FIFTH ANNUAL
ASAP '97 WORKSHOP

Processing Considerations
and Performance Results
for a Real-Time STAP
Flight Demonstration

John Samson and David Grimm
Honeywell Corporation

Arnold Bramson, Bruce Havlicsek, and Robert Warta
Northrop Grumman Corporation

Abstract Space-Time Adaptive Processing (STAP) presents a challenge to digital signal processors which are required to operate in real time. Available processing hardware, given limits on size, weight and power, must achieve a substantial percentage of its peak operating capacity to perform this task which involves careful attention to programming details. This paper discusses a real-time demonstration of STAP clutter suppression in the May 1996 Real Time Multichannel Airborne Radar Measurement (RT-MCARM) flight demonstration program. The system collected and processed IF sampled data from 16 phased array radar receiver channels, performing Doppler processing, adaptive clutter cancellation, pulse compression, and formation of multiple receive beams within an intentionally broadened transmit beam. Selective PRI stagger was implemented in Doppler regions of strong clutter to provide increased cancellation. An efficient method of adaptation with variable beam constraints was developed, which allows preservation of target phase across boundaries between range intervals with differing clutter statistics. In addition to improved angle estimation compared to unconstrained adaptive processing, the constraint makes it possible to perform pulse compression after beamforming thereby reducing the computational requirements. Rome Laboratory STAP algorithms were ported and mapped by Honeywell and Rome Laboratory to the Rugged Touchstone processor, a rugged COTS product derived from the Intel Paragon High Performance Computer. These algorithms were implemented using two different mapping approaches and compared with a mainbeam clutter suppression technique hosted on an onboard Mercury Computer.