PERFORMANCE OF SHARED-MEMORY SWITCHES UNDER MULTICAST BURSTY TRAFFIC

Citation
Fm. Chiussi et al., PERFORMANCE OF SHARED-MEMORY SWITCHES UNDER MULTICAST BURSTY TRAFFIC, IEEE journal on selected areas in communications, 15(3), 1997, pp. 473-487
Citations number
19
Categorie Soggetti
Telecommunications,"Engineering, Eletrical & Electronic
ISSN journal
07338716
Volume
15
Issue
3
Year of publication
1997
Pages
473 - 487
Database
ISI
SICI code
0733-8716(1997)15:3<473:POSSUM>2.0.ZU;2-Y
Abstract
In this paper, we study shared-memory switches under multicast bursty traffic and characterize the relation between their performance and th e multicast distribution that defines the mix of multicast traffic arr iving at the switches, We consider two schemes that have been used in practical realizations of these switches to replicate multicast cells: 1) replication-at-receiving (RAR), where multiple copies of a multica st cell are stored in the buffer and served independently, and 2) repl ication-at-sending (RAS), where a single instance of a multicast cell is stored in the buffer, and the cell is replicated as it is transmitt ed to the output ports, For each scheme, we study two configurations: 1) the shared-memory-only (SMO) configuration, where the bandwidth of the replication mechanism is sufficient to accommodate even the worst- case replication requirements, and 2) the shared-memory-with-replicati on-first-in-first-out (SM+RFIFO) configuration, where the bandwidth of the replication mechanism is lower than what required by the worst ca se, and thus an additional buffer is used in front of the shared memor y to temporarily store cells while they are replicated, For all cases, using simulation, we find upper bounds for the buffer requirements to achieve a desired cell-loss rate, We show that these upper bounds are significantly larger than the buffer requirements under unicast traff ic and are approached even for very small volumes of multicast traffic ; thus, these upper bounds are what should be used in practice to size the buffers to achieve desired performance under traffic with general multicast distributions, We also study shared-memory switches with ou tput demultiplexers and characterize and compare the different multica sting schemes that are used in these switches.