Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.703554
Title: Spatial and social paradigms for coverage analysis in device to device networks
Author: Mustafa, Hafiz Atta Ul
ISNI:       0000 0004 6062 2100
Awarding Body: University of Surrey
Current Institution: University of Surrey
Date of Award: 2017
Availability of Full Text:
Access through EThOS:
Access through Institution:
Abstract:
The traditional approaches towards capacity gain, ubiquitous coverage and energy efficient green cellular communication call for network densification using small cell deployments. This network densification has been a successful strategy; however network operational/capital expenditure (OPEX/CAPEX) and energy efficiency concerns, due to large number of deployments, cannot be ignored. The network controlled underlay device-to-device (D2D) densification is one alternate solution that can offer lower OPEX/CAPEX, huge capacity, improved energy efficiency, increased area spectral efficiency, ubiquitous coverage, and very low end-to-end latency. However, this coexistence poses challenging interference management due to intra-cell cross-tier interference. In order to analyse the effect of intra-cell cross-tier interference, I have provided Stochastic framework for coverage analysis of primary cellular user. The basic model considers homogeneous Poisson point process (PPP) as spatial distribution of mobile users in the coverage area. This model has been extended to capture inhomogeneity and clustering effect. Therefore, I have used Permanental Cox process, a subclass of doubly Stochastic Poisson process where inhomogeneity is dictated by random intensity measure and clustering effect is governed by homophilic relation that exists due to spatial correlation between mobile users. The D2D pairing is realized by nearest neighbour distribution function. In this context, I have proposed Euler Characteristic based approach to approximate intractable random intensity measure and subsequently derive nearest neighbour distribution function. I have proposed the threshold and spatial extent of excursion set of chi-squared random field as interference control parameters to select different cluster sizes for D2D communication. Further, I have considered both spatial and social layers for D2D pairing. In social domain, I consider common content requests and model the joint spatial and social distribution as proximity based independently marked homogeneous PPP. The proximity considers physical distance between D2D nodes whereas social relationship is modelled as Zipf based marks. I apply these two paradigms to analyse the effect of intra-cell cross-tier interference on average coverage probability of distance-proportional power-controlled primary cellular user.
Supervisor: Imran, Muhammad A. ; Rahim, Tafazolli Sponsor: University of Surrey
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.703554  DOI: Not available
Share: