By Topic

A class of interconnection networks for multicasting

Sign In

Cookies must be enabled to login.After enabling cookies , please use refresh or reload or ctrl+f5 on the browser for the login options.

Formats Non-Member Member
$31 $13
Learn how you can qualify for the best price for this item!
Become an IEEE Member or Subscribe to
IEEE Xplore for exclusive pricing!
close button

puzzle piece

IEEE membership options for an individual and IEEE Xplore subscriptions for an organization offer the most affordable access to essential journal articles, conference papers, standards, eBooks, and eLearning courses.

Learn more about:

IEEE membership

IEEE Xplore subscriptions

1 Author(s)
Yuanyuan Yang ; Dept. of Comput. Sci. & Electr. Eng., Vermont Univ., Burlington, VT, USA

Multicast or one-to-many communications arise frequently in parallel computing applications and in other communication environments. Multicast networks can simultaneously support multiple multicast connections between the network inputs and network outputs. However, due to the much more complex communication patterns and routing control in multicast networks, there is still a considerably large gap in network cost between even the currently best known multicast networks and permutation networks. In this paper, we will present a class of interconnection networks which can support a substantial amount of well-defined multiple multicast connections in a nonblocking fashion and yet have a comparable low cost to permutation networks. We also provide an efficient routing algorithm for satisfying multicast connection requests in such networks. Moreover, the multicast connecting capability of the networks is represented as a function of fundamental network structural parameters, so that the trade-off between the network multicast capability and the network cost can be determined. This enables different system designers to choose the multicast networks which fit their particular application needs. By utilizing a network with well-defined multicast capability in a parallel computing system, software or algorithm designers of the system can make full use of the multicast capability provided by the network, and substantial improvements in the performance of the system can be achieved due to significantly shortened delays in data transfer and simplified synchronization mechanisms for shared data

Published in:

Parallel Processing Symposium, 1996., Proceedings of IPPS '96, The 10th International

Date of Conference:

15-19 Apr 1996