We present here a novel approach where we identify a
within which a node is
to be found, in contrast to the existing approaches where no such confining region for a node can be guaranteed, but only the location could be estimated either with no definitive error bound or only with some probabilistic error. The location identification algorithm presented here minimizes the size of this region, using computational geometric methods. The proposed technique iteratively improves the
region of residence
of all the nodes in the network through the exchange of region information among neighbors in
) time, where
are the number of nodes and diameter of the network respectively. Simulation results also show encouraging results with this approach.