Projeto INFRAS

Internet do Futuro: Nuvens, Mobilidade e Segurança

  • Increase font size
  • Default font size
  • Decrease font size
Página Inicial

Projeto INFRAS

Internet do Futuro: Nuvens, Mobilidade e Segurança

INFRAS Project

Future Internet: Clouds, Mobility, and Security


Resumo

O Projeto INFRAS: Internet do Futuro: Nuvens, Mobilidade e Segurança, aborda três aspectos fundamentais para a Internet do Futuro: segurança, mobilidade e virtualização. Esses temas são fundamentalmente correlacionados, e por isso mesmo, a abordagem isolada não representa a melhor solução. O Projeto INFRAS aborda todos os três temas levando em consideração os requisitos das aplicações atuais e as vislumbradas para a Internet do Futuro.
Em segurança, os defensoresdo paradigma fim-a-fim defendiam que os problemas de segurança deveriam ser tratados pelas estações-finais. No entanto, o enorme crescimento dos ataques distribuídos de negação de serviço indicou que pelo menos alguns mecanismos de segurança devem ser providos pelo núcleo da rede e, portanto, fazer parte da arquitetura da rede.
A característica de mobilidade é hoje um requisito fundamental e as expectativas apontam para um crescente acesso de dispositivos sem fio e, assim, a forma de comunicação estabelecida no projeto original da Internet, baseada no princípio do fim-a-fim, com conexão ponto-a-ponto e entrega imediata, já não atende satisfatoriamente os requisitos atuais da rede e, portanto, deve ser revista.
O uso das técnicas de virtualização abre novos horizontes e atende a diversos requisitos no desenvolvimento de uma nova arquitetura para a Internet do Futuro. A virtualização é fundamental na computação em nuvens já que ela possibilita o deslocamento de recursos de rede e computacionais conforme a demanda.  Na computação em nuvem os usuários solicitam um serviço da Internet, mas não tem influência sobre onde e como esse serviço será prestado. A resposta à solicitação é providenciada de acordo com a melhor disponibilidade de recursos encontrada na nuvem. O gerenciamento de tais recursos bem como os algoritmos utilizados são temas de alta relevância para a Internet do Futuro.  O desafio é manter a generalidade e a capacidade de evolução da rede e, ao mesmo tempo, minimizar os custos de implantação.

Abstract

The Internet model has two main architectural pillars: the end-to-end principle for data transfers and the dependence on the TCP/IP protocol stack. Such a model was a catalyst for the adoption and growth of the Internet, as it does not require changing the core of the network when new applications are created. The simplicity of this model, however, also resulted in the “ossification” of the Internet, making it hard to solve structural problems that had not been thought of at the time of its design.
The current Internet does not succeed anymore in supporting today’s requirements in a satisfactory way. Its evolution has been achieved through patches and artifacts. It is now clear that a “new Internet” is required so that today’s and future requirements be fully satisfied. The challenges are however manifold. In particular, proposed solutions must be evaluated, simulated, and tested at a scale that is compatible with the current Internet. Several worldwide projects focus on the evaluation of solutions in real testbeds. The main initiatives on the design of the future Internet advocate that the new architecture must include features able to change and evolve when new network technologies appear. The INFRAS project proposes to develop new techniques to address problems related to three of the most fundamental challenges before the future Internet becomes a reality: security, mobility, and resource virtualization.

 

  • Security. Threats like virus dissemination, Trojan horses, deny-of-service (DoS) attacks, and spams are security problems that proliferate in part due to the fact that the original Internet architecture did not predict mechanisms to limit the operation of malicious users. When the first attacks were observed in the Internet, the supporters of the end-to-end paradigm stated that security problems should be considered only at the end points. Nevertheless, the huge growth of distributed deny-of-service attacks (DDoS) clearly indicated that at least part of the security mechanisms must be provided at the network core. Hence, security must be part of the network architecture to guarantee an identification and accountability system without compromising user anonymity. Furthermore, trustiness mechanisms must be provided to raise the trust and availability of services offered by the network both for data and control.
  • Mobility. It is a fact today that wireless devices are becoming more and more ubiquitous as users are mobile by nature. In fact, the availability of Internet access on mobile phones, together with the increase of social network applications, makes us believe that the number of mobile devices connected to the Internet will overcome the number of fixed devices in a near future. In this way, the end-to-end principle originally included in the Internet architecture, which relies on point-to-point communications and immediate delivery, is no further adapted and must be revisited.
  • Virtualization. The use of virtualization techniques open new horizons and has been considered as an integrating part of the design of the future Internet. One of the possibilities is the paradigm of cloud computing, where users are served with computation resources but do not know where and how the service is provided. The response answer to the user request is provided in accordance with the best resource availability at the entrance of the cloud. Virtualization strongly contributes to cloud computing as it is directly related to the displacement of computational resources as a response to the demand. The management of such resources as well as the underlying algorithms are of major importance to the Internet of the future.

 


Informações do Projeto

Processo: 490694/2010-1 
Modalidade: Auxílio a Pesquisa-APQ
Edital: CNPq N.º 053/2010
Cooperação Internacional Brasil-França CNPq/CNRS
Custeio: R$ 76.000,00