Radar clutter mitigation via space-time wavefront adaptive sensing

Space-time adaptive processing (STAP) in complex radar propagation and inhomogeneous clutter environments is often precluded because neither the target wavefront is sufficiently known nor is signal-free training data available. In recent work, wavefront adaptive sensing (WAS) was proposed to overcome these challenges by combining minimum variance (MV) adaptive processing and blind source separation (BSS) for distributed sources. In this paper, WAS is compared with conventional BSS and MV processing both analytically and via simulation. In an over- the-horizon radar (OTHR) spread-Doppler clutter environment, WAS is shown to avoid MV signal cancellation at high SNR and poor BSS threshold performance at low SNR.

1. INTRODUCTION

Spatially inhomogeneous and Doppler-spread clutter compromises target detection performance of radar systems. Classical space-time adaptive processing (STAP) methods achieve clutter suppression by estimating a clutter-plus-noise covariance matrix from “signal-free” training data. Signal-free snapshots are typically obtained by using data from range bins which are identically distributed but well separated from the hypothesized target range bin “under test”. In highly inhomogeneous environments, such as HF over-the-horizon radar (OTHR), the clutter statistics change significantly across neighboring range bins [1], [2]. An alternative approach is to estimate the signal-plus-clutter-plus-noise covariance matrix at the range bin under test and employ minimum variance distortionless response (MVDR) adaptive beamforming [3], [4].