Fundamental limits for distributed estimation using a sensor field

In distributed statistical inference, it is of interest to relate the statistical properties of different estimates of a parameter obtained by users who have access to different sets of observations. Suppose there are a number of sources of interest, and each user has access to observations that are a combination of data emerging from a particular subset of sources. For a given class of users, the minimax risks achievable by the users are related to each other, in the special case when the observations may be thought of as coming from a location family. Applications are given to design and resource allocation problems in sensor networks.