<?xml version="1.0" encoding="UTF-8"?><?xml-stylesheet type="text/xsl" href="static/style.xsl"?><OAI-PMH xmlns="http://www.openarchives.org/OAI/2.0/" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://www.openarchives.org/OAI/2.0/ http://www.openarchives.org/OAI/2.0/OAI-PMH.xsd"><responseDate>2026-11-22T18:13:01Z</responseDate><request verb="GetRecord" identifier="oai:uwspace.uwaterloo.ca:10012/10253" metadataPrefix="dim">https://dspacemainprd01.lib.uwaterloo.ca/server/oai/request</request><GetRecord><record><header><identifier>oai:uwspace.uwaterloo.ca:10012/10253</identifier><datestamp>2025-03-26T23:38:41Z</datestamp><setSpec>com_10012_9900</setSpec><setSpec>com_10012_1</setSpec><setSpec>col_10012_9908</setSpec><setSpec>col_10012_6</setSpec></header><metadata><dim:dim xmlns:dim="http://www.dspace.org/xmlns/dspace/dim" xmlns:doc="http://www.lyncode.com/xoai" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://www.dspace.org/xmlns/dspace/dim http://www.dspace.org/schema/dim.xsd">
   <dim:field mdschema="dc" element="contributor" qualifier="advisor">Mazumdar, Ravi R.</dim:field>
   <dim:field mdschema="dc" element="contributor" qualifier="author" authority="70de3f7f-c08d-4215-9727-318a7e453cb6" confidence="600">Mukhopadhyay, Arpan</dim:field>
   <dim:field mdschema="dc" element="date" qualifier="accessioned">2016-02-10T18:45:29Z</dim:field>
   <dim:field mdschema="dc" element="date" qualifier="available">2016-02-10T18:45:29Z</dim:field>
   <dim:field mdschema="dc" element="date" qualifier="issued">2016-02-10</dim:field>
   <dim:field mdschema="dc" element="date" qualifier="submitted">2016-02-08</dim:field>
   <dim:field mdschema="dc" element="identifier" qualifier="uri">http://hdl.handle.net/10012/10253</dim:field>
   <dim:field mdschema="dc" element="description" qualifier="abstract" lang="en">In the context of complex networks, we often encounter systems&#xd;
in which the constituent entities randomly interact with each other as they&#xd;
evolve with time. Such random interactions can be described by Markov processes,&#xd;
constructed on suitable state spaces. For many practical systems (e.g. server farms, &#xd;
cloud data centers, social networks), the&#xd;
Markov processes, describing the time-evolution of their constituent entities,&#xd;
become analytically intractable&#xd;
as a result of the complex interdependence&#xd;
among the interacting entities. &#xd;
However, if the `strength&amp;apos; of&#xd;
these interactions converges to a constant as the size of the system&#xd;
is increased, then&#xd;
in the large system limit the underlying Markov process converges to a deterministic process, known&#xd;
as the mean field limit of the corresponding&#xd;
system. Thus, the mean field limit &#xd;
provides a deterministic approximation of the randomly evolving system. &#xd;
Such approximations are accurate for large system sizes.&#xd;
&#xd;
Most prior works on mean field &#xd;
techniques have analyzed systems in which the constituent entities       &#xd;
are identical or homogeneous.&#xd;
In this dissertation, we use mean field techniques to analyze large complex systems&#xd;
composed of heterogeneous entities. &#xd;
&#xd;
First, we consider a class of large multi-server systems, that&#xd;
arise in the context of web-server farms and cloud data centers. In such systems, &#xd;
servers with heterogeneous capacities work in parallel to process&#xd;
incoming jobs or requests. &#xd;
We study schemes to assign the incoming jobs to the servers &#xd;
with the goal of achieving optimal performance in terms of &#xd;
certain metrics of interest while requiring the state information&#xd;
of only a small number of servers in the system. &#xd;
To this end, we consider  randomized dynamic job assignment schemes&#xd;
which sample a small random subset of servers at every job arrival instant&#xd;
and assign the incoming job to one of the sampled servers based &#xd;
on their instantaneous states. &#xd;
We show that for heterogeneous systems,&#xd;
naive sampling of the servers may result in an `unstable&amp;apos; system. &#xd;
We propose schemes which maintain stability&#xd;
by suitably sampling the servers. &#xd;
The performances of these schemes are studied via the corresponding mean field limits,&#xd;
that are shown to exist.&#xd;
The existence and uniqueness of an asymptotically stable&#xd;
equilibrium point of the mean field is established in every case.&#xd;
Furthermore, it is shown that, in the large system limit,&#xd;
the servers become independent of each other and the stationary&#xd;
distribution of occupancy of each server can be obtained from the unique&#xd;
equilibrium point of the mean field. The stationary tail distribution&#xd;
of server occupancies is shown to have a fast decay rate which suggests significantly&#xd;
improved performance for the appropriate metrics relevant to the application. Numerical studies&#xd;
are presented which show that the proposed randomized dynamic schemes significantly outperform&#xd;
randomized static schemes where job assignments are made independently&#xd;
of the server states. In certain scenarios, the randomized dynamic schemes are observed &#xd;
to be nearly optimal in terms of their performances.&#xd;
&#xd;
Next, using mean field techniques, we study &#xd;
a different class of models&#xd;
that arise in the context of social networks.  &#xd;
More specifically, we study the impact of social interactions &#xd;
on the dynamics of opinion formation in a social network&#xd;
consisting of a large number of interacting social agents.&#xd;
The agents are assumed to be mobile and hence do not have any fixed set of&#xd;
neighbors.&#xd;
Opinion of each agent is treated as a binary random variable,&#xd;
taking values in the set {0,1}. This represents scenarios,&#xd;
where the agents have to choose from two available options. &#xd;
The interactions between the agents are modeled using&#xd;
1) the `voter&amp;apos; rule and 2) the `majority&amp;apos; based rule.&#xd;
Under each rule, we consider two scenarios, (1) where the agents &#xd;
are biased towards a specific opinion and&#xd;
(2) where the agents have different propensities to change&#xd;
their past opinions. &#xd;
For each of these scenarios, we characterize the &#xd;
equilibrium distribution of opinions in the network&#xd;
and the convergence rate to the equilibrium &#xd;
by analyzing the corresponding mean field limit. &#xd;
Our results show that the presence of biased agents can significantly&#xd;
reduce the  rate of convergence to the equilibrium. It is also observed that,&#xd;
under the dynamics of the majority rule, the presence of `stubborn&amp;apos; agents &#xd;
(those who do not update their opinions)&#xd;
may result in a metastable network,&#xd;
where the opinion distribution of the non-stubborn agents&#xd;
fluctuates among multiple stable configurations.</dim:field>
   <dim:field mdschema="dc" element="language" qualifier="iso" lang="en">en</dim:field>
   <dim:field mdschema="dc" element="publisher" lang="en">University of Waterloo</dim:field>
   <dim:field mdschema="dc" element="subject" lang="en">Mean field</dim:field>
   <dim:field mdschema="dc" element="subject" lang="en">Heterogeneous</dim:field>
   <dim:field mdschema="dc" element="subject" lang="en">Processor sharing</dim:field>
   <dim:field mdschema="dc" element="subject" lang="en">Loss systems</dim:field>
   <dim:field mdschema="dc" element="subject" lang="en">Opinion dynamics</dim:field>
   <dim:field mdschema="dc" element="subject" lang="en">Majority rule</dim:field>
   <dim:field mdschema="dc" element="title" lang="en">Mean Field Interactions in Heterogeneous Networks</dim:field>
   <dim:field mdschema="dc" element="type" lang="en">Doctoral Thesis</dim:field>
   <dim:field mdschema="dc" element="pending">false</dim:field>
   <dim:field mdschema="uws-etd" element="degree" qualifier="department" lang="en">Electrical and Computer Engineering</dim:field>
   <dim:field mdschema="uws-etd" element="degree" qualifier="discipline" lang="en">Electrical and Computer Engineering</dim:field>
   <dim:field mdschema="uws-etd" element="degree" qualifier="grantor" lang="en">University of Waterloo</dim:field>
   <dim:field mdschema="uws-etd" element="degree" lang="en">Doctor of Philosophy</dim:field>
   <dim:field mdschema="uws" element="comment" qualifier="hidden" lang="en">This is an updated version of the thesis with all the editorial changes incorporated.</dim:field>
   <dim:field mdschema="uws" element="contributor" qualifier="advisor">Mazumdar, Ravi R.</dim:field>
   <dim:field mdschema="uws" element="contributor" qualifier="affiliation1" lang="en">Faculty of Engineering</dim:field>
   <dim:field mdschema="uws" element="published" qualifier="city" lang="en">Waterloo</dim:field>
   <dim:field mdschema="uws" element="published" qualifier="country" lang="en">Canada</dim:field>
   <dim:field mdschema="uws" element="published" qualifier="province" lang="en">Ontario</dim:field>
   <dim:field mdschema="uws" element="typeOfResource" lang="en">Text</dim:field>
   <dim:field mdschema="uws" element="peerReviewStatus" lang="en">Unreviewed</dim:field>
   <dim:field mdschema="uws" element="scholarLevel" lang="en">Graduate</dim:field>
   <dim:field mdschema="others" element="access-status">open.access</dim:field>
</dim:dim>
</metadata></record></GetRecord></OAI-PMH>