Clock and Orientation-Robust Simultaneous Radio Localization and Mapping at Millimeter Wave Bands

TL;DR

This paper introduces a millimeter wave radio-SLAM method that accurately localizes and maps without clock or orientation info, using sparse channel estimation and compressed sensing.

Contribution

It presents a novel radio-SLAM algorithm that operates without clock synchronization or multiple anchors, leveraging high-resolution multipath parameter estimation at millimeter wave bands.

Findings

Achieves sub-meter localization accuracy in simulations.

Operates without clock synchronization or multiple anchors.

Uses compressed sensing for high-resolution multipath estimation.

Abstract

This paper proposes a radio simultaneous location and mapping (radio-SLAM) scheme based on sparse multipath channel estimation. By leveraging sparse channel estimation schemes at millimeter wave bands, namely high resolution estimates of the multipath angle of arrival (AoA), time difference of arrival (TDoA), and angle of departure (AoD), we develop a radio-SLAM algorithm that operates without any requirements of clock synchronization, receiver orientation knowledge, multiple anchor points, or two-way protocols. Thanks to the AoD information obtained via compressed sensing (CS) of the channel, the proposed scheme can estimate the receiver clock offset and orientation from a single anchor transmission, achieving sub-meter accuracy in a realistic typical channel simulation.