Performance comparison of fast-scan GMTI/STAP architectures

This paper presents a study on multibeam radar architectures to provide wide-area fast-scan ground moving target indicator (GMTI) coverage. The minimum dwell time of a GMTI radar system is often determined by the Doppler processor. The revisit rate can be lowered, while still meeting dwell time constraints, by using multibeam techniques to cover more area in each dwell. The two multibeam techniques examined here are: (1) to spoil the transmit beam with a quadratic phase and digitally beamform on receive; and (2) to break each transmitted pulse into sub-pulses each with different center frequencies, pointed in different directions, and to beamform in the corresponding direction on receive. Space time adaptive processing (STAP) is used with these GMTI architectures to mitigate main beam clutter and to enhance detection of targets in severe interference environments. The antenna sub-arraying, used for beamforming and STAP, generates grating lobes in the receive pattern in multibeam systems. This can degrade performance because grating lobes cannot be suppressed in more than one direction at a time. Two different antenna designs are studied to determine the effect of grating lobes for each of the above fast-scan architectures; non-overlapping (contiguous) sub-arrays and 75 percent overlapping sub-arrays which allows greater control of grating lobes at the cost of greater array complexity. This paper presents signal-to-interference ratio loss (SIRL), probability of detection (P _d) and minimum detectable velocity (mdv) for each system. A narrowband interference source is swept in angle to determine system performance degradation due to the grating lobes