Reference : Deep Learning-Based Device-Free Localization in Wireless Sensor Networks

	Document type :	Scientific congresses, symposiums and conference proceedings : Unpublished conference
	Discipline(s) :	Engineering, computing & technology : Electrical & electronics engineering
	Focus Areas :	Security, Reliability and Trust
	To cite this reference:	http://hdl.handle.net/10993/51597

Title :	Deep Learning-Based Device-Free Localization in Wireless Sensor Networks
Language :	English
Author, co-author :	Abdullah, Osamah [Alma'moon University College, Baghdad, Iraq > Department of Electrical Engineering]
	Al-Hraishawi, Hayder [University of Luxembourg > Interdisciplinary Centre for Security, Reliability and Trust (SNT) > SigCom >]
	Chatzinotas, Symeon [University of Luxembourg > Interdisciplinary Centre for Security, Reliability and Trust (SNT) > SigCom >]
Publication date :	2023
Peer reviewed :	Yes
Audience :	International
Event name :	IEEE Wireless Communications and Networking Conference (WCNC)
Event date :	26-03-2023 to 29-03-2023
Event organizer :	Institute of Electrical and Electronics Engineers
Event place (city) :	Glasgow
Event country :	Scotland, UK
Keywords :	[en] Device-free localization ; deep learning ; data dimensionality reduction ; convolutional deep belief network ; autoencoder ; wireless sensor networks
Abstract :	[en] Location-based services are witnessing a rise in popularity owing to their key features of delivering personalized digital experience. The recent developments in wireless sensing techniques make the realization of device-free localization (DFL) feasible within wireless sensor network (WSN) architectures. The DFL is an emerging technology that utilizes radio signal information for detecting and positioning a passive movable target without attached devices. However, determining the characteristics of the massive raw signals and extracting meaningful discriminative features relevant to the localization are highly intricate tasks due to the different patterns associated with different locations. To overcome these issues, deep learning (DL) techniques can be utilized here owing to their remarkable performance gains in similar practical problems. In this direction, we propose a DFL framework consists of multiple convolutional neural network (CNN) layers along with deep autoencoders based on the restricted Boltzmann machines (RBM) to construct a convolutional deep belief network (CDBN) for features recognition and extracting. Each CNN layer has stochastic pooling to sample down the feature map and reduced the dimensions of the required data without losing important information. This dimensionality reduction can alleviate the heavy computation while ensuring precise localization. The proposed framework is validated using real experimental dataset. The results show that the proposed model is able to achieve a high accuracy of 98% with reduced data dimensions and low signal-to-noise ratios (SNRs).
Target :	Researchers ; Professionals ; Students
Permalink :	http://hdl.handle.net/10993/51597

	File(s) associated to this reference
	Fulltext file(s): File Commentary Version Size Access Open access Convolutional_Deep_Belief_Network.pdf Author preprint 385.8 kB View/Open

All documents in ORBi^lu are protected by a user license.

 

University of Luxembourg Library |  Feedback |  Legal notices Site Map