Pseudo-linear localization using perturbed RSSI measurements and inaccurate anchor positions

Abstract

We consider the problem of RSSI-based self-localization by a resource-constrained mobile node given only a single perturbed observation of each RSSI measurement and inaccurate anchor positions. Most existing solutions assume additive independent zero-mean Gaussian perturbations in the observations. We consider a more realistic log-normal shadowing path-loss model for the radio propagation in which RSSI-based distance estimates follow log-normal distribution. We propose a bias-compensated pseudo-linear solution (PLS) using the weighted least-squares (WLS) method. The weights are estimated using the statistical properties of the perturbations in the RSSI-induced distance estimates and anchor position observations. The performance of PLS is evaluated over arbitrarily selected network geometries and compared with an existing WLS-based solution, which only accounts for the perturbations in the distance estimates. Simulation results show that PLS can substantially reduce the root-mean-square error and bias of the existing solution to almost half in many scenarios.