On 1-aug-04, at 16:47, Pekka Savola wrote:
if folks thought multicast on WLAN had such
big problems, they might have devised a different mechanism for it -- for
example, if pure broadcast does not suffer from this issue, certain
multicast addresses, which are expected to have a larger number of
recipients, could be mapped to the broadcast address instead.
Sorry to jump back more than two months in the past, but I think this is an important issue and it doesn't look like it has been resolved in later discussions.
I've been using wireless at home (where I do most of my work) for more than a year now, using Apple equipment, and I haven't experienced any problems with ND. Apple also extensively uses multicasts for service discovery through multicast DNS / zeroconf, I don't remember seeing problem with that either.
I've searched through "802.11 Wireless Networks: The Definitive Guide" and although it doesn't say so specifically, it's pretty clear that broadcasts and multicasts are treated the same in 802.11. This means no acknowledgments and no fragmentation. Which in turn means that broad/multicast packets are delivered less reliably than unicast packets. (So this goes for IPv4 ARP too.)
However, the difference isn't as big as it may seem at first glance. With unicast, transmissions initially happen at a high speed, and if that doesn't work, they're resent at a lower speed. Since non-unicast packets must be received by all stations, they are generally transmitted at 1 or 2 Mbps, which conforms to the original 802.11 (pre-b) standard. Because of this, the chance of a non-multicast packet actually making it through to a given host are quite good, given even low but reasonable radio coverage.
. dnsop resources:_____________________________________________________ web user interface: http://darkwing.uoregon.edu/~llynch/dnsop.html mhonarc archive: http://darkwing.uoregon.edu/~llynch/dnsop/index.html
