Title: Solving very large scale SDPs that arise from ad hoc wireless sensor
         network localization and other Euclidean geometry problems

Abstract:
We describe an iterative distributed and decomposed semidefinite programming
(SDP) method for solving localization problems that arise from ad hoc wireless
sensor network and other Euclidean distance geometry. Using the method, we 
can solve very large scale semidefinite programs which are intractable 
for the centralized methods.
Our decomposition scheme seems applicable to solving other semidefinite 
programs, and, like linear programming, is a most effective way to 
solver very large-scale problems in general.