Indian Journal of Engineering & Materials Sciences

Open Access Open Access  Restricted Access Subscription Access

Conventional System to Deep Learning Based Indoor Positioning System


Affiliations
1 University Institute of Engineering and Technology, Panjab University, Chandigarh 160 014, India
2 Centre for Development of Advanced Computing, Mohali 160 071, India

This review article presents the key fundamentals of indoor positioning system (IPS) and its progressing footprints. The need of IPS and work done with methodology adopted to implement IPS for various applications have been discussed. The evolution from conventional to deep learning (DL) has been presented, addressing various challenges existing in conventional IPS like poor localization, improper accuracy, non-line-of-sight problems, instability of signal due to fading, requirements of large infrastructure, data-set and labour, high cost, and their existing solutions have been disclosed. Furtherin order to compute the indoor positioning with acute precision various advanced positioning technologies including sensor fusion, artificial Intelligence (AI), and hybrid technologies have been explored. The issues and challenges existing in currentIPS technology have been presented and future insights to work in this direction have also been provided.

Keywords

Artificial intelligence (AI), Deep learning (DL), Global positioning system (GPS), Indoor positioning (IP), Reliability, Sensor fusion (SF)
User
Notifications
Font Size

Abstract Views: 27




  • Conventional System to Deep Learning Based Indoor Positioning System

Abstract Views: 27  | 

Authors

Shiva Sharma
University Institute of Engineering and Technology, Panjab University, Chandigarh 160 014, India
Naresh Kumar
University Institute of Engineering and Technology, Panjab University, Chandigarh 160 014, India
Manjit Kaur
Centre for Development of Advanced Computing, Mohali 160 071, India

Abstract


This review article presents the key fundamentals of indoor positioning system (IPS) and its progressing footprints. The need of IPS and work done with methodology adopted to implement IPS for various applications have been discussed. The evolution from conventional to deep learning (DL) has been presented, addressing various challenges existing in conventional IPS like poor localization, improper accuracy, non-line-of-sight problems, instability of signal due to fading, requirements of large infrastructure, data-set and labour, high cost, and their existing solutions have been disclosed. Furtherin order to compute the indoor positioning with acute precision various advanced positioning technologies including sensor fusion, artificial Intelligence (AI), and hybrid technologies have been explored. The issues and challenges existing in currentIPS technology have been presented and future insights to work in this direction have also been provided.

Keywords


Artificial intelligence (AI), Deep learning (DL), Global positioning system (GPS), Indoor positioning (IP), Reliability, Sensor fusion (SF)