BlueDetect: An iBeacon-Enabled Scheme for Accurate and Energy-Efficient Indoor-Outdoor Detection and Seamless Location-Based Service

Han Zou, Hao Jiang, Yiwen Luo, Jianjie Zhu, Xiaoxuan Lu, Lihua Xie

TL;DR

This paper introduces BlueDetect, a novel system that accurately and efficiently determines whether a mobile device is indoors, outdoors, or in a semi-outdoor environment using low-power iBeacon technology. This allows for intelligent management of power-hungry sensors like GPS and WiFi, enabling seamless location-based services and significantly reducing battery consumption.

What Problem Does This Paper Address?

Accurate location and environmental context (indoor/outdoor) are crucial for many applications, from personalized services to smart building management. While GPS excels outdoors and various Indoor Positioning Systems (IPS) exist for indoor environments, a significant challenge lies in reliably distinguishing between these states and, importantly, handling "semi-outdoor" environments (e.g., covered corridors, parking garages). These areas exhibit characteristics of both indoor and outdoor settings, making sole reliance on GPS or traditional IPS insufficient and often leading to poor performance and high battery drain. Existing solutions for indoor-outdoor detection often suffer from accuracy issues, slow response times, or excessive power consumption, particularly when dealing with these ambiguous semi-outdoor zones.

Methods

BlueDetect employs a multi-stage approach leveraging iBeacon technology, which is built on Bluetooth Low Energy (BLE).

  • Indoor-Outdoor (IO) Detection: The system transitions between GPS (outdoors), Bluetooth/iBeacon (semi-outdoors), and WiFi (indoors).
  • Outdoor to Semi-Outdoor: This transition is triggered by a significant decrease (over 20%) in the mean GPS Signal-to-Noise Ratio (SNR) over a short duration. Upon detection, Bluetooth is activated. If at least two iBeacons are detected with sufficient signal strength, the system confirms a semi-outdoor environment and turns off GPS.
  • Semi-Outdoor to Indoor: This transition is managed by a pair of iBeacons placed at entrances. By comparing the received signal strength (RSS) of these two beacons, the system infers whether the user is entering or exiting an indoor space. Upon confirming entry, WiFi is activated for indoor positioning.
  • Indoor to Semi-Outdoor: When exiting an indoor environment, WiFi is turned off, and the iBeacon mode is activated.
  • Semi-Outdoor Localization: In semi-outdoor environments, where GPS and WiFi are unreliable and power-hungry, BlueDetect utilizes iBeacons. A weighted path loss (WPL) model is employed to estimate the device's location based on RSS measurements from deployed iBeacons. The model accounts for Non-Line-of-Sight (NLOS) conditions and variations in mobile device orientation, which were found to significantly impact RSS.
  • Indoor Localization: For indoor environments, the system leverages existing WiFi infrastructure and WiFi-based Indoor Positioning Systems (IPS).

Datasets

Not explicitly discussed in terms of pre-trained datasets. The evaluation relies on experimental data collected in a university campus environment. The authors designed a specific walking route that encompassed all three environment types (indoor, semi-outdoor, outdoor) and their transitions.

Key Insights and Ideas

  • iBeacon as a Low-Power Enabler: The core innovation is the effective use of iBeacons not just for localization but as a crucial component for intelligent sensor management and energy saving in semi-outdoor environments.
  • Semi-Outdoor Environment as a Critical Gap: The paper highlights the inadequacy of existing GPS and IPS solutions in semi-outdoor areas and proposes a practical solution.
  • SNR as a Robust Transition Trigger: Using the decrease in mean GPS SNR as a trigger for transitioning from outdoor to semi-outdoor is a simple yet effective method that accounts for signal degradation without requiring precise GPS coordinates.
  • Dual-iBeacon Entrance Detection: Employing two iBeacons at building entrances provides a robust and intuitive way to detect entry/exit transitions, outperforming simple signal strength thresholds.
  • Accounting for Real-World iBeacon Imperfections: The detailed analysis of NLOS and device orientation effects on iBeacon RSS is crucial for developing an accurate localization model in challenging semi-outdoor settings.

Key Results

  • IO Detection Accuracy: BlueDetect achieved an overall IO detection accuracy of 96.2%, significantly outperforming GPS-IO (56.1%) and IODetector (78.8%). It accurately captured all transitions between environments.
  • Semi-Outdoor Localization Accuracy: BlueDetect achieved an average localization error of 2.18 m in semi-outdoor areas, an improvement of approximately 89.87% compared to GPS (21.53 m). The 90th percentile error for BlueDetect was 7.94 m, compared to 37.33 m for GPS.
  • Power Consumption: BlueDetect demonstrated the lowest power consumption among the evaluated methods, consuming only 119 mAh in 30 minutes. This is significantly less than IODetector (162 mAh) and GPS-IO (213 mAh).

Limitations and Future Work

  • Limitations:
    • The accuracy of the iBeacon localization model is dependent on the calibration of the path loss parameters (Po and α), which may need to be re-calibrated for different environments or beacon types.
  • The performance of the system relies on the presence and proper deployment of iBeacons in semi-outdoor areas.
  • While the paper mentions the influence of weather on GPS, it notes that the influence of weather conditions on BLE beacon RSS is negligible based on their experimental statistics, which might not hold true in all extreme conditions.
  • Future Work:
    • The authors suggest extending BlueDetect for other services like geo-fencing and floor identification.
  • Further research could explore more sophisticated machine learning models for iBeacon localization to potentially improve accuracy further.
  • Investigating the scalability of iBeacon deployment and management for larger areas.

Summary updated

BibTeX
@inproceedings{zou2016bluedetect,
  author = {Han Zou and Hao Jiang and Yiwen Luo and Jianjie Zhu and Xiaoxuan Lu and Lihua Xie},
  title = {{BlueDetect: An iBeacon-Enabled Scheme for Accurate and Energy-Efficient Indoor-Outdoor Detection and Seamless Location-Based Service}},
  booktitle = {Sensors},
  year = {2016},
  url = {https://openreview.net/forum?id=kG14_MOda9Rd},
}
Figure