Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

ML-based Approaches for Wireless NLOS Localization: Input Representations and Uncertainty Estimation

Rafayel Darbinyan, Hrant Khachatrian, Rafayel Mkrtchyan, Theofanis P. Raptis

TL;DR

Two convolutional neural networks are designed and it is demonstrated that, although not significantly improving the NLOS localization performance, they are able to support richer prediction outputs, thus allowing deeper analysis of the predictions.

Abstract

The challenging problem of non-line-of-sight (NLOS) localization is critical for many wireless networking applications. The lack of available datasets has made NLOS localization difficult to tackle with ML-driven methods, but recent developments in synthetic dataset generation have provided new opportunities for research. This paper explores three different input representations: (i) single wireless radio path features, (ii) wireless radio link features (multi-path), and (iii) image-based representations. Inspired by the two latter new representations, we design two convolutional neural networks (CNNs) and we demonstrate that, although not significantly improving the NLOS localization performance, they are able to support richer prediction outputs, thus allowing deeper analysis of the predictions. In particular, the richer outputs enable reliable identification of non-trustworthy predictions and support the prediction of the top-K candidate locations for a given instance. We also measure how the availability of various features (such as angles of signal departure and arrival) affects the model's performance, providing insights about the types of data that should be collected for enhanced NLOS localization. Our insights motivate future work on building more efficient neural architectures and input representations for improved NLOS localization performance, along with additional useful application features.

ML-based Approaches for Wireless NLOS Localization: Input Representations and Uncertainty Estimation

TL;DR

Two convolutional neural networks are designed and it is demonstrated that, although not significantly improving the NLOS localization performance, they are able to support richer prediction outputs, thus allowing deeper analysis of the predictions.

Abstract

The challenging problem of non-line-of-sight (NLOS) localization is critical for many wireless networking applications. The lack of available datasets has made NLOS localization difficult to tackle with ML-driven methods, but recent developments in synthetic dataset generation have provided new opportunities for research. This paper explores three different input representations: (i) single wireless radio path features, (ii) wireless radio link features (multi-path), and (iii) image-based representations. Inspired by the two latter new representations, we design two convolutional neural networks (CNNs) and we demonstrate that, although not significantly improving the NLOS localization performance, they are able to support richer prediction outputs, thus allowing deeper analysis of the predictions. In particular, the richer outputs enable reliable identification of non-trustworthy predictions and support the prediction of the top-K candidate locations for a given instance. We also measure how the availability of various features (such as angles of signal departure and arrival) affects the model's performance, providing insights about the types of data that should be collected for enhanced NLOS localization. Our insights motivate future work on building more efficient neural architectures and input representations for improved NLOS localization performance, along with additional useful application features.
Paper Structure (17 sections, 1 equation, 9 figures, 3 tables)

This paper contains 17 sections, 1 equation, 9 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Related works
  3. Problem formal definition
  4. ML-driven approaches
  5. A baseline method with vector inputs
  6. A CNN with matrix inputs
  7. A CNN with image inputs
  8. Performance evaluation
  9. The synthetic dataset
  10. Benchmark
  11. Implementation details
  12. Results
  13. Predicting top five locations
  14. Identifying Incorrect Predictions
  15. Ablation over input variables
  16. ...and 2 more sections

Figures (9)

  • Figure 1: An example instance of the problem, with $l = 1, i = 3, j = 5, K_{3,5} = 2$.
  • Figure 2: Input and output representations for UNet
  • Figure 3: CNN-based neural architectures defined in \ref{['sec:models']}.
  • Figure 4: Accuracy of the MLP and UNet models with respect to the maximum tolerated error in meters.
  • Figure 5: Accuracy of the UNet model with various hyperparameters on the validation set of WAIR-D dataset.
  • ...and 4 more figures