Wi-Fi Fingerprint Location Estimation System Based on Reliability

Sanghyun Son, Young‐Joon Park, Beomjun Kim, Yunju BaekThe Journal of Korean Institute of Communications and Information Sciences 2013 · DOI

Abstract

Wi-Fi fingerprinting is a standard technique for indoor positioning, yet its accuracy is often limited by the labor-intensive requirement for numerous physical reference points (RPs). This paper proposes a novel system that utilizes a reliability-based "scoremap" to enhance positioning precision while minimizing manual data collection. By dividing the service area into a fine grid and applying a Log-Distance Path Loss Model to generate virtual reference points, the system assigns reliability weights based on signal proximity and strength. Experimental results in a 20m × 22m indoor environment demonstrate that the proposed system reduces average distance error by 1.74 meters compared to the traditional RADAR method.

Introduction

Indoor positioning remains a significant challenge as GPS signals cannot effectively penetrate building structures. Wi-Fi fingerprinting has emerged as a popular alternative, relying on a database of Received Signal Strength Indicators (RSSI) collected at various reference points. However, traditional methods face a trade-off: high resolution requires a massive number of physical measurements, making the setup phase costly and time-consuming. Furthermore, environmental noise and signal fluctuations often degrade the accuracy of simple matching algorithms. This research addresses these limitations by introducing a reliability metric that bridges the gap between physics-based signal modeling and empirical data collection.

Methodology

The proposed system operates in two distinct phases: map generation and location estimation.

Map Generation (Offline Phase)

The service area is first divided into a high-resolution grid (e.g., 10cm × 10cm). Instead of measuring every grid cell manually, the system uses a limited set of physical RPs to calibrate a Log-Distance Path Loss Model. This model predicts signal strengths for every cell in the grid, effectively creating "virtual" reference points.

Grid division and the mapping of locations to signal vectors.

Reliability-Based Scoremap

The core innovation is the introduction of a reliability metric ((\beta)). This metric quantifies the trustworthiness of a predicted signal value based on two factors:

  1. Proximity: Predictions are weighted more heavily when they are closer to a physical measurement point.
  2. Signal Strength: Stronger signals are mathematically more stable within path-loss models, whereas weak signals are more susceptible to noise.

Multiple models from different RPs are merged using these reliability weights to create a final "Scoremap."

Location Estimation

During the online phase, when a device reports its current Wi-Fi signals, the system groups RSSI values into ranges to mitigate noise. It then identifies candidate groups on the scoremap and sums their reliability scores. The grid cell with the highest cumulative reliability is selected as the estimated location.

The process of selecting reference points and calculating reliability for location estimation.

Experimental Results

The system was evaluated in a 20m × 22m laboratory at Pusan National University equipped with 8 Wi-Fi Access Points. The performance was compared against the industry-standard RADAR method, which uses Mean Squared Error (MSE) for fingerprint matching.

  • RADAR Average Error: 6.5 meters
  • Proposed System Average Error: 3.9 meters
  • Total Improvement: 1.74-meter reduction in average distance error.

The results indicate that the reliability-based approach is significantly more resilient to signal noise and provides higher granularity than traditional methods, even with a sparse set of physical reference points.

Conclusion

By integrating physics-based path-loss modeling with a novel reliability metric, this system provides a scalable solution for high-precision indoor positioning. Future work will focus on refining the path-loss variables to account for specific indoor obstacles like walls and glass, and optimizing the computational overhead for large-scale deployments in environments such as airports or shopping malls.

Summary updated

Abstract

Wi-Fi 기반 위치 추적 기술은 다양한 방법들이 제시되어 있으나 일반적으로 지정한 참조위치에서 측정한 신호세기 정보를 활용하는 핑거프린트 기법을 주로 이용한다. 그러나 핑거프린트 기법이 높은 해상도와 정밀도를 가지기 위해서는 다수의 참조위치를 필요로 하지만, 다수의 정보를 위해서는 많은 비용의 소모가 발생한다. 본 논문에서는 서비스 지역을 격자로 구분하고 가상의 참조위치를 바탕으로 신호 세기 모델을 만들어서 격자의 신호 세기정보를 예측하고 신뢰도를 부여하여 핑거프린트 맵을 생성한다. 이렇게 생성된 맵은, 위치 추정을 위해 입력된 신호 세기 리스트와 비교하여 추정하고자 하는 대상이 어떠한 격자에 있는지 비교하는데 사용한다. 제안하는 기법과 시스템의 성능을 평가하기 위해 <TEX>$20m{\times}22m$</TEX>의 실내 환경에서 기존의 핑거프린트 기법(RADAR)과 비교한 결과 제안하는 시스템을 이용할 때 위치 측정 에러가 감소하는 결과를 확인하였다. 그리고 실험 결과 기존의 방법에 비해 1.74미터의 에러가 감소한 것을 확인했다. Fingerprinting technique uses the radio signal strength measured reference locations is typically used, although there are many Wi-Fi based location tracking techniques. However, it needs numerous reference locations for precision and accuracy. This paper the analyzes problems of previous techniques and proposes a fingerprinting system using reliability based on a signal strength map. The system collects the signal strength data from a number of reference locations designated by the developer. And then it generates path-loss models to one of the access points for each reference location. These models calculate the predicted signal strength and reliability for a lattice. To evaluate proposed method and system performance, We perform experiments in a <TEX>$20m{\times}22m$</TEX> real indoor environment installed access points. According to the result, the proposed system reduced distance error than RADAR. Comparing the existing system, it reduced about 1.74m.
BibTeX
@article{son2013wi,
  author = {Sanghyun Son and Young‐Joon Park and Beomjun Kim and Yunju Baek},
  title = {{Wi-Fi Fingerprint Location Estimation System Based on Reliability}},
  journal = {The Journal of Korean Institute of Communications and Information Sciences},
  year = {2013},
  doi = {10.7840/kics.2013.38c.6.531},
}
Figure