulation of clutter returns is made by using an airship-borne half-C60 molecule array and simulated 3D terrain. The proposed clutter model is validated by the numerical simulations.3. The elevation-azimuth-Doppler three-dimensional Space-Time Adaptive Processing (3D-STAP) techniques are studied. First, the principle of STAP and STAP performance metrics are introduced. Subsequently, based on traditional two-dimensional STAP, four 3D-STAP techniques, included Matched Filter(MF), Adaptive Matched Filter(AMF),Extended Factored Approach(EFA or mDT) and Joint Domain Localized(JDL) approach, are derived with elevation information incorporated. Finally, the proposed 3D-STAP techniques are applied to the 3D terrain simulation clutter data. The results indicate that by choosing proper parameters, both 3D-EFA and 3D-JDL can obtain near optimum performance with less training samples.