Three-dimensional sparse recovery space-time adaptive processing for airborne radar

Abstract

Conventional 3-dimensional (3D) space-time adaptive processing (STAP) has achieved good performance for non-stationary clutter suppression. However, the performance of 3D STAP rapidly degrades when applied to heterogeneous clutter environment due to the requirement of a large number of independent and identically distributed training samples to estimate the clutter covariance matrix. This study applies an efficient sparse recovery algorithm, i.e. multiple sparse Bayesian learning (MSBL), with tuning, to solve the limited sample problem of 3D STAP in airborne radar. Differing with the conventional 2D sparsity-based STAP, the proposed method utilises the sparsity of clutter in elevation-azimuth-Doppler domain and recovers the 3D clutter spectrum. Whereas in its’ large computational load, a fast algorithm extending the relevance vector machine is also considered.

Keywords

• radar clutter
• radar signal processing
• airborne radar
• covariance matrices
• space-time adaptive processing
• learning (artificial intelligence)
• Bayes methods
• radar detection
• dimensional sparse recovery space-time adaptive processing
• airborne radar
• 3-dimensional space-time adaptive processing
• nonstationary clutter suppression
• heterogeneous clutter environment
• independent distributed training samples
• identically distributed training samples
• clutter covariance matrix
• efficient sparse recovery algorithm
• multiple sparse Bayesian learning
• conventional 2D sparsity-based STAP
• 3D clutter spectrum