Long-term integration for radar detection of small targets in clutter

This report investigates long term noncoherent integration to detect radar targets with very small cross sections. Because of the long integration times, the target travels over several resolvable volumes; hence the integration must cover multiple resolvable volumes. Long term integration can provide at least 10 dB detection sensitivity increase over conventional cumulative detection for a shipboard radar against sea skimming cruise missiles. Physically, this improvement comes from the ability of long term integration to smooth over spatial variations of sea clutter, as well as over target fluctuations.; Since the actual target path differs from each of stored integration paths, collapsing loss from addition of interference results. The number and kinds of paths also affect false alarm rate. For our implementation of constant radial velocity paths within volumes illuminated by single beams, the method of path placement optimizing these geometric aspects is developed.; Sea clutter at low grazing angles is modeled with correlated temporal fluctuations whose local mean is also random and varies across resolvable areas. We next develop the first detection analysis that handles this entire clutter model. Detection performance including CFAR loss, is presented parametrically for integration within a single resolvable sea clutter area.; For long term integration across sea clutter areas, the transmission method affects smoothing over clutter spatial variations and target fluctuations, as well as the geometric collapsing losses. Example computations showed that long term integration performance is maximized by transmitting the shortest dwells possible, as long as the target is in different resolvable areas from dwell to dwell. This allows maximal smoothing over clutter spatial variations and target fluctuations, and can be markedly superior to coherent integration within a single resolvable area.; Using the above result, we examined long term integration performance for a hypothetical shipboard phased array radar performing horizon search of sea skimming cruise missiles. We found that the sensitivity increase was enough to detect {dollar}-30{dollar} dBsm targets with 90% probability at 10 km range in sea state 4. We then illustrated that, to detect targets with large acceleration at the same range, a more complex implementation including paths with radial acceleration is needed.