4. Lack of the Golden Standard and Missing Measurements

We propose a new error modeling and optimization-based localization approach for sensor networks in presence of range measurement noise. The approach is solely based on the concept of consistency. The error models are constructed using nonparametric statistical techniques; they do not only indicate the most likely error but also provide the likelihood distribution of particular errors occurring. The models are evaluated using the learn-and-test method and served as the OFs for the task of localization. In addition, we also present a localized localization algorithm where a specified communication cost or the location accuracy is guaranteed while optimizing the other. We evaluate the approach (1) in both GPS-based and GPS-less scenarios; (2) in both centralized and localized manners; (3) in 1D, 2D, and 3D spaces; and (4) in the case when error models are not available a priori, on sets of acoustic ranging-based distance measurements recorded by actual deployed sensor networks. The experimental evaluation indicates that localization of only a few centimeters is consistently achieved when the average and median distance measurement errors are more than a meter, even when the nodes have a low connectivity. Furthermore, we compare the relative performance in terms of location accuracy with several state-of-the-art localization approaches. Finally, several insightful implications about the required conditions for accurate Location Discovery are concluded by analyzing the experimental results.