With the wide popularity of smart devices such as smartphones, smartwatches, and smart cameras, Mobile Crowdsensing (MCS) and Crowdsourcing (CS) have been broadly applied for collecting data from a large group of ordinary participants. The quality of participants' contributed data, however, is hard to guarantee, and as such it is critical to develop efficient and effective methods to automatically improve data quality over MCS/CS platforms. In this thesis, we propose three machine learning-based solutions for data quality enhancement in different participatory MCS/CS scenarios. Our solutions aim at the data extraction phase as well as the data collection phase of participatory MCS/CS, including (1) trustworthy information extraction from conflicting data, (2) recognition of learning patterns, and (3) worker recruitment based on interactive training and learning pattern extraction. The first one is designed for the data extraction phase and the other two for the data collection phase. First, to derive reliable data from diverse or even conflicting labels from the crowd, we design a mechanism to infuse knowledge from domain experts into the labels from the crowd to automatically make correct decisions on classification-based MCS tasks. Our solution, named EFusion, utilizes a probabilistic graphical model and the expectation maximization (EM) algorithm to infer the most likely expertise level of each crowd worker, the difficulty level of tasks, and the ground truth answers. Furthermore, we introduce a method to extend EFusion from solving binary classification problems to handling multi-class classification problems. We evaluate EFusion using real-world case studies as well as simulations. Evaluation results demonstrate that EFusion can return more accurate and stable classification results than the majority voting method and state-of-the-art methods. Second, we propose Goldilocks, an interactive learning pattern recognition framework that can identify suitable participants whose performance follows desired learning patterns. To accurately extract a participant's learning pattern, we first estimate the impact of previous training questions on the participant before she answers a new question. After the participant answers each new question, we adjust the estimation of her capability by considering a quantitative measure of the impact of previous questions and her answer to the new question. Based on the extracted learning curve of each participant, we recruit the candidates, who have showed good learning capability and desired learning patterns, for the formal MCS/CS task. We further develop a web service over Amazon Web Services (AWS) that automatically adjusts questions to maximize individual participants' learning performance. This website also profiles the participants' learning patterns, which can be used for task assignment in MCS/CS. Third, we present HybrTraining, a hybrid deep learning framework that captures each candidate’s capability from a long-term perspective and excludes the undesired candidates in the early stage of the training phase. Using two collaborative deep learning networks, HybrTraining can dynamically match participants and MCS/CS tasks. In detail, we build a deep Q-network (DQN) to match the candidates and training batches in the training phase, and develop a long short-term memory (LSTM) model that extracts the learning patterns of different candidates and helps the DQN make better worker-task matching decisions. We build HyberTraining on Compute Canada and evaluate it over two scientific datasets. For each dataset, the learning data of candidates is collected with a Python-based Django website over Amazon Elastic Compute Cloud (Amazon EC2). Evaluation results show that HybrTraining can increase data collection efficiency and improve data quality in MCS/CS. / Graduate / 2022-08-19

Location-based services have experienced substantial growth in the last decade. However, despite extensive research efforts, sub-meter location accuracy with low-cost infrastructure continues to be elusive. Comparing with infrastructure-based solutions, infrastructure-free indoor localization has the major advantage of avoiding extra cost for infrastructure deployment. There are two typical types of infrastructure-free indoor localization solutions, i.e., Pedestrian Dead Reckoning (PDR)-based and fingerprint-based. PDR-based solutions rely on inertial measurement units to estimate the user's relative location. Despite the effort, many issues still remain in PDR systems. For example, any deployed smartphone-based PDR system needs to cope with the changing orientation of smartphone that the phone might be putting in a pocket, or being taken out to use, etc. In addition, the outputs of Micro Electro-Mechanical Systems (MEMS) sensors on smart devices vary over time which results in rapidly accumulated localization errors without external references. Fingerprint-based solutions utilize different types of location dependent parameters to estimate user's absolute location. Although fingerprint-based solutions are usually more practical than PDR-based, they suffer from laborious site survey process. In this dissertation, we aim to mitigate these challenges. First of all, illumination intensity is introduced as a new type of fingerprints to provide location references for PDR-based indoor localization. We propose IDyLL -- an indoor localization system using inertial measurement units (IMU) and photodiode sensors on smartphones. Using a novel illumination peak detection algorithm, IDyLL augments IMU-based pedestrian dead reckoning with location fixes. Moreover, we devise a burned-out detection algorithm for simultaneous luminary-assisted IPS and burned-out luminary detection. Experimental study using data collected from smartphones shows that IDyLL is able to achieve high localization accuracy at low costs. As for fingerprint collection, several frameworks are proposed to ease the laborious site survey process, without compromising fingerprint quality. We propose TuRF, a path-based fingerprint collection mechanism for site survey. MobiBee, a treasure hunt game, is further designed to take advantage of gamification and incentive models for fast fingerprint collection. Motivated by applying mobile crowdsensing for fingerprint collection, we propose ALSense, a distributed active learning framework, for budgeted mobile crowdsensing applications. Novel stream-based active learning strategies are developed to orchestrate queries of annotation data and the upload of unlabeled data from mobile devices. Extensive experiments demonstrate that ALSense can indeed achieve higher classification accuracy given fixed data acquisition budgets. Facing malicious behaviors, three types of location-related attacks and their corresponding detection algorithms are investigated. Experiments on both crowdsensed and emulated dataset show that the proposed algorithms can detect all three types of attacks with high accuracy. / Thesis / Doctor of Philosophy (PhD)

Le mobile crowdsensing consiste à acquérir des données géolocalisées et datées d'une foule de capteurs mobiles (issus de ou connectés à des smartphones). Dans cette thèse, nous nous intéressons au traitement des données issues du mobile crowdsensing environnemental. En particulier, nous proposons de revisiter le problème d'étalonnage aveugle de capteurs comme un problème informé de factorisation matricielle à données manquantes, où les facteurs contiennent respectivement le modèle d'étalonnage fonction du phénomène physique observé (nous proposons des approches pour des modèles affines et non linéaires) et les paramètres d'étalonnage de chaque capteur. Par ailleurs, dans l'application de surveillance de la qualité de l'air que nous considérons, nous supposons avoir à notre disposition des mesures très précises mais distribuées de manière très parcimonieuse dans le temps et l'espace, que nous couplons aux multiples mesures issues de capteurs mobiles. Nos approches sont dites informées car (i) les facteurs matriciels sont structurés par la nature du problème, (ii) le phénomène observé peut être décomposé sous forme parcimonieuse dans un dictionnaire connu ou approché par un modèle physique/géostatistique, et (iii) nous connaissons la fonction d'étalonnage moyenne des capteurs à étalonner. Les approches proposées sont plus performantes que des méthodes basées sur la complétion de la matrice de données observées ou les techniques multi-sauts de la littérature, basées sur des régressions robustes. Enfin, le formalisme informé de factorisation matricielle nous permet aussi de reconstruire une carte fine du phénomène physique observé. / Mobile crowdsensing aims to acquire geolocated and timestamped data from a crowd of sensors (from or connected to smartphones). In this thesis, we focus on processing data from environmental mobile crowdsensing. In particular, we propose to revisit blind sensor calibration as an informed matrix factorization problem with missing entries, where factor matrices respectively contain the calibration model which is a function of the observed physical phenomenon (we focus on approaches for affine or nonlinear sensor responses) and the calibration parameters of each sensor. Moreover, in the considered air quality monitoring application, we assume to pocee- some precise measurements- which are sparsely distributed in space and time - that we melt with the multiple measurements from the mobile sensors. Our approaches are "informed" because (i) factor matrices are structured by the problem nature, (ii) the physical phenomenon can be decomposed using sparse decomposition with a known dictionary or can be approximated by a physical or a geostatistical model, and (iii) we know the mean calibration function of the sensors to be calibrated. The proposed approaches demonstrate better performances than the one based on the completion of the observed data matrix or the multi-hop calibration method from the literature, based on robust regression. Finally, the informed matrix factorization formalism also provides an accurate reconstruction of the observed physical field.

Aujourd’hui, il y a une demande croissante de fournir les informations d'environnement en temps réel tels que la qualité de l'air, le niveau de bruit, état du trafic, etc. pour les citoyens dans les zones urbaines a des fins diverses. La prolifération des capteurs de smartphones et la mobilité de la population font des Mobile Crowdsensing (MCS) un moyen efficace de détecter et de recueillir des informations a un coût faible de déploiement. En MCS, au lieu de déployer capteurs statiques dans les zones urbaines, les utilisateurs avec des périphériques mobiles jouent le rôle des capteurs de mobiles à capturer les informations de leurs environnements, et le réseau de communication (3G, WiFi, etc.) pour le transfert des données pour MCS applications. En général, l'application MCS (ou tâche) non seulement exige que chaque participant de périphérique mobile de posséder la capacité de réception missions de télédétection, de télédétection et de renvoi détecte résultats vers un serveur central, il exige également de recruter des participants, attribuer de télédétection tâches aux participants, et collecter les résultats obtenues par télédétection ainsi que représente les caractéristiques de la cible zone de détection. Afin de recruter un nombre suffisant de participants, l'organisateur d'une MCS tâche devrait considérer la consommation énergétique causée par MCS applications pour chaque participant et les questions de protection dans la vie privée, l'organisateur doit donner a chaque participant un certain montant des incitations comme un encouragement. En outre, afin de recueillir les résultats obtenues par télédétection et représentant la région cible, l'organisateur doit s'assurer que les données de télédétection qualité des résultats obtenues par télédétection, p. ex., la précision et la spatio-temporelle la couverture des résultats obtenus par télédétection. Avec la consommation d'énergie, la protection de la vie privée, les mesures d'incitation, de télédétection et qualité des données à l'esprit, dans cette thèse nous avons étudié quatre problèmes d'optimisation de mobile crowdsensing et mené après quatre travaux de recherche [...] / Nowadays, there is an increasing demand to provide real-time environment information such as air quality, noise level, traffic condition, etc. to citizens in urban areas for various purposes. The proliferation of sensor-equipped smartphones and the mobility of people are making Mobile Crowdsensing (MCS) an effective way to sense and collect information at a low deployment cost. In MCS, instead of deploying static sensors in urban areas, people with mobile devices play the role of mobile sensors to sense the information of their surroundings and the communication network (3G, WiFi, etc.) is used to transfer data for MCS applications. Typically, an MCS application (or task) not only requires each participant's mobile device to possess the capability of receiving sensing tasks, performing sensing and returning sensed results to a central server, it also requires to recruit participants, assign sensing tasks to participants, and collect sensed results that well represents the characteristics of the target sensing region. In order to recruit sufficient participants, the organizer of the MCS task should consider energy consumption caused by MCS applications for each individual participant and the privacy issues, further the organizer should give each participant a certain amount of incentives as encouragement. Further, in order to collect sensed results well representing the target region, the organizer needs to ensure the sensing data quality of the sensed results, e.g., the accuracy and the spatial-temporal coverage of the sensed results. With the energy consumption, privacy, incentives, and sensing data quality in mind, in this thesis we have studied four optimization problems of mobile crowdsensing and conducted following four research works: • EEMC - In this work, the MCS task is splitted into a sequence of sensing cycles, we assume each participant is given an equal amount of incentive for joining in each sensing cycle; further, given the target region of the MCS task, the MCS task aims at collecting an expected number of sensed results from the target region in each sensing cycle.Thus, in order to minimize the total incentive payments and the total energy consumption of the MCS task while meeting the predefined data collection goal, we propose EEMC which intends to select a minimal number of anonymous participants to join in each sensing cycle of the MCS task while ensuring an minimum number of participants returning sensed results. • EMC3 - In this work, we follow the same sensing cycles and incentives assumptions/settings from EEMC; however, given a target region consisting of a set of subareas, the MCS task in this work aims at collecting sensed results covering each subarea of the target region in each sensing cycle (namely full coverage constraint).Thus, in order to minimize the total incentive payments and the total energy consumption of the MCS task under the full coverage constraint, we propose EMC3 which intends to select a minimal number of anonymous participaNts to join in each sensing cycle of the MCS task while ensuring at least one participant returning sensed results from each subarea. • CrowdRecruiter - In this work, we assume each participant is given an equal amount of incentive for joining in all sensing cycles of the MCS task; further, given a target region consisting of a set of subareas, the MCS task aims at collecting sensed results from a predefined percentage of subareas in each sensing cycle (namely probabilistic coverage constraint).Thus, in order to minimize the total incentive payments the probabilistic coverage constraint, we propose CrowdRecruiter which intends to recruit a minimal number of participants for the whole MCS task while ensuring the selected participants returning sensed results from at least a predefined percentage of subareas in each sensing cycle. • CrowdTasker - In this work, we assume each participant is given a varied amount of incentives according to [...]

Smart devices have become one of the fundamental communication and computing devices in people's everyday lives over the past decade. Their various sensors and wireless connectivity have paved the way for a new application area called mobile crowdsensing where sensing services are provided by using the sensor outputs collected from smart devices. A mobile crowdsensing system's service quality heavily depends on the participation of smart device users who probably expect to be compensated in return for their participation. Therefore, mobile crowdsensing applications need incentive mechanisms to motivate such people into participating. In this thesis, we first defined a reverse auction based incentive mechanism for a representative mobile crowdsensing system. Then, we integrated the Vickrey-Clarke- Groves mechanism into the initial incentive mechanism so as to investigate whether truthful bidding would become the dominant strategy in the resulting incentive mechanism. We demonstrated by theoretical analysis that overbidding was the dominant strategy in the base incentive mechanism, whereas truthful bidding was the dominant strategy in the derived incentive mechanism when the VCG mechanism was applicable. Finally, we conducted simulations of both incentive mechanisms in order to measure the fairness of service prices and the fairness of cumulative participant earnings using Jain's fairness index. We observed that both the fairness of service prices and the fairness of cumulative participant earnings were generally better in the derived incentive mechanism when the VCG mechanism was applied. We also found that at least 70% of service requests had fair prices, while between 5% and 85% of participants had fair cumulative earnings in both incentive mechanisms.

L’ubiquité des terminaux intelligents équipés de capteurs a donné naissance à un nouveau paradigme de collecte participative des données appelé Crowdsensing. Pour mener à bien les tâches de collecte, divers défis relatifs à l’implication des participants et des demandeurs de services doivent être relevés. Dans ce contexte, nous abordons quatre questions majeures inhérentes à ce problème: Comment affecter les tâches de collecte afin de maximiser la qualité des données d’une façon éco-énergétique ? Comment minimiser le temps nécessaire à la collecte et au traitement des tâches? Comment inciter les participants à dédier une partie de leurs ressources pour la collecte? et Comment protéger la vie privée des participants tout en préservant la qualité des données reportées ? Tout d’abord, nous nous intéressons au fait que les ressources énergétiques des terminaux mobiles restent limitées. Nous introduisons alors des modèles de déploiement de tâches qui visent à maximiser la qualité des données reportées tout en minimisant le coût énergétique global de la collecte. Ainsi, notre première contribution se matérialise en un modèle d’allocation appelé, QEMSS. QEMSS définit des métriques de qualité de données et cherche à les maximiser en se basant sur des heuristiques utilisant la recherche taboue. De plus, afin de rendre le processus d’allocation résultante plus équitable, nous faisons appel à un deuxième algorithme, F-QEMSS, extension de QEMSS. Les deux solutions ont permis d’obtenir des niveaux de qualité de données compétitifs principalement dans les situations défavorables des zones de faible densité ou de ressources limitées. En outre, afin de minimiser le temps moyen de collecte et de traitement des données, une deuxième phase d’allocation distribuée est ajoutée. Plus précisément, nous proposons dans cette deuxième contribution de désigner des participants responsables de déléguer des tâches. Ces derniers prédisent le comportement d’autres utilisateurs en termes de mobilité et de préférences de collecte. Par conséquent, nous développons deux types d’allocation; MATA qui ne tient compte que de la mobilité et P-MATA qui tient compte à la fois de la mobilité et des préférences des participants. Les deux allocations démontrent que l’estimation des préférences des utilisateurs minimise le temps de collecte et évite le rejet des tâches. La troisième contribution de cette thèse, IP-MATA+, propose des incitations aux participants, ce qui favorise leur engagement aux campagnes de collecte notamment quand le budget dédié est partagé en fonction de la qualité des contributions. Pour finir, nous considérons la problématique de la vie privée des participants au crowdsensing. Particulièrement, nous ciblons la minimisation du risque de divulgation de la vie privée durant la phase du déchargement tout en veillant à l’utilité des données collectées. Ainsi, la quatrième contribution de cette thèse vise à assurer simultanément deux objectifs concurrents, à savoir assurer l’utilité des données nécessaire aux demandeurs et protéger les informations sensibles des participants. Pour ce faire, nous introduisons une entité de confiance dans le système de collecte ayant pour rôle d’exécuter un mécanisme qui génère une version altérée de la donnée collectée qui répond au compromis de protection et d’utilité. La solution développée, appelée PRUM, a été évaluée sur des datasets de collecte participative en variant les scénarios d’attaque et de déchargement des données. Les résultats obtenus prouvent qu’une altération limitée des données collectées peut assurer une protection des informations sensibles des participants tout en préservant environ 98% de l’utilité des données obtenue pour les demandeurs. Pour conclure, nos contributions abordent diverses problématiques complémentaires inhérentes à la collecte participative des données ouvrant la voie à des mises en œuvre réelles et facilitant leur déploiement / The ubiquity of sensors-equipped mobile devices has enabled people to contribute data via crowdsensing systems. This emergent paradigm comes with various applications. However, new challenges arise given users involvement in data collection process. In this context, we introduce collaborative sensing schemes which tackle four main questions: How to assign sensing tasks to maximize data quality with energy-awareness? How to minimize the processing time of sensing tasks? How to motivate users to dedicate part of their resources to the crowdsensing process ? and How to protect participants privacy and not impact data utility when reporting collected sensory data ? First, we focus on the fact that smart devices are energy-constrained and develop task assignment methods that aim to maximize sensor data quality while minimizing the overall energy consumption of the data harvesting process. The resulting contribution materialized as a Quality and Energy-aware Mobile Sensing Scheme (QEMSS) defines first data quality metrics then models and solves the corresponding optimization problem using a Tabu-Search based heuristic. Moreover, we assess the fairness of the resulted scheduling by introducing F-QEMSS variant. Through extensive simulations, we show that both solutions have achieved competitive data quality levels when compared to concurrent methods especially in situations where the process is facing low dense sensing areas and resources shortcomings. As a second contribution, we propose to distribute the assignment process among participants to minimize the average sensing time and processing overload com- pared to a fully centralized approach. Thus, we suggest to designate some participants to carry extra sensing tasks and delegate them to appropriate neighbors. The new assign- ment is based on predicting users local mobility and sensing preferences. Accordingly, we develop two new greedy-based assignment schemes, one only Mobility-aware (MATA) and the other one accounting for both preferences and mobility (P-MATA), and evaluate their performances. Both MATA and P-MATA consider a voluntary sensing process and show that accounting for users preferences minimize the sensing time. Having showing that, our third contribution in this thesis is conceived as an Incentives-based variant, IP-MATA+. IP-MATA+ incorporates rewards in the users choice model and proves their positive impact on enhancing their commitment especially when the dedicated budget is shared function of contributed data quality. Finally, our fourth and last contribution addresses the seizing of users privacy concerns within crowdsensing systems. More specifically, we study the minimization of the incurred privacy leakage in data uploading phase while accounting for the possible quality regression. That is, we assess simultaneously the two competing goals of ensuring queriers required data utility and protecting participants’ sensitive information. Thus, we introduce a trust entity to the crowdsensing traditional system. This entity runs a general privacy-preserving mechanism to release a distorted version of sensed data that responds to a privacy-utility trade-off. The proposed mechanism, called PRUM, is evaluated on three sensing datasets, different adversary models and two main data uploading scenarios. Results show that a limited distortion on collected data may ensure privacy while maintaining about 98% of the required utility level.The four contributions of this thesis tackle competing issues in crowdsensing which paves the way at facilitating its real implementation and aims at broader deployment

[pt] A área das telecomunicações tem presenciado consideráveis avanços tecnológicos dos dispositivos móveis (e. g. telefones inteligentes, relógios inteligentes, tablets, reprodutores de música, entre outros) e da sua crescente popularidade. Dado que esses equipamentos possuem uma série de sensores embutidos como o sistema de posicionamento global, câmera, microfone, bússola, acelerômetro entre outros, e ao mesmo tempo mantêm acesso contínuo às redes de comunicações, eles apresentam uma oportunidade para realizar sensoriamento em grande escala de possíveis eventos do mundo físico e compartilhar os dados obtidos através da internet. Este novo tipo de sensoriamento é conhecido como sensoriamento coletivo móvel ou MCS (Mobile crowd sensing) por sua sigla em inglês, e recentemente tem sido o foco de diversas pesquisas. O maior potencial do MCS é a opção de desenvolver um inúmero de funcionalidades a partir dos próprios recursos internos dos dispositivos móveis e aproveitar seu modelo de mobilidade, baseado no comportamento humano. Em contrapartida existem questões que devem ser consideradas na hora de desenvolver uma rede baseada em MCS. Neste trabalho, apresenta-se a análise realizada sobre os pontos críticos para a criação de uma rede MCS e a partir dos mesmos são apresentadas soluções que permitam criar uma implementação o mais próximo possível da realidade. Foi desenvolvido um modelo de mobilidade para a cidade de Rio de Janeiro baseado na teoria dos grafos, considerando que as atividades diárias das pessoas serão as que definam seu padrão de movimento. A cooptação de usuários é outro dos principais problemas que tem que ser abordado quando se pensa no sensoriamento coletivo móvel. Propõe-se dois modelos de incentivo que consideram e modelam como variáveis as motivações intrínsecas e extrínsecas dos usuários na decisão de participação em uma rede MCS. Consideram também distintos graus de motivação para cada usuário com a finalidade de demonstrar que a resposta dos participantes aos incentivos não é homogênea. O primeiro modelo baseia-se nas respostas consecutivas dos usuários e o segundo baseia-se na teoria de jogos. Em ambos, as decisões tomadas pelos usuários só consideram informações locais ou próprias. Os resultados obtidos permitiram comprovar que os modelos de incentivo propostos conseguem estimar satisfatoriamente o tipo de usuário com o qual está se interagindo e a quantidade de incentivo que deve ser oferecido a cada um deles, além de demonstrar as vantagens de um sistema de incentivo que considera pagamentos variáveis. Também foram analisadas as vantagens de considerar a mobilidade humana neste tipo de abordagem e como a mesma reflete nos modelos de incentivo. / [en] The world of telecommunications has witnessed the growing popularity of mobile devices and its huge technological advancements and innovations (e.g. smartphones, smartwatches, tablets, music players among others). These devices have a series of built-in sensors that measure motion, orientation, and various environmental conditions (e.g. Global Positioning System, camera, microphone, compass, accelerometer, among others). In addition, these devices have continuous network connectivity. So these devices can be seen as a huge opportunity to carry out large-scale sensing of events in the physical world and have the ability of sharing the data obtained through the internet. This new kind of sensor application is known as Mobile crowd sensing (MCS) and it has been a research focus lately. The greatest potential of MCS is found on the versatility that the embedded resources of the mobile devices offer in the development of innumerable functionalities and its mobility model that is based on human behavior. On the other hand, there are issues that must be considered when a MCS-based network is developed. This work presents the analysis performed in order to define issues that are considered critical for the creation and development of an MCS network. From these definitions solutions are proposed that allow to create an implementation as close as possible to reality. A mobility model was developed for the Rio de Janeiro city based on graph theory, and assuming that daily activities of the people will define their movement pattern. Attracting and convincing users is another problem that has to be addressed. Two user incentive models are proposed. Both consider and model the decision of a user to participate in an MCS network based on the intrinsic and extrinsic motivations of the user. The idea is to comprise different levels of motivation for each user in order to demonstrate that the response of the participants to the incentives is not homogeneous. Thus, the first model is based on the consecutive answers of the users and the second model is based on game theory. The results obtained allowed us to prove that the proposed incentive models can satisfactorily estimate the type of user with which we are interacting and the amount of incentive that should be offered to each one of them, besides demonstrating the advantages of an incentive system that considers variable payments. The advantages of considering human mobility in this type of approach and how it affects the incentive models was also analyzed.

Submitted by Cássia Santos (cassia.bcufg@gmail.com) on 2015-10-26T10:10:03Z No. of bitstreams: 2 Dissertação - Paulo César Ferreira Melo - 2014.pdf: 3222791 bytes, checksum: f18cd58c678bb5d11f8bd0cabb32f099 (MD5) license_rdf: 23148 bytes, checksum: 9da0b6dfac957114c6a7714714b86306 (MD5) / Approved for entry into archive by Luciana Ferreira (lucgeral@gmail.com) on 2015-10-26T13:03:05Z (GMT) No. of bitstreams: 2 Dissertação - Paulo César Ferreira Melo - 2014.pdf: 3222791 bytes, checksum: f18cd58c678bb5d11f8bd0cabb32f099 (MD5) license_rdf: 23148 bytes, checksum: 9da0b6dfac957114c6a7714714b86306 (MD5) / Made available in DSpace on 2015-10-26T13:03:05Z (GMT). No. of bitstreams: 2 Dissertação - Paulo César Ferreira Melo - 2014.pdf: 3222791 bytes, checksum: f18cd58c678bb5d11f8bd0cabb32f099 (MD5) license_rdf: 23148 bytes, checksum: 9da0b6dfac957114c6a7714714b86306 (MD5) Previous issue date: 2014-10-15 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES / Recent advances in ubiquitous computing have contributed to the rise of an emerging category of mobile devices that have computational and sensing capabilities, such as smartphones and wearable devices. The widespread use of these devices connected by communication networks contribute to the evolution of the Internet of Things. The presence of these mobile devices increases the chance for the development of applications using the sensing ability of these devices to measure, and understand the environmental indicators. Furthemore, data sensed by these applications can be shared among different mobile devices, giving rise to a paradigm called mobile crowdsensing. The complexity of applications in this domain is associated with factors such as interoperability between different mobile devices, data identification and capture from these devices, and adaptation of their use in heterogeneous and dynamic environments. Software engineering approaches such as Model-Driven Engineering (MDE) and, more specifically, models at runtime are an effective way of dealing with this complexity. We propose the use of an approach based on models at runtime for creating and processing mobile crowdsensing queries.We show how this approach can be used by defining a domain-specific modeling language for the mobile crowdsensing domain, called CSML. We built and validated the CSML metamodel which captures the main aspects of the domain, and its execution environment, which consists of an execution engine for models described in CSML, called CSVM. This approach facilitates the specification of mobile crowdsensing queries, also enabling their dynamic change during their processing. / Recentes avanços na computação ubíqua colaboraram para a ascensão de uma categoria emergente de dispositivos móveis que apresentam capacidades computacionais e de sensoriamento, tais como smartphones e dispositivos vestíveis. A proliferação desses dispositivos e sua conexão por meio de redes de comunicação contribui para a evolução da Internet das Coisas. A presença desses dispositivos móveis aumenta a oportunidade para o desenvolvimento de aplicações que utilizam sua capacidade de sensoriamento a fim de medir, inferir e entender os indicadores do ambiente. Por sua vez, os dados sensoriados por essas aplicações podem ser compartilhados entre diferentes dispositivos móveis, dando origem ao paradigma denominado CrowdSensing móvel. A complexidade de aplicações pertencentes ao domínio de CrowdSensing móvel está associada a fatores como interoperabilidade entre diferentes dispositivos móveis, identificação e captação de dados provenientes desses dispositivos e adaptação de seu uso em ambientes heterogêneos e dinâmicos. Abordagens baseadas na Engenharia Dirigida por Modelos (MDE), como modelos em tempo de execução constituem uma forma de lidar com complexidade desse domínio de aplicações. Neste trabalho propomos o uso de uma abordagem dirigida por modelos em tempo de execução para criação e processamento de consultas de crowdsensing móvel que são um importante elemento de aplicações de crowdsensing. Mostramos como essa abordagem pode ser empregada por meio da definição de uma linguagem de modelagem específica para o domínio de crowdsensing móvel, denominada CSML. Neste sentido, construímos e validamos o metamodelo da CSML, que captura os principais aspectos do domínio e seu ambiente de execução, que consiste em uma máquina de execução de modelos descritos em CSML, denominada CSVM . Essa abordagem dirigida por modelos facilita a especificação de consultas de crowdsensing móvel, além de possibilitar a alteração dinâmica dessas consultas durante seu processamento.

Participatory Sensing Systems (PSS) are a type of Mobile Crowdsensing System where users voluntarily participate in contributing information. Task initiators create tasks, targeting specific data that needs to be gathered by the users’ device sensors. Such systems have been designed with different requirements, such as data trustworthiness, accountability, and incentives, in a secure and private way. However, it is complex to protect user privacy without affecting the performance of the rest of the system. For example, with task assignment, either the user authenticates anonymously, or discloses its sensors for an efficient allocation. If the user identity is hidden from the system, it could receive a task it cannot perform. This thesis aims to design an anonymous authentication model for PSS based on privacy-preserving attribute-based signatures. The proposed solution allows the Participatory Sensing System to enforce sensor requirements for efficient task allocation. In addition to the design, experiments measuring the performance of the operations are included in the thesis, to prove it is suitable for real-world scenarios. / Participatory Sensing Systems (PSS) är en typ av mobilt Crowdsensing-system där användare frivilligt deltar i att bidra med information. Aktivitetsinitiatorer skapar uppgifter, inriktade på specifik data som behöver samlas in av användarnas enhetssensorer. Sådana system har utformats med olika krav, såsom datatillförlitlighet, ansvarighet och incitament, på ett säkert och privat sätt. Det är dock komplicerat att skydda användarnas integritet utan att påverka prestandan för resten av systemet. Till exempel, med uppgiftstilldelning, antingen autentiserar användaren anonymt eller avslöjar sina sensorer för en effektiv tilldelning. Om användaridentiteten är dold från systemet kan den få en uppgift som den inte kan utföra. Denna avhandling syftar till att designa en anonym autentiseringsmodell för PSS baserad på integritetsbevarande attributbaserade signaturer. Den föreslagna lösningen gör det möjligt för Participatory Sensing System att upprätthålla sensorkrav för effektiv uppgiftsallokering. Utöver designen ingår experiment som mäter verksamhetens prestanda i avhandlingen, för att bevisa att den är lämplig för verkliga scenarier.

La collecte participative des données mobiles est un nouveau paradigme dédié aux applications de détection urbaines utilisant une foule de participants munis de téléphones intelligents. Pour mener à bien les tâches de collecte participative des données mobiles, diverses préoccupations relatives aux participants et aux organisateurs doivent être soigneusement prises en considération. Pour les participants, la principale préoccupation porte sur la consommation d'énergie, le coût des données mobiles, etc. Pour les organisateurs, la qualité des données et le budget sont les deux préoccupations essentielles. Dans cette thèse, deux mécanismes de collecte participative des données mobiles sont proposés : le téléchargement montant collaboratif des données et la collecte clairsemée des données mobiles. Pour le téléchargement montant collaboratif des données, deux procédés sont proposés 1) « effSense », qui fournit la meilleure solution permettant d’économiser la consommation d'énergie aux participants ayant un débit suffisant, et de réduire le coût des communications mobiles aux participants ayant un débit limité; 2) « ecoSense », qui permet de réduire le remboursement incitatif par les organisateurs des frais associés au coût des données mobiles des participants. Dans la collecte clairsemée des données mobiles, les corrélations spatiales et temporelles entre les données détectées sont exploitées pour réduire de manière significative le nombre de tâches allouées et, par conséquent, le budget associé aux organisateurs, tout en assurant la qualité des données. De plus, l’intimité différentielle est afin de répondre au besoin de préservation de la localisation des participants / Mobile crowdsensing is a novel paradigm for urban sensing applications using a crowd of participants' sensor-equipped smartphones. To successfully complete mobile crowdsensing tasks, various concerns of participants and organizers need to be carefully considered. For participants, primary concerns include energy consumption, mobile data cost, privacy, etc. For organizers, data quality and budget are two critical concerns. In this dissertation, to address both participants' and organizers' concerns, two mobile crowdsensing mechanisms are proposed - collaborative data uploading and sparse mobile crowdsensing. In collaborative data uploading, participants help each other through opportunistic encounters and data relays in the data uploading process of crowdsensing, in order to save energy consumption, mobile data cost, etc. Specifically, two collaborative data uploading procedures are proposed (1) effSense, which helps participants with enough data plan to save energy consumption, and participants with little data plan to save mobile data cost; (2) ecoSense, which reduces organizers' incentive refund that is paid for covering participants' mobile data cost. In sparse mobile crowdsensing, spatial and temporal correlations among sensed data are leveraged to significantly reduce the number of allocated tasks thus organizers' budget, still ensuring data quality. Specifically, a sparse crowdsensing task allocation framework, CCS-TA, is implemented with compressive sensing, active learning, and Bayesian inference techniques. Furthermore, differential privacy is introduced into sparse mobile crowdsensing to address participants' location privacy concerns