Indoor Localization Based WiFi Signal Strength Using Support Vector Machine
Hani Rubiani (a*), Sulidar Fitri (b), Muhammad Taufiq (b), Mujiarto (c*)
a) Department of Electrical Engineering, Universitas Muhammadiyah Tasikmalaya
*hani.rubiani[at]umtas.ac.id
b) Department of Information Technology Education, Universitas Muhammadiyah Tasikmalaya
c) Department of Mechanical Engineering, Universitas Muhammadiyah Tasikmalaya, Jl. Tamansari Km. 2,5, Mulyasari, Tamansari, Tasikmalaya 46196, Indonesia
*mujiarto[at]umtas.ac.id
Abstract
Estimating the location of Object in an indoor environment poses a fundamental challenge in ubiquitous computing. Indoor localization based on signal strength by utilizing devices in buildings such as WiFi signals is increasingly being done. To determine the user-s position using the algorithm of the received signal strength. This paper shows that contrary to popular belief an indoor localization system based on WiFi fingerprints using Support Vector Machine (SVM) method. The performed experiments using 14480 datasets and 302 classes, collected from real world environments in building, and the comparison with Naïve Bayes confirm the effectiveness of SVM-based localization proposal. Experimental results show that the system achieves a correct classification rate of around 88% and minimum average error distance 4.61 meters compared to Naïve Bayes for correct classification rate of around 67% and minimum average error distance 6.21 meters.
Keywords: Support vector machine; SVM; Naive bayes; Indoor localization
Topic: Computer and Communication Engineering