A greedy approach to beacon placement for localization

The need for location aware applications is increasing with the popularity of ubiquitous computing. Localization which is the process of estimating the location of a node is an important building block in these location aware applications. In a typical localization technique, beacons whose locations are known a priori, acts as reference in estimating the location of user nodes. It has been observed that the placement and density of these beacons have strong impact on the accuracy of localization. The intuitive approaches - uniform and dense beacon placement will not suffice in practical situations due to the presence of nonuniform obstructions and self interference of beacons which results in non-uniform signal propagation error. The effect of beacon density on localization error is characterized as decrease in error with increase in beacon density up to a point termed as threshold density and there after the effect is minimal. The number of beacons that will be in range at this threshold density is threshold degree mu_thresh. To address the problems of uniform and dense placement, we will present a novel beacon placement algorithm in this paper which uses greedy approach, to achieve threshold degree condition at all points and observed an improvement in accuracy over uniform placement.