CLEI 2009

Social Program + Program

- PAINEL: Interiorização da Tecnologia: O Papel dos Parques Tecnológicos

- Social Program CLEI 2009

- Sessão de Abertura

- Transfers schedule [Pdf]

Program

- General Program

* CLEI 2009 - General Program (PDF file, 322kb) *

- Technical Sessions Program

- CLEI 2009 (PDF file, 532Kb)

- CIESC 2009 (PDF file, 139kb)

- CLTM 2009 (PDF file, 122kb)

- LAWC 2009 (PDF file, 39kb)

TUTORIAL CLEI A
TÍTULO: Computación Cuántica. Una ventana al futuro.
MINISTRANTE Prof. Dr. Benjamín Barán (Universidad Nacional de Asunción, Paraguai)

RESUMEN: La tecnología de semiconductores ha permitido impresionantes avances tecnológicos, anticipados por la Ley de Moore, pero ya está llegando a su límite teórico, por lo que se requiere de una nueva tecnología para sostener el nivel de crecimiento de los últimos años. La computación cuántica se presenta como la mejor alternativa para satisfacer este requerimiento, unificando dos de las áreas de mayor desarrollo del siglo XX, la Mecánica Cuántica y la Computación. Esta nueva tecnología muestra tener un potencial extraordinario al posibilitar resolver problemas considerados hasta ahora intratables, potencialmente casi sin gastar energía, logrando incluso demostrar la plausibilidad de fenómenos que hasta ahora solo veíamos en películas de ciencia a ficción, como la teletransportación.

TUTORIAL CLEI B
TÍTULO: Virtualização: o que é e para que serve
MINISTRANTE: Alexandre da Silva Carissimi (Universidade Federal do Rio Grande do Sul, Brasil)

Virtualização é a técnica que permite particionar um único sistema computacional em vários outros denominados de máquinas virtuais. Cada máquina virtual oferece um ambiente completo muito similar a aquele oferecido por uma máquina física. Com isso, cada máquina virtual pode ter seu próprio sistema operacional, aplicativos e serviços de rede (protocolos Internet). As máquinas virtuais, por emularem um ambiente computacional sobre outro, impõem algumas restrições de implementação e de desempenho. É nesse ponto que, dada a importância e a gama de aplicações em que a virtualização pode ser empregada, se observa um investimento maciço nesta tecnologia por parte de fabricantes de hardware e de software. Os processadores mais recentes da Intel e da AMD possuem mecanismos e soluções de hardware especialmente destinados a oferecerem suporte a virtualização. Em nível de software, se encontram diversos produtos e ambientes, onde se pode destacar as soluções de virtualização da Microsoft, da VMware e Citrix (Xen Source).

O tutorial proposto tem como objetivo capacitar o participante a entender o que é a virtualização e as vantagens que sua utilização pode trazer para uma infra-estrutura de TI. Serão abordadas as principais técnicas de virtualização, o suporte de hardware fornecido pelos processadores atuais, as principais ferramentas e ambientes existentes e sua aplicação em ambientes de ensino, como infra-estrutura de TI e em aplicações em processamento paralelo como grids e clusters.

TUTORIAL LANC A
TITLE: Signaling protocols for IP networks
MINISTRANTE: Edmundo Monteiro (Universidade de Coimbra, Portugal)

Abstract:
Signaling can be defined as the exchange of data needed among network elements and end systems to support the setup, maintenance, and release of the network services. In IP-based networks, signaling is needed at the application, transport, and network levels. At the application layer, signaling is needed to manage application sessions (e.g. VoIP or videoconferencing session) and to negotiate service parameters between the applications. At the transport level, signaling is needed to set up and tear down connections between the end systems, and to perform flow, congestion, error and synchronization control in the active connections. Signaling also is needed at the network level to support mobility or for dynamic discrimination of flows (or flow aggregates) to give these flows a treatment different from the normal flows (e.g., different route, priority, delay, or security treatment).

Signaling always has been a controversial issue when the Internet is concerned. Ideally, signaling on the Internet should be reduced to a minimum and should be performed by end-systems, to keep the network as simple as possible. Reality has shown that this is not possible. Although the current Internet is still data-driven — as opposed to the signaling-driven nature of, for instance, the telephone network — signaling is present in virtually all its components, for network operations support, quality of service support, management, and application/user support.

IP companion protocols such as ICMP (Internet Control Message Protocol), ARP (Address Resolution Protocol) and DHCP (Dynamic Host Configuration Protocol) perform network operation signaling. Routing protocols such as OSPF (Open Shortest Path First) and BGP (Border Gateway Protocol) can carry signaling information used for network operations signaling. RSVP (Resource Reservation Protocol), MPLS (Multi-Protocol Label Switching) signaling extensions and NSIS (Next Steps in Signaling) are examples of signaling protocols that support quality of service and resource control. SNMP (Simple Network Management Protocol) and COPS (Common Open Policy Service) are examples of management protocols. SIP (Session Initiation Protocol) and H.323 are examples of application/user support signaling protocols. Session and transport protocols such as TCP (Transmission Control Protocol), RTSP (Real Time Streaming Protocol) and RTCP (Real Time Control Protocol) also include signaling mechanisms for application/user support, to ensure end-to-end connectivity with flow, error and synchronization control.

All of the above protocols and mechanisms have the common objective of providing some form of control and support of user traffic and communication services. Thus, they contribute to the good operation of the network. Signaling mechanisms can be on-path or off-path. When signaling messages follow the same path as data messages, it is said that on-path signaling is being performed (also mentioned as path-coupled or in-band signaling). Sometimes, however, entities that are not on the data path have to be signaled. Signaling protocols that allow the signaling of entities that are not on the data path are called off-path signaling protocols (also mentioned as path-decoupled or out-band signaling). Hybrid signaling solutions using on-path signaling inside each domain and off-path signaling between domains also are possible.

Signaling also can be explicit or implicit. Explicit signaling is achieved though explicit messages that can be carried together with user data, in specific packet fields, as in TCP header fields, or in separate signaling packets, as in RSVP and NSIS. Implicit signaling information can be extracted from the normal application flow to trigger reservation mechanisms in network devices. Some signaling solutions combine implicit signaling near the end-systems (to reduce the complexity of end-systems) and explicit signaling inside the network.

Signaling is closely related to the amount of state information stored in the network elements. State information can be stored per flow or per flow aggregates. State information can be maintained until explicitly released (hard-state) or expire after a given time limit if not refreshed before (soft-state). If state information is needed per flow, signaling also has to be performed on a per-flow basis; otherwise signaling is needed only to manage the flow aggregates. In soft-state approaches, signaling has to be issued periodically to refresh the state information. In hard-state approaches signaling is needed only for the establishment and release of state information.

The definition of a signaling scheme is a trade-off between the application needs and the complexity introduced in the network. Best effort applications need basic signaling mechanisms for flow, error, and congestion control like the ones included in TCP only. Multimedia applications, with quality of service requirements, can operate smoothly if the network is over-provisioned. In a limited resource environment they will require per flow of aggregated signaling to support resource reservation and traffic differentiation mechanisms. Thus, there is also a trade-off between the amount of signaling and network resources. Overall, the definition of the signaling mechanisms for the support of real-time applications over the global Internet with quality and security is still an open issue. The solution is expected to emerge from the signaling proposals being discussed and standardized at the Internet Engineering Task Force (IETF) and settlement of the above trade-offs.

This tutorial describes the main signaling protocols and mechanisms for IP networks. The focus is on signaling protocols for network operation and quality of service support. Signaling mechanisms embedded in routing and transport protocols will also be briefly mentioned. For each signaling protocol described, the main advantages, limitations, and deployment trade-offs are discussed. The tutorial will include the discussion of the signaling protocols currently under standardization in the IETF, ITU-T and other standardization bodies, and the discussion of the signaling approaches recently developed in the scope of international projects like EuQoS (End-to-End Quality of Service over Heterogeneous Networks), WEIRD (WiMax Extensions to Isolated Research Data networks), and Q3M (QoS for Mobile Multimedia Multiuser).

TUTORIAL LANC B
TITULO: Metaheurísticas y Optimización Combinatoria
MINISTRANTE: Pablo Rodríguez Bocca (Universidad de la República, UruguaI)

ABSTRACT:
El tutorial presenta a las metaheurísticas como mecanismo para la resolución de problemas de optimización combinatoria. De forma auto-contenida, se comienza presentando que son los problemas de optimización combinatoria y porque la importancia de poder resolverlos con métodos heurísticos. Se definen y presentan las heurísticas de mayor difusión, como por ejemplo: Búsqueda local (LS-Local Search), Recocido simulado (SA-Simulated Annealing), Búsqueda Tabú (TS-Tabu search), GRASP (Greedy Randomized Adaptive Search Procedure), Computación evolutiva (EC-Evolutionary Computation) y Colonias de Hormigas (ANT-Ant colony optimization). Para cada meta heurística se presentan sus principales características, su pseudo-código y su aplicación en ejemplos prácticos. Luego de presentadas todas las metaheurísticas se realiza una síntesis y clasificación de ellas, con la intención de generar una base para que el estudiante pueda discernir que método usar para cada caso específico. Se presentará también un ejemplo de aplicación a un problema de telecomunicaciones.

PAINEL CLEI
TITULO: Historia de la Informática en Latinoamérica

MOTIVACIÓN: El vertiginoso desarrollo de la Informática a generado también el Latinoamérica un importante progreso en las últimas décadas, pero poco se ha escrito de esta historia. Sus principales precursores y actores están llegando a una etapa madura de su vida profesional por lo que es menester dejar un testimonio vivo de los caminos transitados en las últimas décadas para que la Informática Latinoamericana llegue a su sitial actual, una historia a ser contada por sus mismos protagonistas.

MODERADOR: Benjamín Barán (Universidad Nacional de Asunción, Paraguai)

PANELISTAS: (em construção)
1- Jorge Vidart
2- Raúl Carnota
3- Jorge Aguirre
4- Ernesto Cuadros-Vargas (Universidad Católica San Pablo, Perú)
5- Philippe Navaux (Universidade Federal do Rio Grande do Sul, Brasil)