Enhanced indoor localization through crowd sensing

Abstract

In localization tasks, one typically assumes a statistical model of the observations, where the model quantifies the observations by exploiting interrelationships based on geometry. These models might incorporate unknown parameters that, in general, are functions of space. In this article, we propose a crowd sensing method for estimating a spatial field of a quantity (e.g., ranging biases due to line-of-sight/non-line-of-sight or path-loss parameter) allowing for improved indoor localization. Our method takes advantage of the information provided by various users that navigate the area of interest. The proposed learning approach is based on Gaussian processes and its computational cost does not increase with the number of measurements. We present numerical results that show how the proposed method estimates a spatial field of biases and how these estimates lead to much improved performance in estimation of user positions.

Estimation Of Spatial Fields Of Nlos/Los Conditions For Improved Localization In Indoor Environments

“…[n] ⌫ = 0.1 m. For each user, we ran the proposed algorithm with 500 particles and the method in [22] with an extended Kalman filter (EKF). For comparison, a simple technique assuming always LOS (and thus, zero bias) and a method with perfect knowledge of bias at every instant, both based on EKF, were also considered.…”

“…For each iteration, the crowd based methods started without any knowledge about the true bias field, taking for the first user as initial value b c c,0 = I. It is worth noting that in [22] an initial training phase was considered in which several collaborating users navigated the area along prescribed trajectories. By contrast, in this paper we focus on the challenging situation where no training phase is considered.…”

“…Specifically, Fig. 2 shows the results for the crowd-based method with EKF in [22]. It can be observed that as the number of users grows, the crowd mapping approach outperforms the simple method assuming LOS.…”

“…Based on these measurements, this unit has to estimate the state vector x t,k and to update the estimate of the coefficients c [n] , 8n. To that end, in [22] a "loose coupling" approach was proposed in which two separate methods for estimating the two sets of unknowns "communicate" with each other throughout the estimation. Here we propose a different and improved version of this approach.…”

“…GPs have already been used in the literature of cognitive radio for detecting the presence of primary users, e.g., [21]. Our method follows a line of reasoning presented in our previous paper [22], where we have adopted a "loose coupling" approach with two separate estimators, one for the position and one for the field. These estimators "ping-pong" their estimates as new measurements are being acquired.…”