David A. Selby
IBM J. Res. Dev
Traditional broadcast protocols are inappropriate for the high-speed networks of the future. Such protocols are limited by the speed of software processing, which becomes a bottleneck as network speeds increase. This paper presents broadcast protocols that are appropriate for high-speed networks, and are tolerant of failures involving the loss of messages. The protocols are based primarily on the simple hardware functions present in a high-speed network node. This leads to message delivery at hardware speeds. In the unlikely event of a failure, software intervention is required to guarantee the timely termination of the protocol; however, this software processing does not interfere with message delivery.
David A. Selby
IBM J. Res. Dev
Pradip Bose
VTS 1998
Michael C. McCord, Violetta Cavalli-Sforza
ACL 2007
S.F. Fan, W.B. Yun, et al.
Proceedings of SPIE 1989