Channel Charting in Real-World Coordinates
🤖 AI Summary
To address the misalignment between channel charting outputs and true geographic coordinates, this paper proposes a label-free, geometry-agnostic real-coordinate mapping method. Leveraging only known access point (AP) locations and received signal reference power (RSRP) ratios, it achieves alignment from pseudo-coordinates to real-world Cartesian coordinates via self-supervised learning. The key contribution is a novel bilateration loss—introduced for the first time—that enables end-to-end coordinate system alignment without requiring ground-truth user equipment (UE) positions or explicit channel propagation models. The method jointly optimizes channel state information (CSI) embeddings and ray-tracing-based channel data. Evaluated on synthetic channel datasets, it significantly improves global geometric fidelity and reduces relative localization error by over 40%.
📝 Abstract
Channel charting is an emerging self-supervised method that maps channel state information (CSI) to a low-dimensional latent space, which represents pseudo-positions of user equipments (UEs). While this latent space preserves local geometry, i.e., nearby UEs are nearby in latent space, the pseudo-positions are in arbitrary coordinates and global geometry is not preserved. In order to enable channel charting in real-world coordinates, we propose a novel bilateration loss for multipoint wireless systems in which only the access point (AP) locations are known—no geometrical models or ground-truth UE position information is required. The idea behind this bilateration loss is to compare the received power at pairs of APs in order to determine whether a UE should be placed closer to one AP or the other in latent space. We demonstrate the efficacy of our method using channel vectors from a commercial ray-tracer.
🔎 Similar Papers
No similar papers found.
