RSSI-based indoor tracking using the extended Kalman filter and circularly polarized antennas

Abstract

A tracking scenario comprising a mobile emitter node moving through an indoor environment covered by multiple anchor receivers is investigated in this work. A localization method based on received signal strength indicators (RSSI) and making use of the extended Kalman filter (EKF) and circularly polarized (CP) antennas is proposed. The EKF implements the position-velocity (PV) model, which assumes that the target is moving at a near-constant velocity during any given short time interval Δt. The measurement vector is composed of velocities in addition to RSSI values, which allow to deal with the error term between measurements and the propagation model directly. CP antennas are used on both the anchor nodes and the mobile node. These antennas are known to reduce the effects of multipath, especially those caused by single reflections. As a result, the RSSI values received in line of sight are more accurate and stable than those received from linearly polarized antennas. We tested our approach by tracking the movement of a robot following a predefined trajectory. The maximum location estimation error (LEE) is found to be 0.52 m. In addition, velocity changes are easily tracked during the target movement, which demonstrates the effectiveness of the proposed approach.