INVESTIGADORES
ZUNINO SUAREZ Alejandro Octavio
artículos
Título:
A Peer-to-Peer Communication Infrastructure for Groupware Applications (Indexed SCI, JCR2008 IF=0.714)
Autor/es:
GOTTHELF, P.; ZUNINO, A.; CAMPO, M.
Revista:
INTERNATIONAL JOURNAL OF COOPERATIVE INFORMATION SYSTEMS
Editorial:
World Scientific
Referencias:
Año: 2008 vol. 17 p. 523 - 554
ISSN:
0218-8430
Resumen:
Many advances have been done to allow groups of people to work together and collaborate in the Internet. Collaborative systems are characterized by the way participants interact. In many cases equal standing members should cooperate in a non authoritative environment, where no entity or authority is or should be in charge of regulating the group. Therefore, decentralized communication infrastructures have been hailed as promising alternatives. Recently, decentralized infrastructures based on P2P approaches have drawn the attention of the research community, because of their benefits in terms of scalability, robustness, availability and potentials for leveraging computational resources distributed across the Internet. In this paper, a scalable peer-to-peer (P2P) communication Infrastructure for groupware applications is presented. It enables a large number of people to join and cooperate in a robust, decentralized and easy deployable way, without requiring high capacity servers or any other special network infrastructure. The communication infrastructure is based on a binary tree as overlay structure, which implements all groupware communication functionality, including membership management and packet forwarding, at application level, making it an inexpensive and fast deployable solution for equal standing members, such as home users with a domestic connection to the Internet. Two applications, one for synchronous groupware and other for asynchronous collaboration, have been developed to validate the approach. Comparisons with other communication infrastructures in aspects such as end to end propagation delay, group latency, throughput, protocol overhead, failure recovery and link stress, show that our approach is a scalable and robust alternative.