TABLE OF CONTENTS Yuri I. Abramovich, Stuart J. Anderson, Alexei Y. Gorokhov and Nicholas K. Spencer
We consider the use of spatio-temporal adaptive array processing in over-the-horizon radar applications in order to remove non-stationary multipath interference ('hot clutter'). Since the spatio-temporal properties of hot clutter cannot be assumed constant over the coherent processing interval, conventional adaptive techniques fail to provide effective hot clutter mitigation without simultaneously degrading the properties of the backscattered sea/terrain radar signals ('cold clutter'). The approach presented here incorporates multiple stochastic (i.e. data-dependent) constraints to achieve effective hot clutter suppression, while maintaining distortionless output cold clutter post-processing stationarity.
We discuss the use of stochastically constrained spatial and spatio-temporal adaptive processing for hot clutter mitigation in scenarios that both do and do not allow access to a group of range cells that are free of cold clutter (supervised and unsupervised training, respectively). Theoretical and simulation results are complemented by surface-wave over-the-horizon data processing, collected during experimental trials in northern Australia.
The final section discusses convergence properties and convergence rate issues for stochastically constrained adaptive algorithms based on loaded sample matrix inversion routines.
This chapter is concerned with the adaptive processing of data from high frequency (HF) over-the-horizon radar (OTHR) [1, 2]. HF skywave radars exploit the ionised region of the upper atmosphere (the ionosphere) to reflect upwardly transmitted radiowaves back to the distant earth's surface.
