Deep Learning-Based Indoor Positioning System using Fingerprints

Citation

Mazlan, Aqilah and Ng, Yin Hoe (2022) Deep Learning-Based Indoor Positioning System using Fingerprints. Periodic Research Publication, Faculty of Engineering. (Unpublished)

[img] Text
13_1161104444 Aqilah Binti Mazlan_YHNg_FYP2 Poster.pdf
Restricted to Repository staff only

Download (1MB)

Abstract

In indoor environments, fingerprint based indoor positioning systems (F-IPS) may just be the solution for issues related to the demand of line of sight Recently, deep neural networks were incorporated to F-IPS because traditional machine learning for position prediction only results in low positioning accuracy. Nevertheless, a DNNIPS fails to guarantee high accuracy in dynamic environments as it is sensitive to changes in the input data. Thus, convolutional neural network (CNN) is recommended to replace DNN due to its capability to learn the overall topology of fingerprinting images and capture highly abstract features. However, CNN-IPS has limitation regarding its storage and processing requirements when it is implemented on resource limited devices due to the convolution process and image representation. In this work, the positioning accuracy of the simplified CNN IPS is improved by incorporating knowledge distillation (KD) so that the more complex CNNs can distil its knowledge to smaller and more simplified CNNs.

Item Type: Other
Subjects: Q Science > QA Mathematics > QA71-90 Instruments and machines > QA75.5-76.95 Electronic computers. Computer science > QA76.75-76.765 Computer software
Divisions: Faculty of Engineering (FOE)
Depositing User: Assoc. Dr Chee Pun Ooi
Date Deposited: 29 Nov 2022 01:02
Last Modified: 29 Nov 2022 01:02
URII: http://shdl.mmu.edu.my/id/eprint/10642

Downloads

Downloads per month over past year

View ItemEdit (login required)