Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Global Scale Self-Supervised Channel Charting with Sensor Fusion

Omid Esrafilian, Mohsen Ahadi, Florian Kaltenberger, David Gesbert

TL;DR

This work tackles global-scale, sub-meter localization in wireless networks via self-supervised channel charting (CC). It introduces a neural mapping f_{ theta} from CSI inputs to a low-dimensional chart, leveraging ToA measurements from nearby TRPs and training-time depth data from a 2D laser scanner to achieve global alignment without ground-truth labels. The method employs a bilateration-inspired loss augmented by ICP-based depth displacement and includes an offset correction step, achieving sub-meter accuracy in 90% of trials with only two LoS TRPs and outperforming both state-of-the-art CC methods and traditional triangulation. This sensor-fusion CC approach advances scalable 6G sensing and positioning, with potential extensions to NLoS settings and the inclusion of AoA measurements for further improvements.

Abstract

The sensing and positioning capabilities foreseen in 6G have great potential for technology advancements in various domains, such as future smart cities and industrial use cases. Channel charting has emerged as a promising technology in recent years for radio frequency-based sensing and localization. However, the accuracy of these techniques is yet far behind the numbers envisioned in 6G. To reduce this gap, in this paper, we propose a novel channel charting technique capitalizing on the time of arrival measurements from surrounding Transmission Reception Points (TRPs) along with their locations and leveraging sensor fusion in channel charting by incorporating laser scanner data during the training phase of our algorithm. The proposed algorithm remains self-supervised during training and test phases, requiring no geometrical models or user position ground truth. Simulation results validate the achievement of a sub-meter level localization accuracy using our algorithm 90% of the time, outperforming the state-of-the-art channel charting techniques and the traditional triangulation-based approaches.

Global Scale Self-Supervised Channel Charting with Sensor Fusion

TL;DR

This work tackles global-scale, sub-meter localization in wireless networks via self-supervised channel charting (CC). It introduces a neural mapping f_{ theta} from CSI inputs to a low-dimensional chart, leveraging ToA measurements from nearby TRPs and training-time depth data from a 2D laser scanner to achieve global alignment without ground-truth labels. The method employs a bilateration-inspired loss augmented by ICP-based depth displacement and includes an offset correction step, achieving sub-meter accuracy in 90% of trials with only two LoS TRPs and outperforming both state-of-the-art CC methods and traditional triangulation. This sensor-fusion CC approach advances scalable 6G sensing and positioning, with potential extensions to NLoS settings and the inclusion of AoA measurements for further improvements.

Abstract

The sensing and positioning capabilities foreseen in 6G have great potential for technology advancements in various domains, such as future smart cities and industrial use cases. Channel charting has emerged as a promising technology in recent years for radio frequency-based sensing and localization. However, the accuracy of these techniques is yet far behind the numbers envisioned in 6G. To reduce this gap, in this paper, we propose a novel channel charting technique capitalizing on the time of arrival measurements from surrounding Transmission Reception Points (TRPs) along with their locations and leveraging sensor fusion in channel charting by incorporating laser scanner data during the training phase of our algorithm. The proposed algorithm remains self-supervised during training and test phases, requiring no geometrical models or user position ground truth. Simulation results validate the achievement of a sub-meter level localization accuracy using our algorithm 90% of the time, outperforming the state-of-the-art channel charting techniques and the traditional triangulation-based approaches.
Paper Structure (10 sections, 15 equations, 3 figures, 2 tables)

This paper contains 10 sections, 15 equations, 3 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Related work
  3. System Model
  4. Data Preprocessing and Feature Extraction
  5. Deep Channel Charting
  6. Channel Charting Using Data Fusion
  7. Offset Estimation
  8. Evaluations
  9. Simulation and Results
  10. Conclusions and Future Work

Figures (3)

  • Figure 1: Training and test phases of our proposed algorithm.
  • Figure 2: True trajectory during the training phase.
  • Figure 3: Figures (a) to (e) are the results for test Dataset 1, and figures (f) to (j) are the results for test Dataset 2.