Using ellipses to improve the estimation of the error bound on indoor positioning problems

IoT-based indoor positioning systems consider the average positioning error over a set of evaluation samples, among other proposals, as the global error metric. However, the error given is the same value for all the positions. Therefore, these systems do not measure uncertainty for each position estimation. Recently, a new method has been proposed to use error propagation theory to describe how the uncertainties of the RSSI measurements propagate through the calculations, using a circle centered on the estimated position. The goal is to give the smallest possible circle with the highest probability of certainty. In this paper, we propose replacing the circle with an ellipse, which better aligns with the inner structure in indoor spaces. We validated our proposed method using an open-access dataset, comparing the performance of both approaches: circles versus ellipses. Results show that ellipses give an area smaller than the circle but with similar certainty.