Decentralizing IP mobility management in future networks

The massive adoption of sophisticated mobile devices and applications led to the increase of mobile data in the last decade, which it is expected to continue. This increase of mobile data negatively impacts the network planning and dimension, since core networks are heavy centralized. Mobile operators are investigating atten network architectures that distribute the responsibility of providing connectivity and mobility, in order to improve the network scalability and performance. Moreover, service providers are moving the content servers closer to the user, in order to ensure high availability and performance of content delivery. Besides the e orts to overcome the explosion of mobile data, current mobility management models are heavy centralized to ensure reachability and session continuity to the users connected to the network. Nowadays, deployed architectures have a small number of centralized mobility anchors managing the mobile data and the mobility context of millions of users, which introduces issues related to performance and scalability that require costly network mechanisms. The mobility management needs to be rethought out-of-the box to cope with atten network architectures and distributed content servers closer to the user, which is the purpose of the work developed in this Thesis. The Thesis starts with a characterization of mobility management into well-de ned functional blocks, their interaction and potential grouping. The decentralized mobility management is studied through analytical models and simulations, in which di erent mobility approaches distinctly distribute the mobility management functionalities through the network. The outcome of this study showed that decentralized mobility management brings advantages. Hence, it was proposed a novel distributed and dynamic mobility management approach, which is exhaustively evaluated through analytical models, simulations and testbed experiments. The proposed approach is also integrated with seamless horizontal handover mechanisms, as well as evaluated in vehicular environments. The mobility mechanisms are also speci ed for multihomed scenarios, in order to provide data o oading with IP mobility from cellular to other access networks. In the pursuing of the optimized mobile routing path, a novel network-based strategy for localized mobility is addressed, in which a replication binding system is deployed in the mobility anchors distributed through the access routers and gateways. Finally, we go further in the mobility anchoring subject, presenting a context-aware adaptive IP mobility anchoring model that dynamically assigns the mobility anchors that provide the optimized routing path to a session, based on the user and network context. The integration of dynamic and distributed concepts in the mobility management, such as context-aware adaptive mobility anchoring and dynamic mobility support, allow the optimization of network resources and the improvement of user experience. The overall outcome demonstrates that decentralized mobility management is a promising direction, hence, its ideas should be taken into account by mobile operators in the deployment of future networks.

Contexts-management strategy in considering the security in urban computing based on urban design

Urban Computing (UrC) provides users with the situation-proper information by considering context of users, devices, and social and physical environment in urban life. With social network services, UrC makes it possible for people with common interests to organize a virtual-society through exchange of context information among them. In these cases, people and personal devices are vulnerable to fake and misleading context information which is transferred from unauthorized and unauthenticated servers by attackers. So called smart devices which run automatically on some context events are more vulnerable if they are not prepared for attacks. In this paper, we illustrate some UrC service scenarios, and show important context information, possible threats, protection method, and secure context management for people.

Dynamic power- and failure-aware cloud resources allocation for Sets of independent tasks

Cloud computing is increasingly being adopted in different scenarios, like social networking, business applications, scientific experiments, etc. Relying in virtualization technology, the construction of these computing environments targets improvements in the infrastructure, such as power-efficiency and fulfillment of users’ SLA specifications. The methodology usually applied is packing all the virtual machines on the proper physical servers. However, failure occurrences in these networked computing systems can induce substantial negative impact on system performance, deviating the system from ours initial objectives. In this work, we propose adapted algorithms to dynamically map virtual machines to physical hosts, in order to improve cloud infrastructure power-efficiency, with low impact on users’ required performance. Our decision making algorithms leverage proactive fault-tolerance techniques to deal with systems failures, allied with virtual machine technology to share nodes resources in an accurately and controlled manner. The results indicate that our algorithms perform better targeting power-efficiency and SLA fulfillment, in face of cloud infrastructure failures.

Critérios de projecto de um data centre

Dissertação de Natureza Científica para obtenção do grau de Mestre em Engenharia Civil na Área de Especialização de Edificações

Capacity sharing and stealing in dynamic server-based real-time systems

This paper proposes a dynamic scheduler that supports the coexistence of guaranteed and non-guaranteed bandwidth servers to efficiently handle soft-tasks’ overloads by making additional capacity available from two sources: (i) residual capacity allocated but unused when jobs complete in less than their budgeted execution time; (ii) stealing capacity from inactive non-isolated servers used to schedule best-effort jobs. The effectiveness of the proposed approach in reducing the mean tardiness of periodic jobs is demonstrated through extensive simulations. The achieved results become even more significant when tasks’ computation times have a large variance.

Estimating Effective Slowdown of Tasks in Energy-Aware Clouds

Consolidation consists in scheduling multiple virtual machines onto fewer servers in order to improve resource utilization and to reduce operational costs due to power consumption. However, virtualization technologies do not offer performance isolation, causing applications’ slowdown. In this work, we propose a performance enforcing mechanism, composed of a slowdown estimator, and a interference- and power-aware scheduling algorithm. The slowdown estimator determines, based on noisy slowdown data samples obtained from state-of-the-art slowdown meters, if tasks will complete within their deadlines, invoking the scheduling algorithm if needed. When invoked, the scheduling algorithm builds performance and power aware virtual clusters to successfully execute the tasks. We conduct simulations injecting synthetic jobs which characteristics follow the last version of the Google Cloud tracelogs. The results indicate that our strategy can be efficiently integrated with state-of-the-art slowdown meters to fulfil contracted SLAs in real-world environments, while reducing operational costs in about 12%.

CPMD-mindful task assignment for NPS-F

The multiprocessor scheduling scheme NPS-F for sporadic tasks has a high utilisation bound and an overall number of preemptions bounded at design time. NPS-F binpacks tasks offline to as many servers as needed. At runtime, the scheduler ensures that each server is mapped to at most one of the m processors, at any instant. When scheduled, servers use EDF to select which of their tasks to run. Yet, unlike the overall number of preemptions, the migrations per se are not tightly bounded. Moreover, we cannot know a priori which task a server will be currently executing at the instant when it migrates. This uncertainty complicates the estimation of cache-related preemption and migration costs (CPMD), potentially resulting in their overestimation. Therefore, to simplify the CPMD estimation, we propose an amended bin-packing scheme for NPS-F allowing us (i) to identify at design time, which task migrates at which instant and (ii) bound a priori the number of migrating tasks, while preserving the utilisation bound of NPS-F.

Análise de Desempenho e Recomendação para Soluções IVR

Mestrado em Engenharia Informática, Área de Especialização em Tecnologias do Conhecimento e da Decisão

Criação e atualização de aplicações para área de retalho por MDA

Nos últimos anos tem-se verificado a constante evolução dos mercados em plataformas na Internet como forma de melhoria não só dos serviços prestados, mas também para o aumento de vendas de produtos e respetiva internacionalização dos mesmos. Este aumento da procura por este tipo de softwares, assim como a constante evolução e atualização dos mesmos tem contribuído para que estas aplicações evoluam em termos de funcionalidades e complexidade. Isto contribui cada vez para a dificuldade de formar equipas capazes de manter e desenvolver estes sistemas sem que comprometa em grandes custos para as organizações. Neste sentido surgiram diversas ferramentas que permitem criar soluções pré desenvolvidas de aplicações na Internet denominadas de "E-commerce applications". Estas plataformas, apesar do utilizador não ter obrigatoriamente que deter conhecimentos de programação para proceder à sua instalação, são bastante restritas tanto aos serviços que podem ser usados, e na sua escalabilidade, visto que normalmente correm em servidores específicos e por vezes as configurações necessárias para instalação tornam-se bastante complexas de ser efetuadas. Pretende-se no âmbito desta dissertação de mestrado propor um modelo de uma arquitetura de um sistema baseado em mecanismos MDA para a área de retalho, particularmente em ambientes de e-commerce. Serão inicialmente sistematizados os principais tipos de ecommerce numa perspetiva de evolução histórica. Será igualmente enquadrado o MDA no desenvolvimento de um sistema de e-commerce. Neste sentido, serão equacionadas as diferenças entre o modelo típico de desenvolvimento de software e o desenvolvimento de software orientado pelas metodologias do MDA. No processo de especificação e desenvolvimento do modelo proposto será realizada uma análise de requisitos, assim como, a proposta do modelo da arquitetura de um sistema baseado em mecanismos MDA, tendo como orientação os requisitos e arquitetura definida na fase de análise. Finalmente no sentido de analisar o resultado esperado para um sistema orientado por metodologias definidas por MDA, serão realizado alguns testes no sistema desenvolvido de forma a analisar o seu desempenho e validar a sua adequabilidade no âmbito do processo de desenvolvimento de sistemas e-commerce

Forcing Usage Rules in PublicWireless LANs

Langattomien lähiverkkojen yleistyessä nopeasti suurten verkkojen teknologiana käyttösääntöjen valvonta tulee tarpeelliseksi. Tässä työssä kuvataan, kuinka käyttäjät voidaan pakottaa noudattamaan käyttösääntöjä julkisissa WLAN-verkoissa. Työssä käsiteltävät ongelmat koskevat menetelmiä epäluotettavien DHCP-palvelinten paljastamiseksi sekä omia IP-osoitteita käyttävien käyttäjien paljastamiseksi tilanteissa, jolloin IP-osoite ei ole virallisen DHCP-palvelimen myöntämä. Jokaisen menetelmän kohdalla pohditaan, kuinka tällaisia käyttäjiä voidaan estää rikkomasta käyttösääntöjä. Lisäksi pohditaan keskitetyn tietojen keruun hyödyntämistä kuvattujen tehtävien suorittamiseksi. Esitetyt ratkaisut on erityisesti suunniteltu testiverkkoa varten, mutta yleiset ideat ovat toimivia missä tahansa langattomassa verkossa.

Particle swarm optimization for two-connected networks with bounded rings

The Two-Connected Network with Bounded Ring (2CNBR) problem is a network design problem addressing the connection of servers to create a survivable network with limited redirections in the event of failures. Particle Swarm Optimization (PSO) is a stochastic population-based optimization technique modeled on the social behaviour of flocking birds or schooling fish. This thesis applies PSO to the 2CNBR problem. As PSO is originally designed to handle a continuous solution space, modification of the algorithm was necessary in order to adapt it for such a highly constrained discrete combinatorial optimization problem. Presented are an indirect transcription scheme for applying PSO to such discrete optimization problems and an oscillating mechanism for averting stagnation.

Dynamics in large scale networks

In this thesis we study the properties of two large dynamic networks, the competition network of advertisers on the Google and Bing search engines and the dynamic network of friend relationships among avatars in the massively multiplayer online game (MMOG) Planetside 2. We are particularly interested in removal patterns in these networks. Our main finding is that in both of these networks the nodes which are most commonly removed are minor near isolated nodes. We also investigate the process of merging of two large networks using data captured during the merger of servers of Planetside 2. We found that the original network structures do not really merge but rather they get gradually replaced by newcomers not associated with the original structures. In the final part of the thesis we investigate the concept of motifs in the Barabási-Albert random graph. We establish some bounds on the number of motifs in this graph.

Le vote par internet à distance : est-il conforme aux exigences du droit électoral québécois et canadien?

"Mémoire présenté à la Faculté des études supérieures en vue de l'obtention du grade de LL.M. en droit option droit des technologies de l'information"

Lutte aux botnets : les politiques de prévention s'avèrent-elles efficaces?

Les alertes que nos logiciels antivirus nous envoient ou encore les différents reportages diffusés dans les médias nous font prendre conscience de l'existence des menaces dans le cyberespace. Que ce soit les pourriels, les attaques par déni de service ou les virus, le cyberespace regorge de menaces qui persistent malgré les efforts qui sont déployés dans leur lutte. Est-ce que cela a à voir avec l'efficacité des politiques en place actuellement pour lutter contre ce phénomène? Pour y répondre, l'objectif général du présent mémoire est de vérifier quelles sont les politiques de prévention (lois anti-pourriel, partenariats publics-privés et démantèlements de botnets) qui influencent le plus fortement le taux de menaces informatiques détectées, tout en s'attardant également à l'effet de différents facteurs socio-économiques sur cette variable. Les données collectées par le logiciel antivirus de la compagnie ESET ont été utilisées. Les résultats suggèrent que les partenariats publics-privés offrant une assistance personnalisée aux internautes s'avèrent être la politique de prévention la plus efficace. Les démantèlements de botnets peuvent également s'avérer efficaces, mais seulement lorsque plusieurs acteurs/serveurs importants du réseau sont mis hors d'état de nuire. Le démantèlement du botnet Mariposa en est un bon exemple. Les résultats de ce mémoire suggèrent que la formule partenariats-démantèlements serait le choix le plus judicieux pour lutter contre les cybermenaces. Ces politiques de prévention possèdent toutes deux des méthodes efficaces pour lutter contre les menaces informatiques et c'est pourquoi elles devraient être mises en commun pour assurer une meilleure défense contre ce phénomène.

Analysis of Some Queueing Models Related To N-Policy: Operations Research

This study is about the analysis of some queueing models related to N-policy.The optimal value the queue size has to attain in order to turn on a single server, assuming that the policy is to turn on a single server when the queue size reaches a certain number, N, and turn him off when the system is empty.The operating policy is the usual N-policy, but with random N and in model 2, a system similar to the one described here.This study analyses “ Tandem queue with two servers”.Here assume that the first server is a specialized one.In a queueing system,under N-policy ,the server will be on vacation until N units accumulate for the first time after becoming idle.A modified version of the N-policy for an M│M│1 queueing system is considered here.The novel feature of this model is that a busy service unit prevents the access of new customers to servers further down the line.It is deals with a queueing model consisting of two servers connected in series with a finite intermediate waiting room of capacity k.Here assume that server I is a specialized server.For this model ,the steady state probability vector and the stability condition are obtained using matrix – geometric method.