Accepted Short Courses

SC1 - Big Data in the Cloud: Challenges and Techniques for Innovation

  • Abstract:
  • The amount of data in the world is rapidly increasing and can no longer be measured in giga or terabytes, but in peta or even in exabytes. Users have become data sources; companies store uncountable information from their clients, suppliers and commercial transactions; millions of sensors monitor the real world; cell phones, electronic power meters, portable devices, and vehicles sensor, create, analyze, and remotely exchange data in the Internet of things. Large amounts of data (big data) are now captured, transferred, aggregated, stored and analyzed. Thus, the innovation, the competition, and the productivity will depend on data processing, since the operations on them require efficiency, speed, and reliability. Applications of huge success in the Internet, such as Yahoo, Amazon, and Facebook, use distributed datacenters where data are replicated to reduce latency and to improve users’ experience. The maintenance of coherent copies of data involves the transmission of large amounts of data in the network. In addition to concerns regarding the provided service, the storage and the transmission of data masses create other challenges. The operation of communication protocols is not always reasonable with typical datacenter applications such as parallel computing. Moreover, the manipulation of data masses incurs in an increasing energy and financial costs. Where to store, when to transfer, and how to move data become primarily questions under an ecological viewpoint as well as economical. Besides the new strategies in database, data mining, parallel computing, and data visualization, new communication protocols and virtualization techniques are key components to support big data in the Internet. In this short-course, problems regarding big data applications and the analysis of big databases are addressed. In addition, issues concerning communication protocols, in both inter- and intra-datacenter environments, proposals and technological solutions, as well as the main initiatives in this research area will be presented.

  • Authors:
  • - Luis Henrique Costa (UFRJ)
    - Marcelo Dias de Amorim (LIP6/CNRS - UPMC)
    - Miguel Elias Mitre Campista (UFRJ)
    - Marcelo Gonçalves Rubinstein (UERJ)
    - Patricia Florissi (EMC Corporation)
    - Otto Carlos Muniz Bandeira Duarte (UFRJ)

SC 2 - Virtual Network Security: Fundamentals, Technologies, and Trends

  • Abstract:
  • Network virtualization is the technique that allows the creation of multiple instances of virtual networks over a single physical substrate. This technique has attracted a great amount of interest from the academic community, in particular due to its applicability in creating favorable conditions for the evaluation of new architectures for the Future Internet. There is also noticeable interest from major companies within the segment of computer networks, which are starting to manufacture devices supporting virtualization, as well as offering virtual network provisioning services. However, the shared use of routing devices and communication channels brings a series of security related concerns. It is necessary to provide protection to virtual network infrastructures in order to enable their use in real environments and on a large scale. In this tutorial, we will present a review of the state-of-the-art concerning virtual network security. We will discuss the main challenges related to this kind of environment, some of the major threats, as well as solutions proposed in the literature which aim to deal with varying security aspects.

  • Authors:
  • - Luciano Paschoal Gaspary (UFRGS)
    - Marinho Barcellos (UFRGS)
    - Edmundo Madeira (UNICAMP)
    - Nelson Fonseca (UNICAMP)
    - Leonardo Bays (UFRGS)
    - Rodrigo Ruas Oliveira (UFRGS)

SC 3 - Survey on Control and Monitoring Frameworks for Experimental Networks Testbeds

  • Abstract:
  • In recent years, the computer network research field has largely focused efforts on innovations that follow two main approaches, clean-slate or incremental deployment, as possible proposals for the Future Internet. However, in order to be considered deployable, new ideas need rigorous research on a large scale, using a variety of available tools such as simulators, emulators and experimentation networks (testbeds). Within the context of these experimentation infrastructures, there has been advances in creating complex architectures that can control and monitor experiments, enabling inexperienced experimenters to test new ideas with minimal effort. The control and monitoring frameworks, a.k.a. CMFs, among other things, allow that network and computational resources are registered as assets to be used, register validated experimenters to control access to the infrastructure, create slices and record resource utilization, manage experiments lifecycle, monitor results, store the results, etc. The objective of this course is to provide a survey of these CMF systems. Thus, we provide a general introduction to the problem, provide details about typical CMF requirements, and explore details about implementations, deployments and test recipes of several experimental testbeds and CMFs around the world. We also discuss recent efforts to federalize these experimentation infrastructures in order to have larger and heterogeneous infrastructure to conduct experiments. Finally, the course is focused on descriptive presentations of these CMFs, and we detail CMF concepts and services in general, and later exemplifying using specific CMF/federalization schemes such as PlanetLab, Emulab / Protogenes, OFELIA, OMF and SFA 2. We intend to present videos and practical examples of how to use this technology, bringing attention to Brazilian researchers of how-to conduct experiments with these frameworks.

  • Authors:
  • - Cesar Augusto Cavalheiro Marcondes (UFSCAR)
    - Joberto Martins (UNIFACS)
    - Augusto Suruagy Monteiro (UNIFACS)
    - Kleber Cardoso (UFG)
    - Antônio Jorge G. Abelém (UFPA)
    - Vagner Nascimento (UFPA)
    - Iara Machado (RNP)
    - Tereza Carvalho (USP)
    - Christian Miers (USP)
    - Marcos Salvador (CPqD)
    - Christian Esteve Rothenberg (CPqD)

SC 4 - Software Defined Networks: a systems approach to research in Computer Networks

  • Abstract:
  • Software Defined Networks (SDN) constitute a new paradigm for research in Computer Networks which has been getting attention from a large fraction of the area's community both in academy and elsewhere. Much of the attention so far has focused on OpenFlow, one of the elements that made SDN possible. However, Software Defined Networks go beyond OpenFlow, creating new options in terms of abstractions, control environments and network applications, which can be build in a more straightforward way, free from limitations dictated by current network technologies. This short course intends to take a systems approach to the topic, covering both theory and practice aspects. In terms of theory, we intend to discuss the multiple components of a Software Defined Network System, including solutions for the virtualization of network elements, network operating systems and new applications, as well as the challenges posed by this new paradigm, and the various ongoing research efforts, in Brazil and around the world. To illustrate the new development possibilities offered by SDN, a part of the course will focus on the POX network operating system, developed with teaching and research in mind. We will present its structure and its main programming principles, as well as some examples of applications that can be easily built in that environment.

  • Authors:
  • - Dorgival Guedes (UFMG)
    - Luiz Filipe Vieira (UFMG)
    - Marcos Vieira (UFMG)
    - Henrique Rodrigues (UFMG)
    - Rogério Vinhal Nunes (UFMG)

SC 5 - Content-Oriented Networks: A New Paradigm to the Internet

  • Abstract:
  • Content-Oriented Networks (CONs) are a new communication paradigm to the Internet. This new paradigm focuses on the content delivery to users regardless of the location of the content rather than the current Internet architecture that focuses on the communication between end systems. In CONs, the network infrastructure also actively contributes to content caching and distribution. CONs are based on the success of Internet applications that are content-oriented in nature. CONs are based on the success of Internet applications that are content-oriented in nature, i.e., for users the most important is to receive the content no matter who sends it or where the content is stored. Thus, the main advantage of this new paradigm is to increase the efficiency of content search and delivery and also the content availability. Furthermore, CONs simplifies the solution of current Internet problems, such as mobility and content security. The main goals of this short course are to present the basic concepts of CONs, to describe the main architecture proposals for these networks and to discuss the main challenges for its development. The challenges include routing, naming, and caching on the network-core elements and also several aspects of content security, users' privacy, and issues to implement CONs in practice.

  • Authors:
  • - Gabriel de Brito (UFF)
    - Pedro Velloso (UFF)
    - Igor Moraes (UFF)

Promotion

Organization

Sbc
Larc
Dcc
Ufmg

Support

Ufop
Inweb
CNPq

Sponsors

cgi.br
nic.br
Capes
Ministerio da Educacao
Google
Engetron
Hp
Globo.com
Ativas
Petrobras
CPqD
NET
Intel
Fapemig
Technicolor