The technology of multisensor distributed data fusion has been gaining importance in recent years. Up to now, this problem is always studied with assumption of conditional independence of observation data. In this paper, we consider Neyman-Pearson criterion with general correlated sensor observations.We first consider multisensor distributed Neyman-Pearson decision with correlated sensor observation data and suggest an efficient algorithm to search for optimum local compression rules for any fixed fusion rule. The results are supported by computer simulations.And, We set up models of multivariate Neyman-Pearson decision and generalized the theory and algorithm of dualistic Neyman-Pearson decision to multivariate.
|