Impact of antennas on the anchor-less indoor localization of a static IR-UWB pair

This paper investigates the impact of realistic antennas on joint anchor-less localization and indoor characterization based on Impulse Radio (IR) Ultra Wideband (UWB) communications. In this frame, the Maximum Averaged Likelihood (MAL) algorithm and its extended version are considered, both relying on a tree approach consisting in two stages. The first part of the process, which is common to both algorithms, exploits the cross-correlation between received and locally predicted paths. The second stage calculates the averaged likelihood of measured path parameters obtained in the previous step, but different measurements are used for MAL and extended MAL (eMAL). In the first algorithm, only the Angle of Incidence (AoI) and the Time of Arrival (ToA) are considered, whereas the eMAL tree algorithm also accounts for two couples of Angles of Departure (AoDs) and Angles of Arrival (AoAs). The estimation errors of both nodes coordinates and room dimension obtained with the two algorithms are compared for three realistic UWB antennas. Finally, the remaining algorithm-independent ambiguities (i.e. resulting from scenario and geometry) are discussed.