2009/7/1 Victor lyamtsev <vlyamt...@gmail.com>: > Is there an analythical dependency for end-to-end unicast udp packet > loss as function of bandwidth utilization?
Errr.. no? > I am trying to find books or papers describing it as clear-cut math > expression. Every piece of hardware along the route is free to drop packets depending on the burstiness of the traffic (including unrelated traffic), the size of its queues, the load if its processor, the software and hardware bugs it has, how good it is at coping with a poor physical signal, etc. And that is only for one piece of hardware. I am not sure what you are looking for... a simplified model of the above, like for instance modeling only burstiness and buffer sizes? Cheers, Marc ------------------------------------------------------------------------------ _______________________________________________ Iperf-users mailing list Iperf-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/iperf-users
