Multisensor Decision And Estimation Fusion

Research activity in multisensor decision and estimation fusion problems has significantly increased over the last years of the 20th century. Distributed decision and estimation fusion problems for cases with statistically independent observations - or observation noises - have received the most attention, while problems with statistically dependent observations have been given much less consideration. This title provides a more complete treatment of the fundamentals of multisensor decision and estimation fusion in order to deal with general random observations or observation noises that are correlated across the sensors. Progress is presented in two ways. For multisensor decision fusion with general sensor observations given a fixed fusion rule, the book demonstrates a necessary condition for optimum sensor rules-they must be a fixed point of an integral operator related to given conditional probability densities. In particular, the book presents unified/optimal fusion rules for some specific decision systems. For the multisensor point estimation fusion problem, a general version of the linear unbiased minimum variance estimation fusion rule is developed. In addition, several alternative interval estimation fusion methods are proposed.