Effective error control of iterative localization for wireless sensor networks

Localization results are vulnerable to measurement errors, which allow for non-Gaussian or Gaussian measurements. In this paper we firstly propose an approximate estimation method for non-Gaussian range errors by analyzing the relationship between residuals and errors. Using the estimated range errors, we correct the residuals and propose a second localization method to improve the localization results. Then we formulate the error estimation caused by Gaussian measurements and evaluate our algorithm under various network topologies, range errors, location errors and number of anchors. We apply the error estimation method to iterative localization and explain the process of error propagation and accumulation. The error analysis under Gaussian measurements provides a guide to choose the anchors reasonably for controlling the localization errors. The experiments demonstrate the high efficiency of our proposed non-Gaussian error control method and evaluate the localization errors caused by range errors and location errors of anchors.