Haryanto, Dhanny Kurniawan and Karyono, Kanisius and Hutagalung, Samuel (2019) The Comparison Between Geo-magnetism and WiFi for Indoor Positioning System for Public Places. In: 2018 IEEE International Conference on Robotics, Biomimetics, and Intelligent Computational Systems (Robionetics ), 8-10 Aug. 2018, Bandung, Indonesia.
Full text not available from this repository.Abstract
The Indoor Positioning Systems (IPS) can be built based on several methods. This work compares two IPS methods which are Geo-magnetism and WiFi. This work compares the accuracy of the methods in public places. The Geo-magnetism method uses Earth's Magnetic Fingerprint which is generated from HCM8533L sensor. The WiFi method uses the Received Signal Strength (RSS) from NodeMCU device. The position references are using existing working access points around the location. All of the noises and disturbances are real noises generated by the traffic. The algorithms for magnetic fingerprint are K-Nearest Neighbors (KNN) and first position. The WiFi method uses multilateration algorithm. Geo-magnetism IPS uses 36 fingerprint data every 1.44m2. This approach results in 4.08m deviation with first position algorithm and 4.09m deviation with KNN. The WiFi RSS IPS uses multiple n values and multiple brand of access points. The n value of 1.8 results in 10.7m deviation and the value of 2 give 10.92m deviation. Based on these real world environment test, the Geo-magnetism IPS has a higher position accuracy than WiFi RSS IPS.
| Item Type: | Conference or Workshop Item (Paper) |
|---|---|
| Keywords: | Indoor Positioning System , Geo-magnetic , WiFi Received Signal Strength , K-Nearest Neighbors , Multilateration |
| Subjects: | 000 Computer Science, Information and General Works > 000 Computer Science, Knowledge and Systems > 006 Special Computer Methods 600 Technology (Applied Sciences) |
| Divisions: | Faculty of Engineering & Informatics > Computer Engineering |
| Depositing User: | Administrator UMN Library |
| Date Deposited: | 02 Dec 2021 18:08 |
| Last Modified: | 09 May 2023 05:40 |
| URI: | https://kc.umn.ac.id/id/eprint/19347 |
Actions (login required)
 |
View Item |
