Some comments on distributed detection and estimation

In this paper we discuss the question of information and accuracy attainable in distributed processing as compared to central processing. An example is presented where distributed detection suffers zero loss in performance as compared to central detection. In the same example, if the problem considered is one of estimation rather than detection, then it is shown that distributed estimation suffers a loss as compared to central estimation. This shows that the distributed detection and the distributed estimation problems cannot be considered on an equivalent footing. Some comments regarding an accuracy bound in detection problems are also provided