RT Dissertation/Thesis
T1 Infrastructure-less D2D Communications through Opportunistic Networks
A1 Pérez Palma, Noelia María
A2 IMDEA Networks Institute, 
AB In recent years, we have experienced several social media blackouts, which haveshown how much our daily experiences depend on high-quality communication services.Blackouts have occurred because of technical problems, natural disasters, hacker attacksor even due to deliberate censorship actions undertaken by governments. In all cases,the spontaneous reaction of people consisted in finding alternative channels and media soas to reach out to their contacts and partake their experiences. Thus, it has clearlyemerged that infrastructured networks—and cellular networks in particular—are wellengineered and have been extremely successful so far, although other paradigms shouldbe explored to connect people. The most promising of today’s alternative paradigmsis Device-to-Device (D2D) because it allows for building networks almost freely, andbecause 5G standards are (for the first time) seriously addressing the possibility of usingD2D communications.In this dissertation I look at opportunistic D2D networking, possibly operating in aninfrastructure-less environment, and I investigate several schemes through modeling andsimulation, deriving metrics that characterize their performance. In particular, I considervariations of the Floating Content (FC) paradigm, that was previously proposed in thetechnical literature.Using FC, it is possible to probabilistically store information over a given restrictedlocal area of interest, by opportunistically spreading it to mobile users while in the area.In more detail, a piece of information which is injected in the area by delivering it to oneor more of the mobile users, is opportunistically exchanged among mobile users wheneverthey come in proximity of one another, progressively reaching most (ideally all) users inthe area and thus making the information dwell in the area of interest, like in a sort ofdistributed storage.While previous works on FC almost exclusively concentrated on the communicationcomponent, in this dissertation I look at the storage and computing components of FC,as well as its capability of transferring information from one area of interest to another.I first present background work, including a brief review of my Master Thesis activity,devoted to the design, implementation and validation of a smartphone opportunisticinformation sharing application. The goal of the app was to collect experimental data that permitted a detailed analysis of the occurring events, and a careful assessment ofthe performance of opportunistic information sharing services. Through experiments, Ishowed that many key assumptions commonly adopted in analytical and simulation worksdo not hold with current technologies. I also showed that the high density of devices andthe enforcement of long transmission ranges for links at the edge might counter-intuitivelyimpair performance.The insight obtained during my Master Thesis work was extremely useful to devisesmart operating procedures for the opportunistic D2D communications considered in thisdissertation. In the core of this dissertation, initially I propose and study a set of schemesto explore and combine different information dissemination paradigms along with realusers mobility and predictions focused on the smart diffusion of content over disjointareas of interest. To analyze the viability of such schemes, I have implemented a Pythonsimulator to evaluate the average availability and lifetime of a piece of information, aswell as storage usage and network utilization metrics. Comparing the performance ofthese predictive schemes with state-of-the-art approaches, results demonstrate the needfor smart usage of communication opportunities and storage. The proposed algorithmsallow for an important reduction in network activity by decreasing the number of dataexchanges by up to 92%, requiring the use of up to 50% less of on-device storage,while guaranteeing the dissemination of information with performance similar to legacyepidemic dissemination protocols.In a second step, I have worked on the analysis of the storage capacity of probabilisticdistributed storage systems, developing a simple yet powerful information theoreticalanalysis based on a mean field model of opportunistic information exchange. I havealso extended the previous simulator to compare the numerical results generated by theanalytical model to the predictions of realistic simulations under different setups, showingin this way the accuracy of the analytical approach, and characterizing the properties ofthe system storage capacity.I conclude from analysis and simulated results that when the density of contents seededin a floating system is larger than the maximum amount which can be sustained by thesystem in steady state, the mean content availability decreases, and the stored informationsaturates due to the effects of resource contention. With the presence of static nodes, ina system with infinite host memory and at the mean field limit, there is no upper boundto the amount of injected contents which a floating system can sustain. However, as withno static nodes, by increasing the injected information, the amount of stored informationeventually reaches a saturation value which corresponds to the injected information atwhich the mean amount of time spent exchanging content during a contact is equal tothe mean duration of a contact.As a final step of my dissertation, I have also explored by simulation the computingand learning capabilities of an infrastructure-less opportunistic communication, storage and computing system, considering an environment that hosts a distributed MachineLearning (ML) paradigm that uses observations collected in the area over which the FCsystem operates to infer properties of the area. Results show that the ML system canoperate in two regimes, depending on the load of the FC scheme. At low FC load, the MLsystem in each node operates on observations collected by all users and opportunisticallyshared among nodes. At high FC load, especially when the data to be opportunisticallyexchanged becomes too large to be transmitted during the average contact time betweennodes, the ML system can only exploit the observations endogenous to each user, whichare much less numerous. As a result, I conclude that such setups are adequate to supportgeneral instances of distributed ML algorithms with continuous learning, only under thecondition of low to medium loads of the FC system. While the load of the FC systeminduces a sort of phase transition on the ML system performance, the effect of computingload is more progressive. When the computing capacity is not sufficient to train allobservations, some will be skipped, and performance progressively declines.In summary, with respect to traditional studies of the FC opportunistic informationdiffusion paradigm, which only look at the communication component over one area ofinterest, I have considered three types of extensions by looking at the performance of FC:over several disjoint areas of interest;in terms of information storage capacity;in terms of computing capacity that supports distributed learning.The three topics are treated respectively in Chapters 3 to 5.
YR 2021
FD 2021-11
LK https://hdl.handle.net/10016/34879
UL https://hdl.handle.net/10016/34879
LA eng
NO Mención Internacional en el título de doctor
NO This work has been supported by IMDEA Networks Institute
DS e-Archivo
RD 31 may. 2024