FM-Based Outdoor Fingerprint Location Using DNN Algorithm for Large-Scale Internet of Things

Yichen Pan, Jae Soo Kim

Research output: Contribution to journalArticlepeer-review

1 Scopus citations

Abstract

The generation of positioning technology has a great impact on human life and the development of science and technology, especially with the rapid growth of wireless networks and communication technology today. The Internet of Things (IoT) technology has penetrated all walks of life and even the daily life of human beings. More and more physical devices are connected to the network for information exchange and sharing. To enable large-scale IoT devices and services, several newly developing IoT technologies, Low Power Wide Area Network(LPWAN)have emerged. The FM signal based fingerprint outdoor positioning technology in this paper is a low-cost and low energy consumption positioning method to adapt to large-scale IoT devices. Through collecting FM signal strength and other effective information, fingerprint databases are constructed and the data are trained by using Deep Neural Networks(DNN) to reduce accuracy differences. The Final location information can be obtained by this method. Experimental results show that the accuracy of this method is 95.57%, which can effectively improve the accuracy of FM outdoor positioning.

Original languageEnglish
Pages (from-to)1650-1657
Number of pages8
JournalJournal of Korean Institute of Communications and Information Sciences
Volume46
Issue number10
DOIs
StatePublished - Oct 2021

Keywords

  • Deep Learning
  • Fingerprints
  • FM radio
  • Internet of Things
  • Positioning

Fingerprint

Dive into the research topics of 'FM-Based Outdoor Fingerprint Location Using DNN Algorithm for Large-Scale Internet of Things'. Together they form a unique fingerprint.

Cite this