Distributed Active Sonar Detection in Dependent $K$ -Distributed Clutter

摘要: Distributed active sonar systems exploit the simultaneous detection of targets on multiple receiver platforms to improve system performance. False alarm performance modeling for such typically assumes independence from sensor sensor; however, false alarms clutter (e.g., shipwrecks, mud volcanoes, or rock outcroppings) are expected produce statistically dependent data. In this paper, a model exhibiting intersensor dependence and having K-distributed marginal probability density functions (pdfs) is proposed analyzed. Under constraint equal sensor-level performance, m-of- n fusion processor seen be optimal most cases interest all but heaviest tailed clutter. The system-level derived approximations developed common target models. analyzing (m = n) perform best in heavy when has some measure consistency while m/n ap 0.25 highly variable Gaussian model. AND, OR 1) , median processors degrade as pdf tails increase, although effect confounded between tail heaviness dependence. importance accounting dependence, illustrated by an up 10-dB overestimation data incorrectly assumed independent across sensors.