Show simple item record

dc.contributor.advisorFapojuwo, Abraham Olatunji
dc.contributor.author., Vatsala
dc.date2022-02
dc.date.accessioned2021-11-23T23:25:45Z
dc.date.available2021-11-23T23:25:45Z
dc.date.issued2021-11
dc.identifier.citationVatsala (2021). Resource allocation for energy harvesting D2D communications underlaying NOMA cellular networks (Unpublished master's thesis). University of Calgary, Calgary, AB.en_US
dc.identifier.urihttp://hdl.handle.net/1880/114139
dc.description.abstractThe fifth generation (5G) cellular networks promise higher data rates, lower latency, higher energy efficiency, and increased bandwidth as compared to the fourth generation (4G) networks. To fulfill requirements raised by 5G networks, notable technologies such as Simultaneous Wireless and Information Power Transfer (SWIPT), device to device (D2D) communications and non-orthogonal multiple access (NOMA) are being extensively researched by the academia and industry. This thesis attempts to fulfill the requirements raised by current users and thus studies these technologies in the form of resource allocation problems for two SWIPT receiver architectures, namely, time switching (TS) and power splitting (PS) enabled D2D communications underlaying a NOMA based network with the objective of maximizing the D2D throughput while the rate requirements of the cellular users are guaranteed. The performance is compared with orthogonal multiple access (OMA) scheme. The problems are solved using two approaches: conventional optimization and deep learning. The conventional optimization entails a large number of iterations and involves significant time to solve the problem. Thus, deep learning is used where neural networks can learn from a dataset provided and used to predict an output. The neural networks involve less computation time and are more efficient. Therefore, a feed forward neural network (FFNN) - a kind of Deep Neural Network (DNN) is used to predict the D2D throughput. It was found that the efficient integration of D2D with the conventional cellular networks depends upon several factors such as environment, density of the network, geographical position of the devices and the rate requirement of the cellular users. Also, deep learning gives almost same results as that of the conventional optimization algorithm but is much more time efficient. In all the scenarios, the NOMA based networks give much better performance than the OMA based networks. The significance of the project lies in adopting D2D communications equipped with TS and PS SWIPT architectures in practical scenarios efficiently by studying the various factors that impact the adoption of D2D communications.en_US
dc.language.isoengen_US
dc.rightsUniversity of Calgary graduate students retain copyright ownership and moral rights for their thesis. You may use this material in any way that is permitted by the Copyright Act or through licensing that has been assigned to the document. For uses that are not allowable under copyright legislation or licensing, you are required to seek permission.en_US
dc.subjectD2D Communicationsen_US
dc.subjectEnergy Harvestingen_US
dc.subjectSWIPTen_US
dc.subjectPower splittingen_US
dc.subjectTime switchingen_US
dc.subject.classificationArtificial Intelligenceen_US
dc.subject.classificationEngineeringen_US
dc.subject.classificationEngineering--Electronics and Electricalen_US
dc.titleResource Allocation for Energy Harvesting D2D Communications Underlaying NOMA Cellular Networksen_US
dc.typemaster thesisen_US
dc.publisher.facultySchulich School of Engineeringen_US
dc.publisher.institutionUniversity of Calgaryen
thesis.degree.nameMaster of Science (MSc)en_US
thesis.degree.disciplineEngineering – Electrical & Computeren_US
thesis.degree.grantorUniversity of Calgaryen_US
dc.contributor.committeememberSesay, Abu-Bakarr B
dc.contributor.committeememberFar, Behrouz H
ucalgary.item.requestcopytrueen_US


Files in this item

Thumbnail
Embargoed until: 2022-02-25

This item appears in the following Collection(s)

Show simple item record

University of Calgary graduate students retain copyright ownership and moral rights for their thesis. You may use this material in any way that is permitted by the Copyright Act or through licensing that has been assigned to the document. For uses that are not allowable under copyright legislation or licensing, you are required to seek permission.