Iterative Maximum Likelihood Estimation Algorithm: Leveraging Building Information and Sensing Infrastructure for Localization during Emergencies

Abstract

Knowing real-time locations of first responders and occupants at building emergency scenes can be extremely helpful in improving the efficiency of first-response operations and reducing the chances of fatalities and injuries. This paper introduces an iterative maximum likelihood estimation (IMLE) indoor localization algorithm for supporting building emergency response operations. The algorithm uses radio frequency (RF) signal data, collected by existing sensing infrastructure in a building, and incorporates building geometric information available in building information models (BIMs). The algorithm integrates a maximum likelihood estimation (MLE) method for estimating the parameter values of a radio signal model and the locations of targets. It also introduces a novel iterative computational process for offsetting the effect of wall-related RF signal attenuations on the localization accuracy. The IMLE algorithm was evaluated in two simulated building emergency scenarios. The results reported a room-level accuracy of over 95% and a coordinate-level accuracy of over 0.84 m. The robustness of the algorithm was also tested to evaluate its applicability during emergencies. In both simulated scenarios, the algorithm yielded acceptable accuracies (room-level accuracy over 80% and coordinate-level accuracy over 2.0 m) with as few as four transmitters or three transceivers.