Matching Pursuit Based Joint Angle and Delay Estimation for Bluetooth Direction Finding

Indoor localization is a desired technique for both end-user and industrial applications. The support for Angle of Arrival (AoA) based Direction Finding with version 5.1 of the Bluetooth Core Specification recently generated a particular interest for AoA-based Bluetooth localization systems. Using state-of-the-art AoA estimation algorithms like the Joint Angle and Delay Estimation MUltiple SIgnal Classification (JADE MUSIC) algorithm, high localization accuracies can be achieved with such systems. However, these MUSIC-based algorithms come along with a rather high computational complexity that makes it challenging to implement them in practical systems. Proposing a compromise between localization accuracy and computational complexity, we present a novel Joint Angle and Delay Estimation Matching Pursuit (JADE MP) approach for AoA-based indoor localization. Using a publicly-available dataset from a 100 m
²
Bluetooth 5.1 localization testbed, an evaluation of the proposed algorithm is presented. It is shown that the proposed JADE MP approach can achieve almost the same accuracy as state-of-the-art systems, while being significantly faster at the same time.