A Map-Augmented Positioning Method for UWB Network in Narrow and Enclosed Spaces

The location of the items in the warehouses forms the fundamental spatial information for the Internet of Things. However, there are numerous corridors consisting of warehouses. The narrow and elongated geometry of a corridor or tunnel, also restricts the shape of the ultrawide band (UWB) anchors’ network, resulting in nearly ill-conditioned positioning estimation and poor positioning precision. To address these issues, this article proposes a novel map-augmented least squares (MLS) method for user positioning in narrow and enclosed spaces (NES). MLS achieves the following improvements: 1) MLS converts the disadvantage of narrow width in NES, into a useful condition for positioning estimation. 2) MLS offers initial coordinates with reasonable precision close to the actual situation, significantly enhancing the positioning precision in vertical and narrow directions. Many tests have been conducted with different numbers of UWB anchors, and diverse aspect ratios. Compared to conventional least squares (LS), vertical-constrained LS (VLS), Levenberg–Marquardt method (LM), and particle filter method (PF), testing results indicated: 1) In the case of three UWB anchors, positioning precision of MLS in narrow direction was improved by two times relative to LS and LM, and by 70% relative to VLS and PF. 2) In the case of two UWB anchors, MLS overcame the issues that, LS could not complete 3-D, and VLS suffered ambiguous solutions. 3) As aspect ratios increased and the number of UWB anchors decreased, positioning precision of MLS remained stable, while positioning precisions of LS, VLS, and LM deteriorated significantly.