WPI precision personnel locator: Inverse synthetic array reconciliation tomography

This paper describes the latest algorithm being developed by the Worcester Polytechnic Institute (WPI) Precision Personnel Location (PPL) project. Our goal is to produce a rapidly deployable, ad-hoc system that can achieve sub meter positioning accuracy in any type of emergency response scenario using available spectrum; specifically we wish to locate first responders in and around buildings. Previous work [1] has led to separate or loosely coupled approaches for fusing RF and inertial positioning data, as well as data from other sensors. The Inverse Synthetic Array Reconciliation Tomography (IS-ART) algorithm is a fusion of RF and inertial measurements that is fundamentally different from previous systems. Instead of fusing positioning results, we use the inertial displacements to coherently fuse successive RF captures. The benefits of this approach are twofold. First, the fusion depends only on relationships between inertial values spanning small time intervals, so that we do not accumulate large inertial drift errors. Second, the RF conditions at each point are expected to be very different because high multi-path environments are very sensitive to spatial perturbations. Since the multi-path signal components are changing rapidly with position we expect that only the direct path signals will be correlated between successive RF data captures taken from unique positions. In order for the direct path signals from different locations to be correlated these captures must be aligned for fusion with the inertial displacement data. The ISART algorithm performs this signal alignment and computes a metric based on this direct path correlation using the sample processing employed by the Singular Value Array Reconciliation Tomography (σART ) algorithm [2]; the ISART solution is the point where the metric is maximal.