Abstract
The Performance of cellular long-term evolution (LTE) signals for indoor localization is evaluated. Two different designs of LTE software-defined receivers (SDRs), namely a code phase-based receiver and a carrier phase-based receiver, are presented and assessed experimentally indoors with LTE signals. A base/navigator framework is presented to deal with the unknown clock biases of the LTE eNodeBs. In this framework, the base receiver is placed outdoors, has knowledge of its own position, and makes pseudorange measurements to eNodeBs in the environment whose positions are known. The base transmits these pseudoranges to the indoor navigating receiver, which is also making pseudorange measurements to the same eNodeBs. The navigating receiver differences the base’s and navigator’s pseudoranges; hence, the unknown eNodeBs’ biases are eliminated. The navigator receiver is equipped with an inertial measurement unit (IMU), and the LTE pseudoranges and IMU measurements are tightly coupled using an extended Kalman filter (EKF). Two sets of experimental results are presented. First, it is demonstrated that the standalone carrier phase-based receiver yielded a more precise navigation solution than the code phase-based receiver, specifically a two-dimensional (2-D) position root mean-squared error (RMSE) of 5.09 m versus 11.76 m for an indoor trajectory of 109 m traversed in 50 seconds. Second, it is demonstrated that coupling the IMU with the carrier phase-based LTE receiver reduced the 2-D position RMSE to 2.92 m. Moreover, it is demonstrated that the proposed LTE-IMU system yielded a maximum error of 5.60 compared to 22.53 m for the IMU-only.