A new look at the Barankin lower bound

This paper is primarily tutorial in nature and presents an alternative derivation of the Barankin estimator. The approach presented here places clearly in evidence the significance of the choice of parameter points that one must make in obtaining a Barankin lower bound. Specifically, for any finite set of parameter points, the corresponding Barankin bound is the variance of the locally best estimator that is constrained to be unbiased at the chosen parameter points. In addition, the derivation of the Barankin theory presented here provides a new interpretation of the Cramer-Rao and Bhattacharyya bounds.