Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

BiCSI: A Binary Encoding and Fingerprint-Based Matching Algorithm for Wi-Fi Indoor Positioning

Pei Tang, Jingtao Guo, Ivan Wang-Hei Ho

TL;DR

BiCSI addresses indoor Wi-Fi positioning by converting CSI into binary sequences and applying fingerprint-based matching with Hamming-distance similarity in a two-stage offline-online workflow. It reports average accuracy above 98% and MAE below 3 cm across scenarios, while reducing feature-vector storage to kilobytes and maintaining robustness over time and varying similarity metrics. The method demonstrates strong performance against traditional similarity-based and ML-based approaches and shows resilience to time shifts and environmental changes. Its edge-friendly design, low data requirements, and flexibility make it a practical alternative for precise indoor localization of semi-stationary targets.

Abstract

Traditional global positioning systems often underperform indoors, whereas Wi-Fi has become an effective medium for various radio sensing services. Specifically, utilizing channel state information (CSI) from Wi-Fi networks provides a non-contact method for precise indoor positioning; yet, accurately interpreting the complex CSI matrix to develop a reliable strategy for physical similarity measurement remains challenging. This paper presents BiCSI, which merges binary encoding with fingerprint-based techniques to improve position matching for detecting semi-stationary targets. Inspired by gene sequencing processes, BiCSI initially converts CSI matrices into binary sequences and employs Hamming distances to evaluate signal similarity. The results show that BiCSI achieves an average accuracy above 98% and a mean absolute error (MAE) of less than three centimeters, outperforming algorithms directly dependent on physical measurements by at least two-fold. Moreover, the proposed method for extracting feature vectors from CSI matrices as fingerprints significantly reduces data storage requirements to the kilobyte range, far below the megabytes typically required by conventional machine learning models. Additionally, the results demonstrate that the proposed algorithm adapts well to multiple physical similarity metrics, and remains robust over different time periods, enhancing its utility and versatility in various scenarios.

BiCSI: A Binary Encoding and Fingerprint-Based Matching Algorithm for Wi-Fi Indoor Positioning

TL;DR

BiCSI addresses indoor Wi-Fi positioning by converting CSI into binary sequences and applying fingerprint-based matching with Hamming-distance similarity in a two-stage offline-online workflow. It reports average accuracy above 98% and MAE below 3 cm across scenarios, while reducing feature-vector storage to kilobytes and maintaining robustness over time and varying similarity metrics. The method demonstrates strong performance against traditional similarity-based and ML-based approaches and shows resilience to time shifts and environmental changes. Its edge-friendly design, low data requirements, and flexibility make it a practical alternative for precise indoor localization of semi-stationary targets.

Abstract

Traditional global positioning systems often underperform indoors, whereas Wi-Fi has become an effective medium for various radio sensing services. Specifically, utilizing channel state information (CSI) from Wi-Fi networks provides a non-contact method for precise indoor positioning; yet, accurately interpreting the complex CSI matrix to develop a reliable strategy for physical similarity measurement remains challenging. This paper presents BiCSI, which merges binary encoding with fingerprint-based techniques to improve position matching for detecting semi-stationary targets. Inspired by gene sequencing processes, BiCSI initially converts CSI matrices into binary sequences and employs Hamming distances to evaluate signal similarity. The results show that BiCSI achieves an average accuracy above 98% and a mean absolute error (MAE) of less than three centimeters, outperforming algorithms directly dependent on physical measurements by at least two-fold. Moreover, the proposed method for extracting feature vectors from CSI matrices as fingerprints significantly reduces data storage requirements to the kilobyte range, far below the megabytes typically required by conventional machine learning models. Additionally, the results demonstrate that the proposed algorithm adapts well to multiple physical similarity metrics, and remains robust over different time periods, enhancing its utility and versatility in various scenarios.

Paper Structure

This paper contains 14 sections, 9 equations, 11 figures, 3 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Backgroud
  3. Fingerprint-Based Localization
  4. RSSI and CSI
  5. Definitions of Target States
  6. THE PROPOSED ALGORITHM
  7. EXPERIMENTAL RESULTS AND EVALUATION
  8. Experiment Setup
  9. Algorithm Performance in Different Scenarios
  10. Impact of Different Similarity Measurements on the Algorithm
  11. The Impact of Different Thresholds on the Algorithm
  12. Algorithm Robustness Against Time
  13. Impact of Transceiver Distances
  14. CONCLUSION

Figures (11)

  • Figure 1: The flowchart of the proposed algorithm.
  • Figure 2: The CDF of CSI amplitudes.
  • Figure 3: Five scenarios. (a) Meeting room with Scenario A1-A3. (b) Scenario B: Lecture hall. (c) Scenario C: Classroom. (d) Experiment setup.
  • Figure 4: The MAE of each algorithm across scenarios.
  • Figure 5: The accuracy of each algorithm on each scenario.
  • ...and 6 more figures